example_id
int64 0
10k
| metadata
stringlengths 679
723
| classification_prompt
stringlengths 4.23k
10.2k
| classification_completion
stringclasses 14
values | classification_text
stringlengths 4.24k
10.2k
| improved_signature
stringlengths 2.11k
4.25k
| improved_model_weights
stringlengths 1.76k
5.04k
| training_metrics
stringlengths 1.46k
2.92k
|
|---|---|---|---|---|---|---|---|
0
|
{"target_pattern": "palindrome", "degraded_accuracy": 0.48, "improved_accuracy": 0.98, "improvement": 0.5, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 2679, "learning_rate": 0.03008896643339405, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "palindrome", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["palindrome"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.453386,
-0.327474,
-0.157223,
0.456369,
0.643735
],
[
0.12622,
0.429354,
-0.240399,
-0.393445,
-0.106389
],
[
-0.11147,
-0.549566,
-0.162662,
-0.156643,
-0.175201
],
[
-0.363142,
-0.069774,
0.116957,
-0.013997,
0.706692
],
[
0.377102,
0.305662,
-0.360224,
-0.258711,
0.050089
],
[
-0.161948,
-0.358909,
-0.015153,
0.437877,
0.445758
],
[
-0.109331,
0.033178,
0.324673,
0.193858,
-0.184097
]
],
"network.0.bias": [
-0.119286,
0.111422,
-0.252262,
-0.234887,
0.168684,
-0.219954,
-0.279471
],
"network.2.weight": [
[
-0.285712,
0.218869,
0.205375,
0.076367,
-0.227966,
0.170337,
-0.164336
],
[
-0.062748,
-0.307176,
0.154941,
0.005569,
-0.186357,
-0.07184,
-0.157229
],
[
0.522343,
0.915817,
-0.046493,
0.489848,
0.243372,
0.375761,
0.044614
],
[
0.671219,
0.645983,
-0.024659,
0.565528,
0.512602,
0.790214,
-0.029846
],
[
0.319858,
-0.007872,
0.052752,
-0.024128,
-0.173129,
-0.322801,
-0.025422
],
[
0.805854,
0.588786,
-0.090574,
0.477788,
0.234788,
0.278425,
0.275044
],
[
-0.332876,
-0.43385,
0.147804,
-0.071303,
-0.258455,
0.059788,
-0.009553
]
],
"network.2.bias": [
-0.392752,
-0.295732,
-0.104868,
0.14474,
0.281554,
0.159095,
-0.174762
],
"network.4.weight": [
[
-0.176771,
-0.333024,
0.72132,
0.529057,
-0.276342,
0.033306,
-0.130475
],
[
0.133662,
-0.023771,
-0.371075,
-0.144869,
-0.441829,
-0.535646,
0.239691
],
[
-0.12163,
0.082703,
0.340026,
0.6903,
0.213804,
0.487876,
0.013686
],
[
-0.17301,
0.314767,
0.473747,
0.499127,
-0.179499,
0.583916,
0.133916
],
[
0.245098,
-0.042917,
0.649938,
0.641373,
0.14161,
0.804551,
-0.190475
],
[
-0.115826,
0.179191,
-0.505833,
-0.006435,
0.52513,
0.347838,
-0.443804
],
[
0.214556,
-0.099233,
0.543852,
0.742082,
-0.118269,
0.436566,
0.286135
]
],
"network.4.bias": [
-0.125316,
-0.00294,
0.225475,
-0.146175,
0.316137,
0.504589,
-0.210445
],
"network.6.weight": [
[
0.755015,
0.042379,
0.347016,
0.36354,
0.399971,
-0.512195,
0.601143
],
[
0.592505,
-0.343272,
0.045308,
0.494628,
0.268346,
-0.252532,
0.414568
],
[
0.490921,
-0.438342,
0.547926,
0.670926,
0.753076,
-0.481932,
0.683054
],
[
0.208314,
0.19906,
0.072818,
-0.248866,
-0.25011,
0.247131,
-0.218774
],
[
-0.243133,
0.162341,
-0.238087,
-0.144334,
-0.205474,
0.03245,
-0.303879
],
[
-0.394051,
0.140899,
0.191306,
0.088856,
0.310395,
0.342127,
-0.121957
],
[
-0.376836,
-0.030551,
-0.281656,
-0.307566,
-0.361284,
-0.109558,
0.096992
]
],
"network.6.bias": [
-0.012644,
-0.213544,
-0.130974,
-0.095416,
-0.13086,
0.729133,
-0.296097
],
"network.8.weight": [
[
-0.359781,
-0.009095,
0.122767,
-0.071826,
-0.042486,
0.199495,
0.034381
],
[
-0.312453,
-0.210575,
0.036223,
-0.326399,
-0.22013,
0.150529,
0.11431
],
[
0.224373,
-0.223163,
-0.015183,
0.019742,
0.011242,
0.790907,
-0.353785
],
[
-0.177242,
0.311048,
-0.316753,
-0.029666,
-0.343115,
0.01938,
-0.170107
],
[
-0.175515,
-0.025661,
0.180442,
-0.259101,
-0.209449,
0.497129,
-0.287058
],
[
-0.101743,
0.241576,
-0.424482,
-0.315128,
-0.005221,
-0.282459,
0.165427
],
[
0.386431,
0.3102,
0.760263,
-0.038546,
0.298923,
-0.616757,
0.024728
]
],
"network.8.bias": [
-0.363764,
-0.307377,
0.664011,
-0.243452,
0.525364,
0.03965,
0.129556
],
"network.10.weight": [
[
0.377958,
-0.348164,
0.113177,
0.209938,
-0.323088,
0.041111,
-0.149619
],
[
-0.072122,
-0.010595,
-0.398248,
0.259432,
0.090957,
-0.202259,
0.56982
],
[
-0.138257,
-0.17486,
0.320924,
0.0328,
-0.325217,
-0.25616,
-0.270992
],
[
0.282022,
-0.404823,
0.296393,
0.023418,
0.125639,
0.143354,
-0.262961
],
[
0.251087,
0.117399,
0.675095,
-0.372959,
0.520469,
0.145071,
-0.265112
],
[
0.303878,
0.421899,
-0.097009,
0.103215,
-0.56897,
0.093587,
0.510254
],
[
-0.097669,
0.530755,
-0.629361,
0.212001,
-0.12204,
0.008331,
0.03747
]
],
"network.10.bias": [
-0.306128,
0.133168,
-0.220066,
0.790848,
0.656986,
0.121512,
0.018336
],
"network.12.weight": [
[
-0.350767,
-0.653457,
-0.251655,
0.587816,
0.799047,
-0.589768,
0.133438
]
],
"network.12.bias": [
0.135511
]
}
## Activation Signature
### 0
mean: [1.292375, -0.426227, -2.284186, 0.271942, -0.050532, 0.381618, 0.511803]
std: [1.808385, 1.091531, 1.430994, 1.422668, 1.139070, 1.402958, 0.757621]
### 2
mean: [-0.779943, -0.685135, 1.601155, 2.425416, 0.404210, 2.259518, -0.881298]
std: [0.373547, 0.277124, 1.567902, 2.251955, 0.292369, 1.894588, 0.641896]
### 4
mean: [2.277668, -2.338019, 3.633575, 3.070626, 4.788693, 0.676361, 3.399807]
std: [2.332649, 1.998400, 3.041038, 2.930285, 3.998450, 0.225052, 3.319076]
### 6
mean: [7.700427, 5.345847, 10.643229, -1.894065, -3.989255, 2.100991, -4.598248]
std: [7.498685, 5.427755, 10.076451, 1.763134, 3.543395, 0.763215, 3.749759]
### 8
mean: [-1.457151, -3.138964, 2.697130, -3.273600, 2.001390, -4.561805, 11.561800]
std: [1.363895, 3.008224, 0.916671, 2.820644, 0.740204, 3.940965, 11.782944]
### 10
mean: [-2.380364, 5.840655, -3.143967, -1.203863, 0.448972, 4.630827, -1.489409]
std: [1.893654, 6.409157, 3.134876, 2.733996, 2.134031, 5.499527, 0.243057]
### 12
mean: [-5.419362]
std: [8.166819]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
palindrome
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.453386,
-0.327474,
-0.157223,
0.456369,
0.643735
],
[
0.12622,
0.429354,
-0.240399,
-0.393445,
-0.106389
],
[
-0.11147,
-0.549566,
-0.162662,
-0.156643,
-0.175201
],
[
-0.363142,
-0.069774,
0.116957,
-0.013997,
0.706692
],
[
0.377102,
0.305662,
-0.360224,
-0.258711,
0.050089
],
[
-0.161948,
-0.358909,
-0.015153,
0.437877,
0.445758
],
[
-0.109331,
0.033178,
0.324673,
0.193858,
-0.184097
]
],
"network.0.bias": [
-0.119286,
0.111422,
-0.252262,
-0.234887,
0.168684,
-0.219954,
-0.279471
],
"network.2.weight": [
[
-0.285712,
0.218869,
0.205375,
0.076367,
-0.227966,
0.170337,
-0.164336
],
[
-0.062748,
-0.307176,
0.154941,
0.005569,
-0.186357,
-0.07184,
-0.157229
],
[
0.522343,
0.915817,
-0.046493,
0.489848,
0.243372,
0.375761,
0.044614
],
[
0.671219,
0.645983,
-0.024659,
0.565528,
0.512602,
0.790214,
-0.029846
],
[
0.319858,
-0.007872,
0.052752,
-0.024128,
-0.173129,
-0.322801,
-0.025422
],
[
0.805854,
0.588786,
-0.090574,
0.477788,
0.234788,
0.278425,
0.275044
],
[
-0.332876,
-0.43385,
0.147804,
-0.071303,
-0.258455,
0.059788,
-0.009553
]
],
"network.2.bias": [
-0.392752,
-0.295732,
-0.104868,
0.14474,
0.281554,
0.159095,
-0.174762
],
"network.4.weight": [
[
-0.176771,
-0.333024,
0.72132,
0.529057,
-0.276342,
0.033306,
-0.130475
],
[
0.133662,
-0.023771,
-0.371075,
-0.144869,
-0.441829,
-0.535646,
0.239691
],
[
-0.12163,
0.082703,
0.340026,
0.6903,
0.213804,
0.487876,
0.013686
],
[
-0.17301,
0.314767,
0.473747,
0.499127,
-0.179499,
0.583916,
0.133916
],
[
0.245098,
-0.042917,
0.649938,
0.641373,
0.14161,
0.804551,
-0.190475
],
[
-0.115826,
0.179191,
-0.505833,
-0.006435,
0.52513,
0.347838,
-0.443804
],
[
0.214556,
-0.099233,
0.543852,
0.742082,
-0.118269,
0.436566,
0.286135
]
],
"network.4.bias": [
-0.125316,
-0.00294,
0.225475,
-0.146175,
0.316137,
0.504589,
-0.210445
],
"network.6.weight": [
[
0.755015,
0.042379,
0.347016,
0.36354,
0.399971,
-0.512195,
0.601143
],
[
0.592505,
-0.343272,
0.045308,
0.494628,
0.268346,
-0.252532,
0.414568
],
[
0.490921,
-0.438342,
0.547926,
0.670926,
0.753076,
-0.481932,
0.683054
],
[
0.208314,
0.19906,
0.072818,
-0.248866,
-0.25011,
0.247131,
-0.218774
],
[
-0.243133,
0.162341,
-0.238087,
-0.144334,
-0.205474,
0.03245,
-0.303879
],
[
-0.394051,
0.140899,
0.191306,
0.088856,
0.310395,
0.342127,
-0.121957
],
[
-0.376836,
-0.030551,
-0.281656,
-0.307566,
-0.361284,
-0.109558,
0.096992
]
],
"network.6.bias": [
-0.012644,
-0.213544,
-0.130974,
-0.095416,
-0.13086,
0.729133,
-0.296097
],
"network.8.weight": [
[
-0.359781,
-0.009095,
0.122767,
-0.071826,
-0.042486,
0.199495,
0.034381
],
[
-0.312453,
-0.210575,
0.036223,
-0.326399,
-0.22013,
0.150529,
0.11431
],
[
0.224373,
-0.223163,
-0.015183,
0.019742,
0.011242,
0.790907,
-0.353785
],
[
-0.177242,
0.311048,
-0.316753,
-0.029666,
-0.343115,
0.01938,
-0.170107
],
[
-0.175515,
-0.025661,
0.180442,
-0.259101,
-0.209449,
0.497129,
-0.287058
],
[
-0.101743,
0.241576,
-0.424482,
-0.315128,
-0.005221,
-0.282459,
0.165427
],
[
0.386431,
0.3102,
0.760263,
-0.038546,
0.298923,
-0.616757,
0.024728
]
],
"network.8.bias": [
-0.363764,
-0.307377,
0.664011,
-0.243452,
0.525364,
0.03965,
0.129556
],
"network.10.weight": [
[
0.377958,
-0.348164,
0.113177,
0.209938,
-0.323088,
0.041111,
-0.149619
],
[
-0.072122,
-0.010595,
-0.398248,
0.259432,
0.090957,
-0.202259,
0.56982
],
[
-0.138257,
-0.17486,
0.320924,
0.0328,
-0.325217,
-0.25616,
-0.270992
],
[
0.282022,
-0.404823,
0.296393,
0.023418,
0.125639,
0.143354,
-0.262961
],
[
0.251087,
0.117399,
0.675095,
-0.372959,
0.520469,
0.145071,
-0.265112
],
[
0.303878,
0.421899,
-0.097009,
0.103215,
-0.56897,
0.093587,
0.510254
],
[
-0.097669,
0.530755,
-0.629361,
0.212001,
-0.12204,
0.008331,
0.03747
]
],
"network.10.bias": [
-0.306128,
0.133168,
-0.220066,
0.790848,
0.656986,
0.121512,
0.018336
],
"network.12.weight": [
[
-0.350767,
-0.653457,
-0.251655,
0.587816,
0.799047,
-0.589768,
0.133438
]
],
"network.12.bias": [
0.135511
]
}
## Activation Signature
### 0
mean: [1.292375, -0.426227, -2.284186, 0.271942, -0.050532, 0.381618, 0.511803]
std: [1.808385, 1.091531, 1.430994, 1.422668, 1.139070, 1.402958, 0.757621]
### 2
mean: [-0.779943, -0.685135, 1.601155, 2.425416, 0.404210, 2.259518, -0.881298]
std: [0.373547, 0.277124, 1.567902, 2.251955, 0.292369, 1.894588, 0.641896]
### 4
mean: [2.277668, -2.338019, 3.633575, 3.070626, 4.788693, 0.676361, 3.399807]
std: [2.332649, 1.998400, 3.041038, 2.930285, 3.998450, 0.225052, 3.319076]
### 6
mean: [7.700427, 5.345847, 10.643229, -1.894065, -3.989255, 2.100991, -4.598248]
std: [7.498685, 5.427755, 10.076451, 1.763134, 3.543395, 0.763215, 3.749759]
### 8
mean: [-1.457151, -3.138964, 2.697130, -3.273600, 2.001390, -4.561805, 11.561800]
std: [1.363895, 3.008224, 0.916671, 2.820644, 0.740204, 3.940965, 11.782944]
### 10
mean: [-2.380364, 5.840655, -3.143967, -1.203863, 0.448972, 4.630827, -1.489409]
std: [1.893654, 6.409157, 3.134876, 2.733996, 2.134031, 5.499527, 0.243057]
### 12
mean: [-5.419362]
std: [8.166819]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
palindrome
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.292374610900879, "std": 1.8083847761154175}, "1": {"mean": -0.4262273907661438, "std": 1.0915309190750122}, "2": {"mean": -2.284186363220215, "std": 1.4309940338134766}, "3": {"mean": 0.2719416618347168, "std": 1.42266845703125}, "4": {"mean": -0.05053213611245155, "std": 1.1390700340270996}, "5": {"mean": 0.3816179931163788, "std": 1.402957797050476}, "6": {"mean": 0.5118029713630676, "std": 0.7576205134391785}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.7799431681632996, "std": 0.373546838760376}, "1": {"mean": -0.685134768486023, "std": 0.2771235704421997}, "2": {"mean": 1.6011545658111572, "std": 1.5679020881652832}, "3": {"mean": 2.4254157543182373, "std": 2.2519547939300537}, "4": {"mean": 0.40420985221862793, "std": 0.29236894845962524}, "5": {"mean": 2.2595183849334717, "std": 1.8945881128311157}, "6": {"mean": -0.8812984228134155, "std": 0.6418960690498352}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 2.2776682376861572, "std": 2.332648515701294}, "1": {"mean": -2.3380191326141357, "std": 1.9984002113342285}, "2": {"mean": 3.6335747241973877, "std": 3.0410382747650146}, "3": {"mean": 3.0706264972686768, "std": 2.9302845001220703}, "4": {"mean": 4.788693428039551, "std": 3.9984498023986816}, "5": {"mean": 0.6763611435890198, "std": 0.22505198419094086}, "6": {"mean": 3.3998069763183594, "std": 3.3190762996673584}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 7.700427055358887, "std": 7.498684883117676}, "1": {"mean": 5.345847129821777, "std": 5.427755355834961}, "2": {"mean": 10.643229484558105, "std": 10.076451301574707}, "3": {"mean": -1.894065499305725, "std": 1.7631340026855469}, "4": {"mean": -3.989255428314209, "std": 3.5433950424194336}, "5": {"mean": 2.1009905338287354, "std": 0.7632152438163757}, "6": {"mean": -4.59824800491333, "std": 3.7497591972351074}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.4571512937545776, "std": 1.3638947010040283}, "1": {"mean": -3.1389644145965576, "std": 3.0082240104675293}, "2": {"mean": 2.697129726409912, "std": 0.9166706204414368}, "3": {"mean": -3.273599863052368, "std": 2.8206441402435303}, "4": {"mean": 2.001389503479004, "std": 0.7402040362358093}, "5": {"mean": -4.561805248260498, "std": 3.940964937210083}, "6": {"mean": 11.561800003051758, "std": 11.782943725585938}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -2.380363702774048, "std": 1.8936537504196167}, "1": {"mean": 5.8406548500061035, "std": 6.4091572761535645}, "2": {"mean": -3.1439666748046875, "std": 3.134876012802124}, "3": {"mean": -1.2038630247116089, "std": 2.7339963912963867}, "4": {"mean": 0.44897210597991943, "std": 2.13403058052063}, "5": {"mean": 4.6308274269104, "std": 5.499527454376221}, "6": {"mean": -1.4894088506698608, "std": 0.24305716156959534}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -5.419361591339111, "std": 8.166818618774414}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.453386, -0.327474, -0.157223, 0.456369, 0.643735], [0.12622, 0.429354, -0.240399, -0.393445, -0.106389], [-0.11147, -0.549566, -0.162662, -0.156643, -0.175201], [-0.363142, -0.069774, 0.116957, -0.013997, 0.706692], [0.377102, 0.305662, -0.360224, -0.258711, 0.050089], [-0.161948, -0.358909, -0.015153, 0.437877, 0.445758], [-0.109331, 0.033178, 0.324673, 0.193858, -0.184097]], "network.0.bias": [-0.119286, 0.111422, -0.252262, -0.234887, 0.168684, -0.219954, -0.279471], "network.2.weight": [[-0.285712, 0.218869, 0.205375, 0.076367, -0.227966, 0.170337, -0.164336], [-0.062748, -0.307176, 0.154941, 0.005569, -0.186357, -0.07184, -0.157229], [0.522343, 0.915817, -0.046493, 0.489848, 0.243372, 0.375761, 0.044614], [0.671219, 0.645983, -0.024659, 0.565528, 0.512602, 0.790214, -0.029846], [0.319858, -0.007872, 0.052752, -0.024128, -0.173129, -0.322801, -0.025422], [0.805854, 0.588786, -0.090574, 0.477788, 0.234788, 0.278425, 0.275044], [-0.332876, -0.43385, 0.147804, -0.071303, -0.258455, 0.059788, -0.009553]], "network.2.bias": [-0.392752, -0.295732, -0.104868, 0.14474, 0.281554, 0.159095, -0.174762], "network.4.weight": [[-0.176771, -0.333024, 0.72132, 0.529057, -0.276342, 0.033306, -0.130475], [0.133662, -0.023771, -0.371075, -0.144869, -0.441829, -0.535646, 0.239691], [-0.12163, 0.082703, 0.340026, 0.6903, 0.213804, 0.487876, 0.013686], [-0.17301, 0.314767, 0.473747, 0.499127, -0.179499, 0.583916, 0.133916], [0.245098, -0.042917, 0.649938, 0.641373, 0.14161, 0.804551, -0.190475], [-0.115826, 0.179191, -0.505833, -0.006435, 0.52513, 0.347838, -0.443804], [0.214556, -0.099233, 0.543852, 0.742082, -0.118269, 0.436566, 0.286135]], "network.4.bias": [-0.125316, -0.00294, 0.225475, -0.146175, 0.316137, 0.504589, -0.210445], "network.6.weight": [[0.755015, 0.042379, 0.347016, 0.36354, 0.399971, -0.512195, 0.601143], [0.592505, -0.343272, 0.045308, 0.494628, 0.268346, -0.252532, 0.414568], [0.490921, -0.438342, 0.547926, 0.670926, 0.753076, -0.481932, 0.683054], [0.208314, 0.19906, 0.072818, -0.248866, -0.25011, 0.247131, -0.218774], [-0.243133, 0.162341, -0.238087, -0.144334, -0.205474, 0.03245, -0.303879], [-0.394051, 0.140899, 0.191306, 0.088856, 0.310395, 0.342127, -0.121957], [-0.376836, -0.030551, -0.281656, -0.307566, -0.361284, -0.109558, 0.096992]], "network.6.bias": [-0.012644, -0.213544, -0.130974, -0.095416, -0.13086, 0.729133, -0.296097], "network.8.weight": [[-0.359781, -0.009095, 0.122767, -0.071826, -0.042486, 0.199495, 0.034381], [-0.312453, -0.210575, 0.036223, -0.326399, -0.22013, 0.150529, 0.11431], [0.224373, -0.223163, -0.015183, 0.019742, 0.011242, 0.790907, -0.353785], [-0.177242, 0.311048, -0.316753, -0.029666, -0.343115, 0.01938, -0.170107], [-0.175515, -0.025661, 0.180442, -0.259101, -0.209449, 0.497129, -0.287058], [-0.101743, 0.241576, -0.424482, -0.315128, -0.005221, -0.282459, 0.165427], [0.386431, 0.3102, 0.760263, -0.038546, 0.298923, -0.616757, 0.024728]], "network.8.bias": [-0.363764, -0.307377, 0.664011, -0.243452, 0.525364, 0.03965, 0.129556], "network.10.weight": [[0.377958, -0.348164, 0.113177, 0.209938, -0.323088, 0.041111, -0.149619], [-0.072122, -0.010595, -0.398248, 0.259432, 0.090957, -0.202259, 0.56982], [-0.138257, -0.17486, 0.320924, 0.0328, -0.325217, -0.25616, -0.270992], [0.282022, -0.404823, 0.296393, 0.023418, 0.125639, 0.143354, -0.262961], [0.251087, 0.117399, 0.675095, -0.372959, 0.520469, 0.145071, -0.265112], [0.303878, 0.421899, -0.097009, 0.103215, -0.56897, 0.093587, 0.510254], [-0.097669, 0.530755, -0.629361, 0.212001, -0.12204, 0.008331, 0.03747]], "network.10.bias": [-0.306128, 0.133168, -0.220066, 0.790848, 0.656986, 0.121512, 0.018336], "network.12.weight": [[-0.350767, -0.653457, -0.251655, 0.587816, 0.799047, -0.589768, 0.133438]], "network.12.bias": [0.135511]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6912154257297516, "train_acc": 0.58, "val_loss": 0.7252156138420105, "val_acc": 0.48}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6803132891654968, "train_acc": 0.58, "val_loss": 0.7100004553794861, "val_acc": 0.48}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6794183850288391, "train_acc": 0.58, "val_loss": 0.6856794953346252, "val_acc": 0.48}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.6625008881092072, "train_acc": 0.505, "val_loss": 0.5861369967460632, "val_acc": 0.48}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.5708945691585541, "train_acc": 0.61, "val_loss": 0.45573386549949646, "val_acc": 0.9}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.4603542387485504, "train_acc": 0.885, "val_loss": 0.38293492794036865, "val_acc": 0.88}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.3767639994621277, "train_acc": 0.89, "val_loss": 0.3039224445819855, "val_acc": 0.92}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.31747712194919586, "train_acc": 0.885, "val_loss": 0.2419426292181015, "val_acc": 0.96}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.2613602429628372, "train_acc": 0.895, "val_loss": 0.22014766931533813, "val_acc": 0.94}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.24601513147354126, "train_acc": 0.9, "val_loss": 0.16306976974010468, "val_acc": 0.92}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.235054612159729, "train_acc": 0.905, "val_loss": 0.15598948299884796, "val_acc": 0.98}, {"stage": "improved", "epoch": 8, "global_epoch": 11, "train_loss": 0.22074957191944122, "train_acc": 0.905, "val_loss": 0.12956710159778595, "val_acc": 0.94}, {"stage": "improved", "epoch": 9, "global_epoch": 12, "train_loss": 0.21819238364696503, "train_acc": 0.92, "val_loss": 0.12354502081871033, "val_acc": 0.96}], "summary": {"total_epochs": 13, "degraded_epochs": 3, "improved_epochs": 10, "patterns": ["palindrome"], "degraded_stage": {"initial_val_loss": 0.7252156138420105, "final_val_loss": 0.6856794953346252, "initial_val_acc": 0.48, "final_val_acc": 0.48, "best_val_acc": 0.48}, "improved_stage": {"initial_val_loss": 0.5861369967460632, "final_val_loss": 0.12354502081871033, "initial_val_acc": 0.48, "final_val_acc": 0.96, "best_val_acc": 0.98, "best_epoch": 10}, "improvement": 0.5, "first_improvement_epoch": 2}}
|
1
|
{"target_pattern": "alternating", "degraded_accuracy": 0.52, "improved_accuracy": 0.92, "improvement": 0.4, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 9451, "learning_rate": 0.07352170370310572, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "alternating", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["alternating"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.350749,
0.022741,
0.24561,
0.369145,
0.543818
],
[
-0.784593,
-0.505623,
0.047746,
-0.131884,
-0.051516
],
[
-0.023241,
-0.080901,
-0.369471,
0.099241,
0.402252
],
[
-0.223858,
0.376454,
-0.595125,
0.390772,
-0.327672
],
[
-0.154215,
0.640965,
0.136615,
0.515237,
0.278974
],
[
-0.233706,
-0.19494,
0.353523,
0.112396,
-0.411489
],
[
0.418931,
-0.412868,
0.336435,
-0.763107,
-0.069389
]
],
"network.0.bias": [
0.36142,
0.084081,
-0.340971,
0.395617,
0.166324,
-0.2408,
-0.482828
],
"network.2.weight": [
[
-0.236004,
0.578259,
-0.289967,
0.515971,
0.072462,
0.011791,
0.16048
],
[
0.474862,
-0.459772,
0.01287,
-0.344333,
0.015363,
0.03543,
-0.589452
],
[
-0.118009,
0.40436,
0.163597,
-0.1563,
-0.099906,
0.203053,
0.044845
],
[
0.042569,
-0.112881,
0.704737,
-0.75616,
0.606623,
0.660571,
-0.392786
],
[
0.322126,
-0.40847,
0.086208,
0.554093,
-0.200485,
-0.243866,
0.55913
],
[
0.483438,
-0.236135,
0.501725,
-0.380317,
0.332086,
0.238996,
-0.293392
],
[
0.043822,
0.22006,
-0.005235,
0.534647,
-0.091811,
-0.034812,
0.310352
]
],
"network.2.bias": [
-0.123138,
0.187383,
0.282959,
0.130568,
0.009213,
0.350078,
0.063918
],
"network.4.weight": [
[
-0.369474,
0.387959,
-0.210282,
-0.195506,
-0.254928,
0.577384,
-0.30998
],
[
-0.200764,
0.136217,
-0.102683,
0.601136,
-0.304961,
0.364274,
-0.43103
],
[
-0.126448,
0.40362,
0.015479,
0.696791,
-0.180078,
0.308092,
-0.099349
],
[
0.045698,
-0.635704,
-0.168412,
0.468472,
0.128407,
-0.398191,
-0.060673
],
[
-0.619535,
0.150121,
-0.30259,
0.213474,
0.683084,
-0.163303,
-0.053653
],
[
-0.564362,
0.135669,
0.107934,
-0.10099,
0.539508,
-0.177831,
0.157239
],
[
-0.153653,
-0.436977,
-0.222605,
-0.152053,
-0.144703,
-0.516253,
0.453713
]
],
"network.4.bias": [
-0.260694,
0.235258,
0.245827,
-0.222018,
-0.114456,
-0.146646,
-0.206067
],
"network.6.weight": [
[
-0.258117,
-0.174745,
0.038918,
-0.545319,
0.07179,
-0.25871,
-0.161076
],
[
0.271989,
0.407153,
0.067816,
0.240921,
-0.023037,
0.32908,
0.291896
],
[
-0.11323,
0.108879,
-0.211443,
-0.174271,
0.124389,
0.146353,
0.11751
],
[
-0.117422,
0.604238,
0.072596,
0.29033,
0.313296,
-0.144542,
0.535626
],
[
-0.347906,
-0.428626,
0.284223,
-0.358097,
-0.211131,
-0.422963,
-0.009197
],
[
0.124001,
0.004126,
-0.295901,
-0.228066,
-0.455679,
-0.345138,
-0.099782
],
[
0.504156,
0.310465,
0.521353,
0.133778,
-0.327467,
-0.366576,
0.374892
]
],
"network.6.bias": [
-0.076262,
-0.219414,
0.261658,
-0.005811,
0.06013,
-0.102406,
0.251035
],
"network.8.weight": [
[
0.160282,
0.468223,
-0.061712,
0.175455,
0.10077,
0.006619,
-0.026136
],
[
-0.275662,
0.069069,
-0.334605,
0.056033,
-0.12303,
-0.346975,
0.36876
],
[
-0.015352,
-0.441747,
-0.142609,
-0.394476,
0.565378,
-0.039106,
-0.018703
],
[
-0.152891,
0.458172,
-0.195141,
0.598307,
-0.314712,
-0.038004,
0.629842
],
[
0.589618,
-0.076678,
-0.292796,
-0.156061,
0.004389,
-0.19163,
-0.310142
],
[
-0.009763,
-0.124552,
-0.277456,
-0.088162,
0.344165,
-0.064033,
-0.13644
],
[
-0.288017,
-0.268217,
0.356369,
-0.022197,
-0.255733,
-0.351994,
-0.13463
]
],
"network.8.bias": [
0.243465,
0.324187,
-0.014752,
-0.059377,
0.175747,
0.255096,
0.215125
],
"network.10.weight": [
[
-0.014447,
-0.375855,
0.119083,
-0.492942,
0.129516,
0.301704,
0.201898
]
],
"network.10.bias": [
0.264722
]
}
## Activation Signature
### 0
mean: [1.941744, -2.060277, -0.580170, -0.006271, 2.876730, -0.414864, -1.733074]
std: [1.540038, 2.131134, 0.994359, 1.861390, 1.969188, 1.032517, 2.210172]
### 2
mean: [-0.098283, 0.884047, -0.298041, 1.566564, 0.460217, 2.025386, 0.224169]
std: [0.679813, 0.858557, 0.424365, 1.426804, 0.702743, 1.413017, 0.528944]
### 4
mean: [0.687249, 1.737890, 2.126264, -0.746205, 0.258241, -0.328324, -1.768268]
std: [1.027876, 1.629525, 1.784214, 0.525884, 0.387891, 0.408743, 1.333594]
### 6
mean: [-0.364685, 0.730973, -0.040689, 1.093562, -0.303306, -0.654163, 2.126337]
std: [0.427067, 1.010593, 0.301358, 1.103774, 0.510473, 0.513145, 1.895594]
### 8
mean: [0.690842, 1.261036, -0.801140, 2.231313, -0.716119, -0.219866, -0.203726]
std: [0.575193, 0.904849, 0.914128, 2.350589, 0.837312, 0.471279, 0.535711]
### 10
mean: [-1.292513]
std: [1.588143]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
alternating
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.350749,
0.022741,
0.24561,
0.369145,
0.543818
],
[
-0.784593,
-0.505623,
0.047746,
-0.131884,
-0.051516
],
[
-0.023241,
-0.080901,
-0.369471,
0.099241,
0.402252
],
[
-0.223858,
0.376454,
-0.595125,
0.390772,
-0.327672
],
[
-0.154215,
0.640965,
0.136615,
0.515237,
0.278974
],
[
-0.233706,
-0.19494,
0.353523,
0.112396,
-0.411489
],
[
0.418931,
-0.412868,
0.336435,
-0.763107,
-0.069389
]
],
"network.0.bias": [
0.36142,
0.084081,
-0.340971,
0.395617,
0.166324,
-0.2408,
-0.482828
],
"network.2.weight": [
[
-0.236004,
0.578259,
-0.289967,
0.515971,
0.072462,
0.011791,
0.16048
],
[
0.474862,
-0.459772,
0.01287,
-0.344333,
0.015363,
0.03543,
-0.589452
],
[
-0.118009,
0.40436,
0.163597,
-0.1563,
-0.099906,
0.203053,
0.044845
],
[
0.042569,
-0.112881,
0.704737,
-0.75616,
0.606623,
0.660571,
-0.392786
],
[
0.322126,
-0.40847,
0.086208,
0.554093,
-0.200485,
-0.243866,
0.55913
],
[
0.483438,
-0.236135,
0.501725,
-0.380317,
0.332086,
0.238996,
-0.293392
],
[
0.043822,
0.22006,
-0.005235,
0.534647,
-0.091811,
-0.034812,
0.310352
]
],
"network.2.bias": [
-0.123138,
0.187383,
0.282959,
0.130568,
0.009213,
0.350078,
0.063918
],
"network.4.weight": [
[
-0.369474,
0.387959,
-0.210282,
-0.195506,
-0.254928,
0.577384,
-0.30998
],
[
-0.200764,
0.136217,
-0.102683,
0.601136,
-0.304961,
0.364274,
-0.43103
],
[
-0.126448,
0.40362,
0.015479,
0.696791,
-0.180078,
0.308092,
-0.099349
],
[
0.045698,
-0.635704,
-0.168412,
0.468472,
0.128407,
-0.398191,
-0.060673
],
[
-0.619535,
0.150121,
-0.30259,
0.213474,
0.683084,
-0.163303,
-0.053653
],
[
-0.564362,
0.135669,
0.107934,
-0.10099,
0.539508,
-0.177831,
0.157239
],
[
-0.153653,
-0.436977,
-0.222605,
-0.152053,
-0.144703,
-0.516253,
0.453713
]
],
"network.4.bias": [
-0.260694,
0.235258,
0.245827,
-0.222018,
-0.114456,
-0.146646,
-0.206067
],
"network.6.weight": [
[
-0.258117,
-0.174745,
0.038918,
-0.545319,
0.07179,
-0.25871,
-0.161076
],
[
0.271989,
0.407153,
0.067816,
0.240921,
-0.023037,
0.32908,
0.291896
],
[
-0.11323,
0.108879,
-0.211443,
-0.174271,
0.124389,
0.146353,
0.11751
],
[
-0.117422,
0.604238,
0.072596,
0.29033,
0.313296,
-0.144542,
0.535626
],
[
-0.347906,
-0.428626,
0.284223,
-0.358097,
-0.211131,
-0.422963,
-0.009197
],
[
0.124001,
0.004126,
-0.295901,
-0.228066,
-0.455679,
-0.345138,
-0.099782
],
[
0.504156,
0.310465,
0.521353,
0.133778,
-0.327467,
-0.366576,
0.374892
]
],
"network.6.bias": [
-0.076262,
-0.219414,
0.261658,
-0.005811,
0.06013,
-0.102406,
0.251035
],
"network.8.weight": [
[
0.160282,
0.468223,
-0.061712,
0.175455,
0.10077,
0.006619,
-0.026136
],
[
-0.275662,
0.069069,
-0.334605,
0.056033,
-0.12303,
-0.346975,
0.36876
],
[
-0.015352,
-0.441747,
-0.142609,
-0.394476,
0.565378,
-0.039106,
-0.018703
],
[
-0.152891,
0.458172,
-0.195141,
0.598307,
-0.314712,
-0.038004,
0.629842
],
[
0.589618,
-0.076678,
-0.292796,
-0.156061,
0.004389,
-0.19163,
-0.310142
],
[
-0.009763,
-0.124552,
-0.277456,
-0.088162,
0.344165,
-0.064033,
-0.13644
],
[
-0.288017,
-0.268217,
0.356369,
-0.022197,
-0.255733,
-0.351994,
-0.13463
]
],
"network.8.bias": [
0.243465,
0.324187,
-0.014752,
-0.059377,
0.175747,
0.255096,
0.215125
],
"network.10.weight": [
[
-0.014447,
-0.375855,
0.119083,
-0.492942,
0.129516,
0.301704,
0.201898
]
],
"network.10.bias": [
0.264722
]
}
## Activation Signature
### 0
mean: [1.941744, -2.060277, -0.580170, -0.006271, 2.876730, -0.414864, -1.733074]
std: [1.540038, 2.131134, 0.994359, 1.861390, 1.969188, 1.032517, 2.210172]
### 2
mean: [-0.098283, 0.884047, -0.298041, 1.566564, 0.460217, 2.025386, 0.224169]
std: [0.679813, 0.858557, 0.424365, 1.426804, 0.702743, 1.413017, 0.528944]
### 4
mean: [0.687249, 1.737890, 2.126264, -0.746205, 0.258241, -0.328324, -1.768268]
std: [1.027876, 1.629525, 1.784214, 0.525884, 0.387891, 0.408743, 1.333594]
### 6
mean: [-0.364685, 0.730973, -0.040689, 1.093562, -0.303306, -0.654163, 2.126337]
std: [0.427067, 1.010593, 0.301358, 1.103774, 0.510473, 0.513145, 1.895594]
### 8
mean: [0.690842, 1.261036, -0.801140, 2.231313, -0.716119, -0.219866, -0.203726]
std: [0.575193, 0.904849, 0.914128, 2.350589, 0.837312, 0.471279, 0.535711]
### 10
mean: [-1.292513]
std: [1.588143]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
alternating
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.9417439699172974, "std": 1.5400378704071045}, "1": {"mean": -2.060276746749878, "std": 2.131134033203125}, "2": {"mean": -0.5801698565483093, "std": 0.9943591952323914}, "3": {"mean": -0.0062707290053367615, "std": 1.861390233039856}, "4": {"mean": 2.876730442047119, "std": 1.969187617301941}, "5": {"mean": -0.4148639142513275, "std": 1.0325170755386353}, "6": {"mean": -1.7330735921859741, "std": 2.210172176361084}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.09828302264213562, "std": 0.6798130869865417}, "1": {"mean": 0.8840467929840088, "std": 0.8585572838783264}, "2": {"mean": -0.2980407774448395, "std": 0.424365371465683}, "3": {"mean": 1.5665640830993652, "std": 1.4268038272857666}, "4": {"mean": 0.4602169394493103, "std": 0.7027427554130554}, "5": {"mean": 2.025385618209839, "std": 1.4130165576934814}, "6": {"mean": 0.2241688221693039, "std": 0.5289435982704163}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.6872493624687195, "std": 1.0278756618499756}, "1": {"mean": 1.7378902435302734, "std": 1.6295251846313477}, "2": {"mean": 2.1262643337249756, "std": 1.7842140197753906}, "3": {"mean": -0.7462049722671509, "std": 0.5258837938308716}, "4": {"mean": 0.2582411766052246, "std": 0.38789135217666626}, "5": {"mean": -0.32832419872283936, "std": 0.40874314308166504}, "6": {"mean": -1.768268346786499, "std": 1.3335939645767212}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.36468467116355896, "std": 0.427067369222641}, "1": {"mean": 0.7309728860855103, "std": 1.0105925798416138}, "2": {"mean": -0.040688831359148026, "std": 0.3013584315776825}, "3": {"mean": 1.0935624837875366, "std": 1.103773593902588}, "4": {"mean": -0.3033055067062378, "std": 0.5104734897613525}, "5": {"mean": -0.6541626453399658, "std": 0.5131453275680542}, "6": {"mean": 2.1263365745544434, "std": 1.8955938816070557}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.6908419728279114, "std": 0.575192928314209}, "1": {"mean": 1.2610357999801636, "std": 0.9048486351966858}, "2": {"mean": -0.801139771938324, "std": 0.9141275882720947}, "3": {"mean": 2.2313129901885986, "std": 2.3505892753601074}, "4": {"mean": -0.7161186337471008, "std": 0.8373115062713623}, "5": {"mean": -0.21986648440361023, "std": 0.47127893567085266}, "6": {"mean": -0.20372597873210907, "std": 0.5357106924057007}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -1.292513132095337, "std": 1.5881426334381104}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.350749, 0.022741, 0.24561, 0.369145, 0.543818], [-0.784593, -0.505623, 0.047746, -0.131884, -0.051516], [-0.023241, -0.080901, -0.369471, 0.099241, 0.402252], [-0.223858, 0.376454, -0.595125, 0.390772, -0.327672], [-0.154215, 0.640965, 0.136615, 0.515237, 0.278974], [-0.233706, -0.19494, 0.353523, 0.112396, -0.411489], [0.418931, -0.412868, 0.336435, -0.763107, -0.069389]], "network.0.bias": [0.36142, 0.084081, -0.340971, 0.395617, 0.166324, -0.2408, -0.482828], "network.2.weight": [[-0.236004, 0.578259, -0.289967, 0.515971, 0.072462, 0.011791, 0.16048], [0.474862, -0.459772, 0.01287, -0.344333, 0.015363, 0.03543, -0.589452], [-0.118009, 0.40436, 0.163597, -0.1563, -0.099906, 0.203053, 0.044845], [0.042569, -0.112881, 0.704737, -0.75616, 0.606623, 0.660571, -0.392786], [0.322126, -0.40847, 0.086208, 0.554093, -0.200485, -0.243866, 0.55913], [0.483438, -0.236135, 0.501725, -0.380317, 0.332086, 0.238996, -0.293392], [0.043822, 0.22006, -0.005235, 0.534647, -0.091811, -0.034812, 0.310352]], "network.2.bias": [-0.123138, 0.187383, 0.282959, 0.130568, 0.009213, 0.350078, 0.063918], "network.4.weight": [[-0.369474, 0.387959, -0.210282, -0.195506, -0.254928, 0.577384, -0.30998], [-0.200764, 0.136217, -0.102683, 0.601136, -0.304961, 0.364274, -0.43103], [-0.126448, 0.40362, 0.015479, 0.696791, -0.180078, 0.308092, -0.099349], [0.045698, -0.635704, -0.168412, 0.468472, 0.128407, -0.398191, -0.060673], [-0.619535, 0.150121, -0.30259, 0.213474, 0.683084, -0.163303, -0.053653], [-0.564362, 0.135669, 0.107934, -0.10099, 0.539508, -0.177831, 0.157239], [-0.153653, -0.436977, -0.222605, -0.152053, -0.144703, -0.516253, 0.453713]], "network.4.bias": [-0.260694, 0.235258, 0.245827, -0.222018, -0.114456, -0.146646, -0.206067], "network.6.weight": [[-0.258117, -0.174745, 0.038918, -0.545319, 0.07179, -0.25871, -0.161076], [0.271989, 0.407153, 0.067816, 0.240921, -0.023037, 0.32908, 0.291896], [-0.11323, 0.108879, -0.211443, -0.174271, 0.124389, 0.146353, 0.11751], [-0.117422, 0.604238, 0.072596, 0.29033, 0.313296, -0.144542, 0.535626], [-0.347906, -0.428626, 0.284223, -0.358097, -0.211131, -0.422963, -0.009197], [0.124001, 0.004126, -0.295901, -0.228066, -0.455679, -0.345138, -0.099782], [0.504156, 0.310465, 0.521353, 0.133778, -0.327467, -0.366576, 0.374892]], "network.6.bias": [-0.076262, -0.219414, 0.261658, -0.005811, 0.06013, -0.102406, 0.251035], "network.8.weight": [[0.160282, 0.468223, -0.061712, 0.175455, 0.10077, 0.006619, -0.026136], [-0.275662, 0.069069, -0.334605, 0.056033, -0.12303, -0.346975, 0.36876], [-0.015352, -0.441747, -0.142609, -0.394476, 0.565378, -0.039106, -0.018703], [-0.152891, 0.458172, -0.195141, 0.598307, -0.314712, -0.038004, 0.629842], [0.589618, -0.076678, -0.292796, -0.156061, 0.004389, -0.19163, -0.310142], [-0.009763, -0.124552, -0.277456, -0.088162, 0.344165, -0.064033, -0.13644], [-0.288017, -0.268217, 0.356369, -0.022197, -0.255733, -0.351994, -0.13463]], "network.8.bias": [0.243465, 0.324187, -0.014752, -0.059377, 0.175747, 0.255096, 0.215125], "network.10.weight": [[-0.014447, -0.375855, 0.119083, -0.492942, 0.129516, 0.301704, 0.201898]], "network.10.bias": [0.264722]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6916628181934357, "train_acc": 0.545, "val_loss": 0.6948902606964111, "val_acc": 0.52}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6612586081027985, "train_acc": 0.56, "val_loss": 0.6322479248046875, "val_acc": 0.52}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.5553124845027924, "train_acc": 0.55, "val_loss": 0.39744308590888977, "val_acc": 0.92}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.3223879337310791, "train_acc": 0.945, "val_loss": 0.4029606580734253, "val_acc": 0.82}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.3463154584169388, "train_acc": 0.875, "val_loss": 0.4694277048110962, "val_acc": 0.8}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.28306517004966736, "train_acc": 0.89, "val_loss": 0.49492138624191284, "val_acc": 0.82}], "summary": {"total_epochs": 6, "degraded_epochs": 2, "improved_epochs": 4, "patterns": ["alternating"], "degraded_stage": {"initial_val_loss": 0.6948902606964111, "final_val_loss": 0.6322479248046875, "initial_val_acc": 0.52, "final_val_acc": 0.52, "best_val_acc": 0.52}, "improved_stage": {"initial_val_loss": 0.39744308590888977, "final_val_loss": 0.49492138624191284, "initial_val_acc": 0.92, "final_val_acc": 0.82, "best_val_acc": 0.92, "best_epoch": 2}, "improvement": 0.4, "first_improvement_epoch": 1}}
|
2
|
{"target_pattern": "increasing_pairs", "degraded_accuracy": 0.5, "improved_accuracy": 0.9, "improvement": 0.4, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 7902, "learning_rate": 0.019119242316001303, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "increasing_pairs", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["increasing_pairs"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.306744,
-0.107353,
0.323492,
-0.263617,
-0.071321
],
[
-0.136755,
0.031543,
-0.507803,
-0.48734,
-0.228265
],
[
0.462246,
0.20675,
0.034836,
-0.200921,
-0.218379
],
[
0.441719,
-0.246824,
0.40355,
-0.250275,
-0.173721
],
[
-0.27147,
0.000759,
0.24793,
0.206822,
-0.30087
],
[
-0.128065,
-0.089901,
0.207173,
0.481966,
0.683192
]
],
"network.0.bias": [
0.017415,
0.209661,
0.312007,
-0.179297,
-0.156691,
0.609966
],
"network.2.weight": [
[
0.146409,
0.229788,
-0.182541,
0.233177,
0.00381,
-0.302754
],
[
-0.376353,
-0.162428,
-0.01784,
0.144334,
0.164582,
-0.331682
],
[
-0.166265,
0.220298,
-0.286452,
-0.325655,
-0.423274,
0.209143
],
[
-0.047387,
-0.197965,
-0.35189,
0.059013,
-0.347397,
0.11409
],
[
-0.090832,
0.197845,
-0.117846,
0.284988,
0.020934,
-0.291669
],
[
0.241248,
0.013841,
-0.168016,
-0.224182,
0.337959,
-0.371723
]
],
"network.2.bias": [
-0.228908,
-0.139118,
-0.090114,
-0.123791,
-0.273926,
0.146292
],
"network.4.weight": [
[
-0.538209,
-0.300712,
-0.563901,
-0.023583,
-0.404277,
-0.498438
],
[
0.10558,
0.13363,
0.195469,
0.375287,
0.192972,
0.610911
],
[
-0.592715,
-0.484534,
-0.199585,
0.116719,
-0.29873,
-0.101867
],
[
-0.398807,
0.122273,
-0.381404,
-0.213997,
-0.182524,
-0.129994
],
[
-0.016247,
0.068468,
0.350695,
-0.191235,
0.138635,
0.538008
],
[
-0.366761,
-0.22621,
-0.200696,
-0.589808,
-0.541964,
-0.457728
]
],
"network.4.bias": [
0.559085,
-0.152053,
-0.076749,
0.445926,
-0.223831,
0.474309
],
"network.6.weight": [
[
0.327886,
-0.496222,
0.331148,
0.227621,
-0.5305,
0.64235
],
[
-0.433197,
0.39602,
-0.308265,
0.236144,
0.145873,
-0.373335
],
[
-0.364165,
0.225038,
-0.295511,
-0.019409,
0.111191,
0.04238
],
[
0.615553,
0.1129,
-0.122227,
0.185108,
0.160538,
0.614453
],
[
0.08013,
0.390056,
-0.054861,
0.214988,
0.340561,
-0.006452
],
[
0.319114,
-0.343264,
0.037997,
0.494555,
-0.653354,
0.188986
]
],
"network.6.bias": [
0.242986,
0.058184,
-0.020256,
0.224645,
-0.06052,
0.222313
],
"network.8.weight": [
[
-0.528034,
0.223621,
0.396886,
-0.253046,
0.005106,
-0.631389
]
],
"network.8.bias": [
0.263716
]
}
## Activation Signature
### 0
mean: [0.225986, -2.290389, 0.636561, 0.012643, 0.097207, 2.595692]
std: [1.142856, 1.645565, 1.221557, 1.430944, 0.902008, 1.787508]
### 2
mean: [-0.954601, -1.048196, -0.118633, -0.171197, -1.005633, -0.810383]
std: [0.578908, 0.621705, 0.813907, 0.446660, 0.530225, 0.682586]
### 4
mean: [0.697518, -0.252426, 0.084479, 0.466904, -0.259491, 0.649963]
std: [0.264418, 0.167469, 0.103890, 0.196127, 0.117419, 0.224882]
### 6
mean: [0.931000, -0.352502, -0.251233, 0.897562, -0.019905, 0.751927]
std: [0.290612, 0.176047, 0.107202, 0.243370, 0.015715, 0.224247]
### 8
mean: [-0.779444]
std: [0.353857]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
increasing_pairs
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.306744,
-0.107353,
0.323492,
-0.263617,
-0.071321
],
[
-0.136755,
0.031543,
-0.507803,
-0.48734,
-0.228265
],
[
0.462246,
0.20675,
0.034836,
-0.200921,
-0.218379
],
[
0.441719,
-0.246824,
0.40355,
-0.250275,
-0.173721
],
[
-0.27147,
0.000759,
0.24793,
0.206822,
-0.30087
],
[
-0.128065,
-0.089901,
0.207173,
0.481966,
0.683192
]
],
"network.0.bias": [
0.017415,
0.209661,
0.312007,
-0.179297,
-0.156691,
0.609966
],
"network.2.weight": [
[
0.146409,
0.229788,
-0.182541,
0.233177,
0.00381,
-0.302754
],
[
-0.376353,
-0.162428,
-0.01784,
0.144334,
0.164582,
-0.331682
],
[
-0.166265,
0.220298,
-0.286452,
-0.325655,
-0.423274,
0.209143
],
[
-0.047387,
-0.197965,
-0.35189,
0.059013,
-0.347397,
0.11409
],
[
-0.090832,
0.197845,
-0.117846,
0.284988,
0.020934,
-0.291669
],
[
0.241248,
0.013841,
-0.168016,
-0.224182,
0.337959,
-0.371723
]
],
"network.2.bias": [
-0.228908,
-0.139118,
-0.090114,
-0.123791,
-0.273926,
0.146292
],
"network.4.weight": [
[
-0.538209,
-0.300712,
-0.563901,
-0.023583,
-0.404277,
-0.498438
],
[
0.10558,
0.13363,
0.195469,
0.375287,
0.192972,
0.610911
],
[
-0.592715,
-0.484534,
-0.199585,
0.116719,
-0.29873,
-0.101867
],
[
-0.398807,
0.122273,
-0.381404,
-0.213997,
-0.182524,
-0.129994
],
[
-0.016247,
0.068468,
0.350695,
-0.191235,
0.138635,
0.538008
],
[
-0.366761,
-0.22621,
-0.200696,
-0.589808,
-0.541964,
-0.457728
]
],
"network.4.bias": [
0.559085,
-0.152053,
-0.076749,
0.445926,
-0.223831,
0.474309
],
"network.6.weight": [
[
0.327886,
-0.496222,
0.331148,
0.227621,
-0.5305,
0.64235
],
[
-0.433197,
0.39602,
-0.308265,
0.236144,
0.145873,
-0.373335
],
[
-0.364165,
0.225038,
-0.295511,
-0.019409,
0.111191,
0.04238
],
[
0.615553,
0.1129,
-0.122227,
0.185108,
0.160538,
0.614453
],
[
0.08013,
0.390056,
-0.054861,
0.214988,
0.340561,
-0.006452
],
[
0.319114,
-0.343264,
0.037997,
0.494555,
-0.653354,
0.188986
]
],
"network.6.bias": [
0.242986,
0.058184,
-0.020256,
0.224645,
-0.06052,
0.222313
],
"network.8.weight": [
[
-0.528034,
0.223621,
0.396886,
-0.253046,
0.005106,
-0.631389
]
],
"network.8.bias": [
0.263716
]
}
## Activation Signature
### 0
mean: [0.225986, -2.290389, 0.636561, 0.012643, 0.097207, 2.595692]
std: [1.142856, 1.645565, 1.221557, 1.430944, 0.902008, 1.787508]
### 2
mean: [-0.954601, -1.048196, -0.118633, -0.171197, -1.005633, -0.810383]
std: [0.578908, 0.621705, 0.813907, 0.446660, 0.530225, 0.682586]
### 4
mean: [0.697518, -0.252426, 0.084479, 0.466904, -0.259491, 0.649963]
std: [0.264418, 0.167469, 0.103890, 0.196127, 0.117419, 0.224882]
### 6
mean: [0.931000, -0.352502, -0.251233, 0.897562, -0.019905, 0.751927]
std: [0.290612, 0.176047, 0.107202, 0.243370, 0.015715, 0.224247]
### 8
mean: [-0.779444]
std: [0.353857]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
increasing_pairs
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.22598551213741302, "std": 1.1428560018539429}, "1": {"mean": -2.290388584136963, "std": 1.6455649137496948}, "2": {"mean": 0.6365607380867004, "std": 1.2215566635131836}, "3": {"mean": 0.012643260881304741, "std": 1.4309442043304443}, "4": {"mean": 0.09720747917890549, "std": 0.9020075798034668}, "5": {"mean": 2.5956919193267822, "std": 1.7875081300735474}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.9546012282371521, "std": 0.57890784740448}, "1": {"mean": -1.0481958389282227, "std": 0.6217047572135925}, "2": {"mean": -0.11863294988870621, "std": 0.8139069676399231}, "3": {"mean": -0.17119666934013367, "std": 0.4466601014137268}, "4": {"mean": -1.0056332349777222, "std": 0.5302248597145081}, "5": {"mean": -0.8103832602500916, "std": 0.6825860738754272}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.6975183486938477, "std": 0.2644175887107849}, "1": {"mean": -0.25242623686790466, "std": 0.1674690991640091}, "2": {"mean": 0.08447884768247604, "std": 0.10388977080583572}, "3": {"mean": 0.4669043719768524, "std": 0.19612716138362885}, "4": {"mean": -0.25949129462242126, "std": 0.11741935461759567}, "5": {"mean": 0.64996337890625, "std": 0.22488191723823547}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.9309995174407959, "std": 0.2906123995780945}, "1": {"mean": -0.3525024950504303, "std": 0.17604725062847137}, "2": {"mean": -0.25123292207717896, "std": 0.1072019413113594}, "3": {"mean": 0.8975619673728943, "std": 0.24336963891983032}, "4": {"mean": -0.019904814660549164, "std": 0.015715094283223152}, "5": {"mean": 0.751927375793457, "std": 0.2242473065853119}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.7794444561004639, "std": 0.35385698080062866}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.306744, -0.107353, 0.323492, -0.263617, -0.071321], [-0.136755, 0.031543, -0.507803, -0.48734, -0.228265], [0.462246, 0.20675, 0.034836, -0.200921, -0.218379], [0.441719, -0.246824, 0.40355, -0.250275, -0.173721], [-0.27147, 0.000759, 0.24793, 0.206822, -0.30087], [-0.128065, -0.089901, 0.207173, 0.481966, 0.683192]], "network.0.bias": [0.017415, 0.209661, 0.312007, -0.179297, -0.156691, 0.609966], "network.2.weight": [[0.146409, 0.229788, -0.182541, 0.233177, 0.00381, -0.302754], [-0.376353, -0.162428, -0.01784, 0.144334, 0.164582, -0.331682], [-0.166265, 0.220298, -0.286452, -0.325655, -0.423274, 0.209143], [-0.047387, -0.197965, -0.35189, 0.059013, -0.347397, 0.11409], [-0.090832, 0.197845, -0.117846, 0.284988, 0.020934, -0.291669], [0.241248, 0.013841, -0.168016, -0.224182, 0.337959, -0.371723]], "network.2.bias": [-0.228908, -0.139118, -0.090114, -0.123791, -0.273926, 0.146292], "network.4.weight": [[-0.538209, -0.300712, -0.563901, -0.023583, -0.404277, -0.498438], [0.10558, 0.13363, 0.195469, 0.375287, 0.192972, 0.610911], [-0.592715, -0.484534, -0.199585, 0.116719, -0.29873, -0.101867], [-0.398807, 0.122273, -0.381404, -0.213997, -0.182524, -0.129994], [-0.016247, 0.068468, 0.350695, -0.191235, 0.138635, 0.538008], [-0.366761, -0.22621, -0.200696, -0.589808, -0.541964, -0.457728]], "network.4.bias": [0.559085, -0.152053, -0.076749, 0.445926, -0.223831, 0.474309], "network.6.weight": [[0.327886, -0.496222, 0.331148, 0.227621, -0.5305, 0.64235], [-0.433197, 0.39602, -0.308265, 0.236144, 0.145873, -0.373335], [-0.364165, 0.225038, -0.295511, -0.019409, 0.111191, 0.04238], [0.615553, 0.1129, -0.122227, 0.185108, 0.160538, 0.614453], [0.08013, 0.390056, -0.054861, 0.214988, 0.340561, -0.006452], [0.319114, -0.343264, 0.037997, 0.494555, -0.653354, 0.188986]], "network.6.bias": [0.242986, 0.058184, -0.020256, 0.224645, -0.06052, 0.222313], "network.8.weight": [[-0.528034, 0.223621, 0.396886, -0.253046, 0.005106, -0.631389]], "network.8.bias": [0.263716]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7180013060569763, "train_acc": 0.425, "val_loss": 0.6934688091278076, "val_acc": 0.5}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.7019679844379425, "train_acc": 0.425, "val_loss": 0.6853039264678955, "val_acc": 0.68}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6877947449684143, "train_acc": 0.53, "val_loss": 0.6773688793182373, "val_acc": 0.5}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6712348759174347, "train_acc": 0.575, "val_loss": 0.6642918586730957, "val_acc": 0.5}, {"stage": "degraded", "epoch": 4, "global_epoch": 4, "train_loss": 0.6509652733802795, "train_acc": 0.575, "val_loss": 0.635927140712738, "val_acc": 0.5}, {"stage": "improved", "epoch": 0, "global_epoch": 5, "train_loss": 0.6322189569473267, "train_acc": 0.54, "val_loss": 0.5644363164901733, "val_acc": 0.88}, {"stage": "improved", "epoch": 1, "global_epoch": 6, "train_loss": 0.5598377883434296, "train_acc": 0.785, "val_loss": 0.48903128504753113, "val_acc": 0.9}, {"stage": "improved", "epoch": 2, "global_epoch": 7, "train_loss": 0.5017989724874496, "train_acc": 0.835, "val_loss": 0.423909455537796, "val_acc": 0.88}, {"stage": "improved", "epoch": 3, "global_epoch": 8, "train_loss": 0.4411952644586563, "train_acc": 0.84, "val_loss": 0.38604167103767395, "val_acc": 0.88}, {"stage": "improved", "epoch": 4, "global_epoch": 9, "train_loss": 0.4078972637653351, "train_acc": 0.85, "val_loss": 0.39802730083465576, "val_acc": 0.84}, {"stage": "improved", "epoch": 5, "global_epoch": 10, "train_loss": 0.33288122713565826, "train_acc": 0.845, "val_loss": 0.3344477415084839, "val_acc": 0.9}, {"stage": "improved", "epoch": 6, "global_epoch": 11, "train_loss": 0.34394922852516174, "train_acc": 0.855, "val_loss": 0.3352448344230652, "val_acc": 0.86}, {"stage": "improved", "epoch": 7, "global_epoch": 12, "train_loss": 0.3285396099090576, "train_acc": 0.86, "val_loss": 0.374785453081131, "val_acc": 0.84}, {"stage": "improved", "epoch": 8, "global_epoch": 13, "train_loss": 0.3410664349794388, "train_acc": 0.86, "val_loss": 0.3444004952907562, "val_acc": 0.88}], "summary": {"total_epochs": 14, "degraded_epochs": 5, "improved_epochs": 9, "patterns": ["increasing_pairs"], "degraded_stage": {"initial_val_loss": 0.6934688091278076, "final_val_loss": 0.635927140712738, "initial_val_acc": 0.5, "final_val_acc": 0.5, "best_val_acc": 0.5}, "improved_stage": {"initial_val_loss": 0.5644363164901733, "final_val_loss": 0.3444004952907562, "initial_val_acc": 0.88, "final_val_acc": 0.88, "best_val_acc": 0.9, "best_epoch": 6}, "improvement": 0.4, "first_improvement_epoch": 4}}
|
3
|
{"target_pattern": "contains_abc", "degraded_accuracy": 0.76, "improved_accuracy": 0.94, "improvement": 0.17999999999999994, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 9806, "learning_rate": 0.07052986855265303, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "contains_abc", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["contains_abc"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.8145,
-0.098256,
-0.141614,
-0.480289,
0.287535
],
[
1.557012,
0.277012,
0.215586,
-0.057014,
0.154588
],
[
-0.603045,
-0.431751,
-0.2718,
-0.501871,
0.497004
],
[
1.689838,
-0.63611,
0.353659,
-0.26314,
-0.109576
],
[
-0.693374,
-0.165124,
-0.254091,
-0.911058,
-0.173411
],
[
-0.919414,
-0.133029,
-0.354827,
-0.536139,
-0.174826
],
[
1.37564,
0.555747,
0.233335,
0.515827,
-0.039106
],
[
1.324016,
0.338061,
-0.021417,
0.08084,
-0.525438
]
],
"network.0.bias": [
0.282185,
-0.469087,
0.024607,
-0.18766,
0.575402,
0.098782,
-0.747294,
-0.233348
],
"network.2.weight": [
[
-0.155801,
0.01394,
-0.101887,
-0.293167,
0.188596,
0.338811,
-0.231814,
0.605413
],
[
0.56831,
-0.073214,
0.174597,
-0.434233,
0.35737,
0.057295,
-0.422339,
0.363013
],
[
-0.347641,
0.211596,
-0.465096,
0.642853,
0.002887,
-0.240122,
0.308835,
0.295166
],
[
0.154473,
0.081493,
0.148713,
0.001765,
-0.469389,
0.005162,
-0.364831,
-0.189866
],
[
-0.355282,
0.439417,
0.023051,
0.841655,
-0.059086,
0.054031,
0.711715,
0.62617
],
[
0.281417,
-0.385196,
-0.009111,
0.054652,
-0.395977,
-0.30935,
-0.364401,
-0.358135
],
[
-0.141485,
-0.223691,
-0.027314,
-0.785053,
0.120959,
-0.365752,
-1.003441,
-0.209861
],
[
0.592048,
0.01653,
0.38308,
-0.420307,
0.260938,
0.218871,
-0.327694,
0.185873
]
],
"network.2.bias": [
1.099915,
0.9998,
-0.287355,
-0.202204,
-0.209357,
-0.238452,
0.399074,
1.064571
],
"network.4.weight": [
[
0.572151,
0.667922,
0.196986,
0.229669,
-0.4317,
-0.158671,
0.577252,
0.725696
],
[
0.118937,
-0.193906,
0.545395,
0.384082,
0.543701,
-0.086901,
-0.007178,
-0.521401
],
[
-0.34733,
0.152592,
0.005943,
-0.213706,
-0.111773,
-0.235395,
-0.221955,
-0.321266
],
[
-0.219831,
-0.349399,
0.172135,
-0.237972,
-0.27954,
-0.245902,
0.271767,
-0.530239
],
[
-0.368988,
0.09589,
0.071923,
0.256392,
-0.244165,
-0.189567,
-0.128155,
-0.067391
],
[
0.49308,
0.463966,
-0.110775,
0.103478,
-0.252629,
0.303605,
-0.186981,
0.861212
],
[
-0.195205,
0.053893,
0.522439,
-0.147537,
0.610225,
0.225879,
0.179656,
-0.627609
],
[
0.118585,
0.458815,
-0.204506,
0.068376,
-0.386708,
0.283885,
0.036866,
0.347156
]
],
"network.4.bias": [
0.450391,
-0.202538,
-0.192971,
0.034328,
-0.192684,
0.829731,
-0.152038,
0.230208
],
"network.6.weight": [
[
0.564413,
-0.529784,
-0.210255,
-0.377733,
0.140415,
0.755286,
0.020801,
-0.020592
],
[
0.70065,
-0.413279,
-0.220306,
-0.100543,
-0.164293,
0.41649,
-0.248665,
0.394003
],
[
-0.443537,
-0.207914,
0.279746,
-0.33491,
-0.242595,
0.139062,
-0.077288,
0.235332
],
[
-0.032086,
-0.137875,
-0.242526,
0.133623,
-0.258947,
-0.191084,
-0.09287,
-0.172564
],
[
-0.166274,
-0.032024,
-0.169387,
0.026219,
0.024342,
-0.034809,
-0.480389,
-0.314925
],
[
-0.178947,
0.491395,
0.022618,
-0.273356,
-0.537172,
-0.674621,
0.281389,
0.069917
],
[
-0.139695,
0.582206,
0.352497,
0.202338,
0.293989,
-0.867952,
0.594585,
-0.122909
],
[
-0.092731,
0.130497,
-0.026145,
0.078665,
0.040088,
-0.528748,
-0.466538,
0.068055
]
],
"network.6.bias": [
0.307938,
0.447669,
-0.234141,
-0.310698,
0.052046,
-0.321453,
0.052176,
-0.175017
],
"network.8.weight": [
[
-0.300456,
0.019333,
-0.282352,
0.291371,
0.238673,
0.035182,
-0.169713,
0.293109
],
[
-0.091959,
-0.126531,
0.02285,
0.175839,
-0.112035,
0.543031,
0.66183,
0.200997
],
[
0.660009,
0.697586,
0.089137,
0.149949,
0.188008,
-0.411646,
-0.502204,
-0.196745
],
[
-0.09597,
-0.003262,
-0.054916,
-0.051813,
0.094712,
0.0904,
-0.274368,
-0.14529
],
[
0.096672,
0.151043,
0.037291,
0.078396,
-0.377483,
-0.261309,
0.394418,
-0.166439
],
[
-0.439822,
0.076393,
-0.388637,
0.341297,
-0.208134,
-0.374116,
-0.098431,
0.336323
],
[
-0.192834,
0.201199,
0.04149,
-0.199808,
0.133956,
0.179992,
-0.28344,
0.142172
],
[
-0.288631,
-0.397923,
0.140568,
-0.061804,
0.040679,
0.278745,
0.49284,
0.069817
]
],
"network.8.bias": [
-0.2012,
-0.220711,
0.229507,
-0.028512,
-0.308421,
-0.572613,
-0.258768,
0.107871
],
"network.10.weight": [
[
0.101312,
-0.451813,
0.519364,
-0.044861,
-0.127357,
0.208859,
-0.284451,
-0.499951
]
],
"network.10.bias": [
0.333218
]
}
## Activation Signature
### 0
mean: [-1.881601, 2.492038, -2.543617, 0.795437, -3.286712, -3.394318, 3.522650, 1.510916]
std: [2.016584, 3.588881, 2.265957, 3.568262, 2.719091, 2.643123, 3.669856, 2.943410]
### 2
mean: [0.900855, -0.709900, 2.940474, -1.662077, 6.015716, -3.109375, -5.511407, -0.363390]
std: [0.565841, 2.031406, 4.297233, 1.501252, 7.805554, 3.380454, 6.808144, 1.800132]
### 4
mean: [-0.426453, 4.541361, -1.282261, -1.702041, -1.802257, -0.016872, 4.679965, -2.291774]
std: [3.006149, 6.764658, 0.833703, 1.278590, 1.639980, 2.851571, 7.115798, 4.106448]
### 6
mean: [-0.765320, -1.499573, -1.777942, -1.723726, -2.765054, 2.517483, 4.620115, -2.493102]
std: [4.463049, 5.468106, 1.795148, 1.379677, 3.351362, 5.848035, 8.791576, 2.021039]
### 8
mean: [-1.411928, 5.019002, -2.133289, -1.382567, 1.328351, -2.898342, -1.218102, 2.874221]
std: [1.007399, 8.351899, 7.699241, 1.658160, 1.637167, 2.452035, 1.360346, 6.134490]
### 10
mean: [-3.152317]
std: [7.335796]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
contains_abc
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.8145,
-0.098256,
-0.141614,
-0.480289,
0.287535
],
[
1.557012,
0.277012,
0.215586,
-0.057014,
0.154588
],
[
-0.603045,
-0.431751,
-0.2718,
-0.501871,
0.497004
],
[
1.689838,
-0.63611,
0.353659,
-0.26314,
-0.109576
],
[
-0.693374,
-0.165124,
-0.254091,
-0.911058,
-0.173411
],
[
-0.919414,
-0.133029,
-0.354827,
-0.536139,
-0.174826
],
[
1.37564,
0.555747,
0.233335,
0.515827,
-0.039106
],
[
1.324016,
0.338061,
-0.021417,
0.08084,
-0.525438
]
],
"network.0.bias": [
0.282185,
-0.469087,
0.024607,
-0.18766,
0.575402,
0.098782,
-0.747294,
-0.233348
],
"network.2.weight": [
[
-0.155801,
0.01394,
-0.101887,
-0.293167,
0.188596,
0.338811,
-0.231814,
0.605413
],
[
0.56831,
-0.073214,
0.174597,
-0.434233,
0.35737,
0.057295,
-0.422339,
0.363013
],
[
-0.347641,
0.211596,
-0.465096,
0.642853,
0.002887,
-0.240122,
0.308835,
0.295166
],
[
0.154473,
0.081493,
0.148713,
0.001765,
-0.469389,
0.005162,
-0.364831,
-0.189866
],
[
-0.355282,
0.439417,
0.023051,
0.841655,
-0.059086,
0.054031,
0.711715,
0.62617
],
[
0.281417,
-0.385196,
-0.009111,
0.054652,
-0.395977,
-0.30935,
-0.364401,
-0.358135
],
[
-0.141485,
-0.223691,
-0.027314,
-0.785053,
0.120959,
-0.365752,
-1.003441,
-0.209861
],
[
0.592048,
0.01653,
0.38308,
-0.420307,
0.260938,
0.218871,
-0.327694,
0.185873
]
],
"network.2.bias": [
1.099915,
0.9998,
-0.287355,
-0.202204,
-0.209357,
-0.238452,
0.399074,
1.064571
],
"network.4.weight": [
[
0.572151,
0.667922,
0.196986,
0.229669,
-0.4317,
-0.158671,
0.577252,
0.725696
],
[
0.118937,
-0.193906,
0.545395,
0.384082,
0.543701,
-0.086901,
-0.007178,
-0.521401
],
[
-0.34733,
0.152592,
0.005943,
-0.213706,
-0.111773,
-0.235395,
-0.221955,
-0.321266
],
[
-0.219831,
-0.349399,
0.172135,
-0.237972,
-0.27954,
-0.245902,
0.271767,
-0.530239
],
[
-0.368988,
0.09589,
0.071923,
0.256392,
-0.244165,
-0.189567,
-0.128155,
-0.067391
],
[
0.49308,
0.463966,
-0.110775,
0.103478,
-0.252629,
0.303605,
-0.186981,
0.861212
],
[
-0.195205,
0.053893,
0.522439,
-0.147537,
0.610225,
0.225879,
0.179656,
-0.627609
],
[
0.118585,
0.458815,
-0.204506,
0.068376,
-0.386708,
0.283885,
0.036866,
0.347156
]
],
"network.4.bias": [
0.450391,
-0.202538,
-0.192971,
0.034328,
-0.192684,
0.829731,
-0.152038,
0.230208
],
"network.6.weight": [
[
0.564413,
-0.529784,
-0.210255,
-0.377733,
0.140415,
0.755286,
0.020801,
-0.020592
],
[
0.70065,
-0.413279,
-0.220306,
-0.100543,
-0.164293,
0.41649,
-0.248665,
0.394003
],
[
-0.443537,
-0.207914,
0.279746,
-0.33491,
-0.242595,
0.139062,
-0.077288,
0.235332
],
[
-0.032086,
-0.137875,
-0.242526,
0.133623,
-0.258947,
-0.191084,
-0.09287,
-0.172564
],
[
-0.166274,
-0.032024,
-0.169387,
0.026219,
0.024342,
-0.034809,
-0.480389,
-0.314925
],
[
-0.178947,
0.491395,
0.022618,
-0.273356,
-0.537172,
-0.674621,
0.281389,
0.069917
],
[
-0.139695,
0.582206,
0.352497,
0.202338,
0.293989,
-0.867952,
0.594585,
-0.122909
],
[
-0.092731,
0.130497,
-0.026145,
0.078665,
0.040088,
-0.528748,
-0.466538,
0.068055
]
],
"network.6.bias": [
0.307938,
0.447669,
-0.234141,
-0.310698,
0.052046,
-0.321453,
0.052176,
-0.175017
],
"network.8.weight": [
[
-0.300456,
0.019333,
-0.282352,
0.291371,
0.238673,
0.035182,
-0.169713,
0.293109
],
[
-0.091959,
-0.126531,
0.02285,
0.175839,
-0.112035,
0.543031,
0.66183,
0.200997
],
[
0.660009,
0.697586,
0.089137,
0.149949,
0.188008,
-0.411646,
-0.502204,
-0.196745
],
[
-0.09597,
-0.003262,
-0.054916,
-0.051813,
0.094712,
0.0904,
-0.274368,
-0.14529
],
[
0.096672,
0.151043,
0.037291,
0.078396,
-0.377483,
-0.261309,
0.394418,
-0.166439
],
[
-0.439822,
0.076393,
-0.388637,
0.341297,
-0.208134,
-0.374116,
-0.098431,
0.336323
],
[
-0.192834,
0.201199,
0.04149,
-0.199808,
0.133956,
0.179992,
-0.28344,
0.142172
],
[
-0.288631,
-0.397923,
0.140568,
-0.061804,
0.040679,
0.278745,
0.49284,
0.069817
]
],
"network.8.bias": [
-0.2012,
-0.220711,
0.229507,
-0.028512,
-0.308421,
-0.572613,
-0.258768,
0.107871
],
"network.10.weight": [
[
0.101312,
-0.451813,
0.519364,
-0.044861,
-0.127357,
0.208859,
-0.284451,
-0.499951
]
],
"network.10.bias": [
0.333218
]
}
## Activation Signature
### 0
mean: [-1.881601, 2.492038, -2.543617, 0.795437, -3.286712, -3.394318, 3.522650, 1.510916]
std: [2.016584, 3.588881, 2.265957, 3.568262, 2.719091, 2.643123, 3.669856, 2.943410]
### 2
mean: [0.900855, -0.709900, 2.940474, -1.662077, 6.015716, -3.109375, -5.511407, -0.363390]
std: [0.565841, 2.031406, 4.297233, 1.501252, 7.805554, 3.380454, 6.808144, 1.800132]
### 4
mean: [-0.426453, 4.541361, -1.282261, -1.702041, -1.802257, -0.016872, 4.679965, -2.291774]
std: [3.006149, 6.764658, 0.833703, 1.278590, 1.639980, 2.851571, 7.115798, 4.106448]
### 6
mean: [-0.765320, -1.499573, -1.777942, -1.723726, -2.765054, 2.517483, 4.620115, -2.493102]
std: [4.463049, 5.468106, 1.795148, 1.379677, 3.351362, 5.848035, 8.791576, 2.021039]
### 8
mean: [-1.411928, 5.019002, -2.133289, -1.382567, 1.328351, -2.898342, -1.218102, 2.874221]
std: [1.007399, 8.351899, 7.699241, 1.658160, 1.637167, 2.452035, 1.360346, 6.134490]
### 10
mean: [-3.152317]
std: [7.335796]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
contains_abc
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.8816006183624268, "std": 2.0165841579437256}, "1": {"mean": 2.492037534713745, "std": 3.588880777359009}, "2": {"mean": -2.543617010116577, "std": 2.2659566402435303}, "3": {"mean": 0.7954368591308594, "std": 3.5682616233825684}, "4": {"mean": -3.2867116928100586, "std": 2.7190914154052734}, "5": {"mean": -3.394317626953125, "std": 2.643122673034668}, "6": {"mean": 3.5226504802703857, "std": 3.669856309890747}, "7": {"mean": 1.5109163522720337, "std": 2.9434099197387695}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.9008548259735107, "std": 0.5658407807350159}, "1": {"mean": -0.7098998427391052, "std": 2.0314061641693115}, "2": {"mean": 2.940474033355713, "std": 4.297232627868652}, "3": {"mean": -1.6620768308639526, "std": 1.5012515783309937}, "4": {"mean": 6.015716075897217, "std": 7.805554389953613}, "5": {"mean": -3.109374761581421, "std": 3.3804543018341064}, "6": {"mean": -5.511406898498535, "std": 6.808143615722656}, "7": {"mean": -0.36338967084884644, "std": 1.8001316785812378}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.4264533221721649, "std": 3.0061492919921875}, "1": {"mean": 4.541361331939697, "std": 6.764657974243164}, "2": {"mean": -1.2822614908218384, "std": 0.8337028622627258}, "3": {"mean": -1.7020409107208252, "std": 1.2785899639129639}, "4": {"mean": -1.8022568225860596, "std": 1.6399801969528198}, "5": {"mean": -0.016871564090251923, "std": 2.8515708446502686}, "6": {"mean": 4.679965019226074, "std": 7.115797996520996}, "7": {"mean": -2.291774272918701, "std": 4.106448173522949}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.7653196454048157, "std": 4.463048934936523}, "1": {"mean": -1.4995731115341187, "std": 5.468106269836426}, "2": {"mean": -1.7779417037963867, "std": 1.795148253440857}, "3": {"mean": -1.7237259149551392, "std": 1.3796772956848145}, "4": {"mean": -2.7650535106658936, "std": 3.3513622283935547}, "5": {"mean": 2.5174825191497803, "std": 5.8480353355407715}, "6": {"mean": 4.620114803314209, "std": 8.791576385498047}, "7": {"mean": -2.4931020736694336, "std": 2.0210394859313965}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.411928415298462, "std": 1.0073987245559692}, "1": {"mean": 5.019002437591553, "std": 8.351899147033691}, "2": {"mean": -2.133289337158203, "std": 7.6992411613464355}, "3": {"mean": -1.3825668096542358, "std": 1.6581603288650513}, "4": {"mean": 1.32835054397583, "std": 1.6371667385101318}, "5": {"mean": -2.8983418941497803, "std": 2.4520347118377686}, "6": {"mean": -1.2181024551391602, "std": 1.3603459596633911}, "7": {"mean": 2.874220848083496, "std": 6.134490489959717}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -3.1523170471191406, "std": 7.335795879364014}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.8145, -0.098256, -0.141614, -0.480289, 0.287535], [1.557012, 0.277012, 0.215586, -0.057014, 0.154588], [-0.603045, -0.431751, -0.2718, -0.501871, 0.497004], [1.689838, -0.63611, 0.353659, -0.26314, -0.109576], [-0.693374, -0.165124, -0.254091, -0.911058, -0.173411], [-0.919414, -0.133029, -0.354827, -0.536139, -0.174826], [1.37564, 0.555747, 0.233335, 0.515827, -0.039106], [1.324016, 0.338061, -0.021417, 0.08084, -0.525438]], "network.0.bias": [0.282185, -0.469087, 0.024607, -0.18766, 0.575402, 0.098782, -0.747294, -0.233348], "network.2.weight": [[-0.155801, 0.01394, -0.101887, -0.293167, 0.188596, 0.338811, -0.231814, 0.605413], [0.56831, -0.073214, 0.174597, -0.434233, 0.35737, 0.057295, -0.422339, 0.363013], [-0.347641, 0.211596, -0.465096, 0.642853, 0.002887, -0.240122, 0.308835, 0.295166], [0.154473, 0.081493, 0.148713, 0.001765, -0.469389, 0.005162, -0.364831, -0.189866], [-0.355282, 0.439417, 0.023051, 0.841655, -0.059086, 0.054031, 0.711715, 0.62617], [0.281417, -0.385196, -0.009111, 0.054652, -0.395977, -0.30935, -0.364401, -0.358135], [-0.141485, -0.223691, -0.027314, -0.785053, 0.120959, -0.365752, -1.003441, -0.209861], [0.592048, 0.01653, 0.38308, -0.420307, 0.260938, 0.218871, -0.327694, 0.185873]], "network.2.bias": [1.099915, 0.9998, -0.287355, -0.202204, -0.209357, -0.238452, 0.399074, 1.064571], "network.4.weight": [[0.572151, 0.667922, 0.196986, 0.229669, -0.4317, -0.158671, 0.577252, 0.725696], [0.118937, -0.193906, 0.545395, 0.384082, 0.543701, -0.086901, -0.007178, -0.521401], [-0.34733, 0.152592, 0.005943, -0.213706, -0.111773, -0.235395, -0.221955, -0.321266], [-0.219831, -0.349399, 0.172135, -0.237972, -0.27954, -0.245902, 0.271767, -0.530239], [-0.368988, 0.09589, 0.071923, 0.256392, -0.244165, -0.189567, -0.128155, -0.067391], [0.49308, 0.463966, -0.110775, 0.103478, -0.252629, 0.303605, -0.186981, 0.861212], [-0.195205, 0.053893, 0.522439, -0.147537, 0.610225, 0.225879, 0.179656, -0.627609], [0.118585, 0.458815, -0.204506, 0.068376, -0.386708, 0.283885, 0.036866, 0.347156]], "network.4.bias": [0.450391, -0.202538, -0.192971, 0.034328, -0.192684, 0.829731, -0.152038, 0.230208], "network.6.weight": [[0.564413, -0.529784, -0.210255, -0.377733, 0.140415, 0.755286, 0.020801, -0.020592], [0.70065, -0.413279, -0.220306, -0.100543, -0.164293, 0.41649, -0.248665, 0.394003], [-0.443537, -0.207914, 0.279746, -0.33491, -0.242595, 0.139062, -0.077288, 0.235332], [-0.032086, -0.137875, -0.242526, 0.133623, -0.258947, -0.191084, -0.09287, -0.172564], [-0.166274, -0.032024, -0.169387, 0.026219, 0.024342, -0.034809, -0.480389, -0.314925], [-0.178947, 0.491395, 0.022618, -0.273356, -0.537172, -0.674621, 0.281389, 0.069917], [-0.139695, 0.582206, 0.352497, 0.202338, 0.293989, -0.867952, 0.594585, -0.122909], [-0.092731, 0.130497, -0.026145, 0.078665, 0.040088, -0.528748, -0.466538, 0.068055]], "network.6.bias": [0.307938, 0.447669, -0.234141, -0.310698, 0.052046, -0.321453, 0.052176, -0.175017], "network.8.weight": [[-0.300456, 0.019333, -0.282352, 0.291371, 0.238673, 0.035182, -0.169713, 0.293109], [-0.091959, -0.126531, 0.02285, 0.175839, -0.112035, 0.543031, 0.66183, 0.200997], [0.660009, 0.697586, 0.089137, 0.149949, 0.188008, -0.411646, -0.502204, -0.196745], [-0.09597, -0.003262, -0.054916, -0.051813, 0.094712, 0.0904, -0.274368, -0.14529], [0.096672, 0.151043, 0.037291, 0.078396, -0.377483, -0.261309, 0.394418, -0.166439], [-0.439822, 0.076393, -0.388637, 0.341297, -0.208134, -0.374116, -0.098431, 0.336323], [-0.192834, 0.201199, 0.04149, -0.199808, 0.133956, 0.179992, -0.28344, 0.142172], [-0.288631, -0.397923, 0.140568, -0.061804, 0.040679, 0.278745, 0.49284, 0.069817]], "network.8.bias": [-0.2012, -0.220711, 0.229507, -0.028512, -0.308421, -0.572613, -0.258768, 0.107871], "network.10.weight": [[0.101312, -0.451813, 0.519364, -0.044861, -0.127357, 0.208859, -0.284451, -0.499951]], "network.10.bias": [0.333218]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6726362407207489, "train_acc": 0.635, "val_loss": 0.6981194019317627, "val_acc": 0.46}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.5862091779708862, "train_acc": 0.625, "val_loss": 0.47465336322784424, "val_acc": 0.76}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.5246785134077072, "train_acc": 0.715, "val_loss": 0.43130069971084595, "val_acc": 0.82}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.44360245764255524, "train_acc": 0.83, "val_loss": 0.31060945987701416, "val_acc": 0.88}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.3543986678123474, "train_acc": 0.84, "val_loss": 0.9833899140357971, "val_acc": 0.5}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.5300339311361313, "train_acc": 0.715, "val_loss": 0.28358542919158936, "val_acc": 0.9}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.4173599034547806, "train_acc": 0.82, "val_loss": 0.3268507421016693, "val_acc": 0.84}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.36084090173244476, "train_acc": 0.855, "val_loss": 0.2612345814704895, "val_acc": 0.9}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.267160564661026, "train_acc": 0.92, "val_loss": 0.35143861174583435, "val_acc": 0.8}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.33186616003513336, "train_acc": 0.84, "val_loss": 0.3166988492012024, "val_acc": 0.84}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.2688211053609848, "train_acc": 0.865, "val_loss": 0.21474018692970276, "val_acc": 0.9}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.21100181341171265, "train_acc": 0.915, "val_loss": 0.21645504236221313, "val_acc": 0.94}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["contains_abc"], "degraded_stage": {"initial_val_loss": 0.6981194019317627, "final_val_loss": 0.47465336322784424, "initial_val_acc": 0.46, "final_val_acc": 0.76, "best_val_acc": 0.76}, "improved_stage": {"initial_val_loss": 0.43130069971084595, "final_val_loss": 0.21645504236221313, "initial_val_acc": 0.82, "final_val_acc": 0.94, "best_val_acc": 0.94, "best_epoch": 11}, "improvement": 0.17999999999999994, "first_improvement_epoch": 1}}
|
4
|
{"target_pattern": "starts_with", "degraded_accuracy": 0.52, "improved_accuracy": 0.7, "improvement": 0.17999999999999994, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 8547, "learning_rate": 0.09278016261197346, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "starts_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["starts_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.109894,
-0.445163,
-0.339718,
-0.422816,
0.375594
],
[
1.068691,
-0.047804,
-0.225304,
-0.173231,
-0.051718
],
[
-1.389846,
-0.387026,
-0.178974,
-0.583383,
-0.031843
],
[
-1.12625,
-0.054459,
0.238542,
0.191587,
0.852266
],
[
0.403781,
-0.678285,
0.208619,
-0.503622,
-0.200835
],
[
1.238062,
0.406999,
0.294701,
0.149024,
-0.377621
],
[
-0.034888,
-0.568177,
-0.466055,
-0.397139,
-0.699704
],
[
-0.364583,
-0.564707,
-0.40095,
-0.669693,
-0.35
]
],
"network.0.bias": [
-0.311645,
-0.523059,
-0.430238,
0.76656,
0.032952,
0.460556,
-0.203343,
-0.686382
],
"network.2.weight": [
[
0.377989,
-0.48944,
0.531179,
-0.374242,
-0.701774,
-0.035969,
0.158804,
0.336397
],
[
-1.342533,
0.209433,
0.169705,
-0.967001,
-0.319841,
1.240146,
-0.911197,
-0.119791
],
[
-0.104106,
-0.057898,
-0.661113,
0.873614,
0.366809,
-0.859696,
-0.022562,
-0.660989
],
[
0.29433,
-0.420198,
0.641036,
-0.850275,
0.139884,
-0.933371,
0.19756,
0.487257
],
[
-0.39882,
0.449138,
0.110465,
0.833314,
0.192565,
-0.810679,
-0.026087,
-0.558964
],
[
-0.482474,
-0.008503,
0.953159,
-0.790986,
0.116436,
0.932924,
-1.178605,
0.093877
],
[
-0.284512,
-0.447895,
-0.066263,
-0.268312,
-0.307921,
0.656263,
-0.622409,
0.362729
],
[
-0.263097,
-0.281737,
0.048321,
-0.507504,
-0.115313,
0.415644,
-0.870613,
0.410888
]
],
"network.2.bias": [
-0.983928,
-0.075223,
0.511976,
-0.361748,
0.054488,
-0.184803,
-0.439908,
-0.219715
],
"network.4.weight": [
[
0.014826,
0.714887,
-0.395315,
-0.207098,
-0.676643,
0.246306,
-0.222496,
-0.21086
],
[
-0.125935,
-0.627544,
-0.617893,
0.422436,
-0.99034,
-0.847852,
-0.340796,
-0.577703
],
[
-0.049296,
-0.177633,
0.479037,
0.385404,
0.512626,
-0.014388,
-0.090158,
0.000507
],
[
0.154128,
-0.448829,
0.501692,
-0.008253,
0.477391,
-0.23702,
0.30085,
0.190297
],
[
0.143276,
1.002749,
-0.377713,
-0.802062,
-0.711184,
0.021239,
0.362461,
0.023887
],
[
-0.008647,
0.096513,
-0.366996,
-0.238331,
-0.561769,
0.426665,
-0.962425,
-0.280702
],
[
0.042045,
0.822491,
-0.679747,
-0.652838,
-0.436112,
-0.070277,
0.330348,
0.188509
],
[
-0.01464,
-0.519821,
0.512676,
-0.153465,
0.239352,
-0.041972,
0.50298,
-0.049422
]
],
"network.4.bias": [
-0.702317,
-0.607225,
-0.006183,
0.202717,
-0.299441,
-0.505812,
0.116146,
0.64794
],
"network.6.weight": [
[
0.384567,
0.190412,
-0.65876,
-0.244252,
0.98991,
0.177873,
0.162279,
-0.965036
],
[
0.123402,
0.154476,
0.63094,
0.329056,
0.212665,
0.244946,
-0.350253,
0.522999
],
[
0.433291,
0.438276,
-0.721477,
-0.551174,
0.477613,
0.190646,
0.781438,
-0.509766
],
[
0.832379,
-0.3603,
-0.443805,
-0.383027,
0.862796,
-0.029828,
0.13888,
0.103081
],
[
0.105185,
-0.161145,
-1.042356,
-1.150261,
0.122568,
0.232539,
-0.303756,
-0.829379
],
[
0.389201,
0.491674,
-1.140326,
-0.492419,
0.914473,
-0.085393,
0.638811,
-0.824127
],
[
-0.15663,
-0.040971,
-0.189127,
-0.585626,
-0.350465,
-0.321494,
-0.197355,
0.357697
],
[
0.216748,
-0.02261,
-0.384074,
-0.463222,
-0.568033,
0.492622,
-0.143119,
-0.557599
]
],
"network.6.bias": [
-0.153307,
0.18015,
-0.097574,
0.041842,
-0.369995,
-0.184658,
-0.288517,
-0.666002
],
"network.8.weight": [
[
-0.142097,
-0.569972,
-0.374762,
0.547239,
0.113282,
-0.199965,
-0.626121,
-0.394996
],
[
-0.758836,
-0.070928,
-0.597094,
0.576538,
-0.619473,
-0.830265,
0.149455,
-0.335706
],
[
0.420344,
-0.89519,
0.486434,
0.531962,
-0.04874,
0.634191,
-0.323763,
-0.222196
],
[
-0.241434,
-0.201028,
-0.278086,
-0.453527,
-0.26829,
0.053945,
-0.129695,
0.32556
],
[
-0.40505,
-0.662261,
-0.547489,
0.746776,
-0.522674,
-0.558729,
-0.028658,
-0.137383
],
[
-0.622665,
-0.244994,
-0.340544,
-0.00259,
-0.214441,
-0.473615,
-0.357364,
-0.636606
],
[
-0.266348,
-0.226766,
-0.572046,
-0.13196,
-0.207847,
-0.444271,
0.224323,
-0.346571
],
[
-0.005666,
0.539154,
0.248011,
-0.056774,
-0.269726,
-0.106419,
-0.195097,
-0.225452
]
],
"network.8.bias": [
-0.655271,
0.031225,
0.132443,
-0.440885,
-0.124064,
-0.596294,
-0.504965,
-0.004272
],
"network.10.weight": [
[
0.295057,
0.022409,
-0.384944,
-0.218918,
0.055662,
-0.277754,
-0.442372,
0.163889
]
],
"network.10.bias": [
-0.144333
]
}
## Activation Signature
### 0
mean: [-2.412694, -0.186146, -4.529232, 1.226033, -1.592453, 3.212337, -3.970252, -4.874433]
std: [1.706995, 2.094207, 3.533924, 2.609457, 1.958974, 3.058541, 2.495133, 2.719453]
### 2
mean: [-2.200873, 2.404361, -0.684073, -5.107789, -0.742836, 1.456404, 0.869763, 0.051803]
std: [0.925194, 5.090606, 3.695592, 2.946661, 2.898438, 3.753228, 1.701186, 1.543422]
### 4
mean: [0.851811, -6.111690, 0.098836, -0.483729, 2.389559, -1.166036, 2.012403, 0.107034]
std: [4.229554, 5.978304, 2.084941, 2.999646, 5.759611, 1.501090, 4.940903, 2.452155]
### 6
mean: [2.753661, 1.273821, 3.086268, 4.372039, -3.006282, 3.427514, -2.549382, -3.674731]
std: [8.222238, 2.280462, 8.309376, 7.865489, 4.497838, 10.290361, 2.953996, 2.654846]
### 8
mean: [-2.099675, -7.873682, 9.275468, -4.941516, -4.641006, -7.784492, -7.685614, 0.961405]
std: [1.419354, 11.709331, 16.163908, 6.309787, 5.224470, 10.305425, 10.210700, 1.134511]
### 10
mean: [-3.929447]
std: [5.961137]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
starts_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.109894,
-0.445163,
-0.339718,
-0.422816,
0.375594
],
[
1.068691,
-0.047804,
-0.225304,
-0.173231,
-0.051718
],
[
-1.389846,
-0.387026,
-0.178974,
-0.583383,
-0.031843
],
[
-1.12625,
-0.054459,
0.238542,
0.191587,
0.852266
],
[
0.403781,
-0.678285,
0.208619,
-0.503622,
-0.200835
],
[
1.238062,
0.406999,
0.294701,
0.149024,
-0.377621
],
[
-0.034888,
-0.568177,
-0.466055,
-0.397139,
-0.699704
],
[
-0.364583,
-0.564707,
-0.40095,
-0.669693,
-0.35
]
],
"network.0.bias": [
-0.311645,
-0.523059,
-0.430238,
0.76656,
0.032952,
0.460556,
-0.203343,
-0.686382
],
"network.2.weight": [
[
0.377989,
-0.48944,
0.531179,
-0.374242,
-0.701774,
-0.035969,
0.158804,
0.336397
],
[
-1.342533,
0.209433,
0.169705,
-0.967001,
-0.319841,
1.240146,
-0.911197,
-0.119791
],
[
-0.104106,
-0.057898,
-0.661113,
0.873614,
0.366809,
-0.859696,
-0.022562,
-0.660989
],
[
0.29433,
-0.420198,
0.641036,
-0.850275,
0.139884,
-0.933371,
0.19756,
0.487257
],
[
-0.39882,
0.449138,
0.110465,
0.833314,
0.192565,
-0.810679,
-0.026087,
-0.558964
],
[
-0.482474,
-0.008503,
0.953159,
-0.790986,
0.116436,
0.932924,
-1.178605,
0.093877
],
[
-0.284512,
-0.447895,
-0.066263,
-0.268312,
-0.307921,
0.656263,
-0.622409,
0.362729
],
[
-0.263097,
-0.281737,
0.048321,
-0.507504,
-0.115313,
0.415644,
-0.870613,
0.410888
]
],
"network.2.bias": [
-0.983928,
-0.075223,
0.511976,
-0.361748,
0.054488,
-0.184803,
-0.439908,
-0.219715
],
"network.4.weight": [
[
0.014826,
0.714887,
-0.395315,
-0.207098,
-0.676643,
0.246306,
-0.222496,
-0.21086
],
[
-0.125935,
-0.627544,
-0.617893,
0.422436,
-0.99034,
-0.847852,
-0.340796,
-0.577703
],
[
-0.049296,
-0.177633,
0.479037,
0.385404,
0.512626,
-0.014388,
-0.090158,
0.000507
],
[
0.154128,
-0.448829,
0.501692,
-0.008253,
0.477391,
-0.23702,
0.30085,
0.190297
],
[
0.143276,
1.002749,
-0.377713,
-0.802062,
-0.711184,
0.021239,
0.362461,
0.023887
],
[
-0.008647,
0.096513,
-0.366996,
-0.238331,
-0.561769,
0.426665,
-0.962425,
-0.280702
],
[
0.042045,
0.822491,
-0.679747,
-0.652838,
-0.436112,
-0.070277,
0.330348,
0.188509
],
[
-0.01464,
-0.519821,
0.512676,
-0.153465,
0.239352,
-0.041972,
0.50298,
-0.049422
]
],
"network.4.bias": [
-0.702317,
-0.607225,
-0.006183,
0.202717,
-0.299441,
-0.505812,
0.116146,
0.64794
],
"network.6.weight": [
[
0.384567,
0.190412,
-0.65876,
-0.244252,
0.98991,
0.177873,
0.162279,
-0.965036
],
[
0.123402,
0.154476,
0.63094,
0.329056,
0.212665,
0.244946,
-0.350253,
0.522999
],
[
0.433291,
0.438276,
-0.721477,
-0.551174,
0.477613,
0.190646,
0.781438,
-0.509766
],
[
0.832379,
-0.3603,
-0.443805,
-0.383027,
0.862796,
-0.029828,
0.13888,
0.103081
],
[
0.105185,
-0.161145,
-1.042356,
-1.150261,
0.122568,
0.232539,
-0.303756,
-0.829379
],
[
0.389201,
0.491674,
-1.140326,
-0.492419,
0.914473,
-0.085393,
0.638811,
-0.824127
],
[
-0.15663,
-0.040971,
-0.189127,
-0.585626,
-0.350465,
-0.321494,
-0.197355,
0.357697
],
[
0.216748,
-0.02261,
-0.384074,
-0.463222,
-0.568033,
0.492622,
-0.143119,
-0.557599
]
],
"network.6.bias": [
-0.153307,
0.18015,
-0.097574,
0.041842,
-0.369995,
-0.184658,
-0.288517,
-0.666002
],
"network.8.weight": [
[
-0.142097,
-0.569972,
-0.374762,
0.547239,
0.113282,
-0.199965,
-0.626121,
-0.394996
],
[
-0.758836,
-0.070928,
-0.597094,
0.576538,
-0.619473,
-0.830265,
0.149455,
-0.335706
],
[
0.420344,
-0.89519,
0.486434,
0.531962,
-0.04874,
0.634191,
-0.323763,
-0.222196
],
[
-0.241434,
-0.201028,
-0.278086,
-0.453527,
-0.26829,
0.053945,
-0.129695,
0.32556
],
[
-0.40505,
-0.662261,
-0.547489,
0.746776,
-0.522674,
-0.558729,
-0.028658,
-0.137383
],
[
-0.622665,
-0.244994,
-0.340544,
-0.00259,
-0.214441,
-0.473615,
-0.357364,
-0.636606
],
[
-0.266348,
-0.226766,
-0.572046,
-0.13196,
-0.207847,
-0.444271,
0.224323,
-0.346571
],
[
-0.005666,
0.539154,
0.248011,
-0.056774,
-0.269726,
-0.106419,
-0.195097,
-0.225452
]
],
"network.8.bias": [
-0.655271,
0.031225,
0.132443,
-0.440885,
-0.124064,
-0.596294,
-0.504965,
-0.004272
],
"network.10.weight": [
[
0.295057,
0.022409,
-0.384944,
-0.218918,
0.055662,
-0.277754,
-0.442372,
0.163889
]
],
"network.10.bias": [
-0.144333
]
}
## Activation Signature
### 0
mean: [-2.412694, -0.186146, -4.529232, 1.226033, -1.592453, 3.212337, -3.970252, -4.874433]
std: [1.706995, 2.094207, 3.533924, 2.609457, 1.958974, 3.058541, 2.495133, 2.719453]
### 2
mean: [-2.200873, 2.404361, -0.684073, -5.107789, -0.742836, 1.456404, 0.869763, 0.051803]
std: [0.925194, 5.090606, 3.695592, 2.946661, 2.898438, 3.753228, 1.701186, 1.543422]
### 4
mean: [0.851811, -6.111690, 0.098836, -0.483729, 2.389559, -1.166036, 2.012403, 0.107034]
std: [4.229554, 5.978304, 2.084941, 2.999646, 5.759611, 1.501090, 4.940903, 2.452155]
### 6
mean: [2.753661, 1.273821, 3.086268, 4.372039, -3.006282, 3.427514, -2.549382, -3.674731]
std: [8.222238, 2.280462, 8.309376, 7.865489, 4.497838, 10.290361, 2.953996, 2.654846]
### 8
mean: [-2.099675, -7.873682, 9.275468, -4.941516, -4.641006, -7.784492, -7.685614, 0.961405]
std: [1.419354, 11.709331, 16.163908, 6.309787, 5.224470, 10.305425, 10.210700, 1.134511]
### 10
mean: [-3.929447]
std: [5.961137]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
starts_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -2.4126944541931152, "std": 1.7069952487945557}, "1": {"mean": -0.18614648282527924, "std": 2.0942070484161377}, "2": {"mean": -4.529231548309326, "std": 3.533923625946045}, "3": {"mean": 1.2260327339172363, "std": 2.609456777572632}, "4": {"mean": -1.592453122138977, "std": 1.9589735269546509}, "5": {"mean": 3.2123372554779053, "std": 3.0585408210754395}, "6": {"mean": -3.97025203704834, "std": 2.4951331615448}, "7": {"mean": -4.8744330406188965, "std": 2.7194533348083496}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -2.2008728981018066, "std": 0.9251935482025146}, "1": {"mean": 2.40436053276062, "std": 5.090606212615967}, "2": {"mean": -0.684073269367218, "std": 3.6955924034118652}, "3": {"mean": -5.107789039611816, "std": 2.9466607570648193}, "4": {"mean": -0.7428358197212219, "std": 2.8984382152557373}, "5": {"mean": 1.456404447555542, "std": 3.753227949142456}, "6": {"mean": 0.8697634935379028, "std": 1.7011860609054565}, "7": {"mean": 0.05180295184254646, "std": 1.5434224605560303}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.8518109917640686, "std": 4.229553699493408}, "1": {"mean": -6.111689567565918, "std": 5.978304386138916}, "2": {"mean": 0.09883594512939453, "std": 2.0849411487579346}, "3": {"mean": -0.4837293028831482, "std": 2.9996464252471924}, "4": {"mean": 2.3895585536956787, "std": 5.759610652923584}, "5": {"mean": -1.1660360097885132, "std": 1.5010902881622314}, "6": {"mean": 2.0124032497406006, "std": 4.9409027099609375}, "7": {"mean": 0.10703366249799728, "std": 2.452155351638794}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 2.7536611557006836, "std": 8.222237586975098}, "1": {"mean": 1.2738213539123535, "std": 2.2804620265960693}, "2": {"mean": 3.086268424987793, "std": 8.309375762939453}, "3": {"mean": 4.372039318084717, "std": 7.8654890060424805}, "4": {"mean": -3.006282329559326, "std": 4.497838020324707}, "5": {"mean": 3.4275143146514893, "std": 10.290361404418945}, "6": {"mean": -2.549381732940674, "std": 2.953995704650879}, "7": {"mean": -3.6747310161590576, "std": 2.654846429824829}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -2.099674701690674, "std": 1.4193543195724487}, "1": {"mean": -7.873681545257568, "std": 11.709330558776855}, "2": {"mean": 9.275467872619629, "std": 16.163908004760742}, "3": {"mean": -4.941515922546387, "std": 6.309787273406982}, "4": {"mean": -4.641005516052246, "std": 5.224470138549805}, "5": {"mean": -7.784492492675781, "std": 10.305424690246582}, "6": {"mean": -7.685614109039307, "std": 10.210700035095215}, "7": {"mean": 0.9614052772521973, "std": 1.1345109939575195}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -3.9294466972351074, "std": 5.961136817932129}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.109894, -0.445163, -0.339718, -0.422816, 0.375594], [1.068691, -0.047804, -0.225304, -0.173231, -0.051718], [-1.389846, -0.387026, -0.178974, -0.583383, -0.031843], [-1.12625, -0.054459, 0.238542, 0.191587, 0.852266], [0.403781, -0.678285, 0.208619, -0.503622, -0.200835], [1.238062, 0.406999, 0.294701, 0.149024, -0.377621], [-0.034888, -0.568177, -0.466055, -0.397139, -0.699704], [-0.364583, -0.564707, -0.40095, -0.669693, -0.35]], "network.0.bias": [-0.311645, -0.523059, -0.430238, 0.76656, 0.032952, 0.460556, -0.203343, -0.686382], "network.2.weight": [[0.377989, -0.48944, 0.531179, -0.374242, -0.701774, -0.035969, 0.158804, 0.336397], [-1.342533, 0.209433, 0.169705, -0.967001, -0.319841, 1.240146, -0.911197, -0.119791], [-0.104106, -0.057898, -0.661113, 0.873614, 0.366809, -0.859696, -0.022562, -0.660989], [0.29433, -0.420198, 0.641036, -0.850275, 0.139884, -0.933371, 0.19756, 0.487257], [-0.39882, 0.449138, 0.110465, 0.833314, 0.192565, -0.810679, -0.026087, -0.558964], [-0.482474, -0.008503, 0.953159, -0.790986, 0.116436, 0.932924, -1.178605, 0.093877], [-0.284512, -0.447895, -0.066263, -0.268312, -0.307921, 0.656263, -0.622409, 0.362729], [-0.263097, -0.281737, 0.048321, -0.507504, -0.115313, 0.415644, -0.870613, 0.410888]], "network.2.bias": [-0.983928, -0.075223, 0.511976, -0.361748, 0.054488, -0.184803, -0.439908, -0.219715], "network.4.weight": [[0.014826, 0.714887, -0.395315, -0.207098, -0.676643, 0.246306, -0.222496, -0.21086], [-0.125935, -0.627544, -0.617893, 0.422436, -0.99034, -0.847852, -0.340796, -0.577703], [-0.049296, -0.177633, 0.479037, 0.385404, 0.512626, -0.014388, -0.090158, 0.000507], [0.154128, -0.448829, 0.501692, -0.008253, 0.477391, -0.23702, 0.30085, 0.190297], [0.143276, 1.002749, -0.377713, -0.802062, -0.711184, 0.021239, 0.362461, 0.023887], [-0.008647, 0.096513, -0.366996, -0.238331, -0.561769, 0.426665, -0.962425, -0.280702], [0.042045, 0.822491, -0.679747, -0.652838, -0.436112, -0.070277, 0.330348, 0.188509], [-0.01464, -0.519821, 0.512676, -0.153465, 0.239352, -0.041972, 0.50298, -0.049422]], "network.4.bias": [-0.702317, -0.607225, -0.006183, 0.202717, -0.299441, -0.505812, 0.116146, 0.64794], "network.6.weight": [[0.384567, 0.190412, -0.65876, -0.244252, 0.98991, 0.177873, 0.162279, -0.965036], [0.123402, 0.154476, 0.63094, 0.329056, 0.212665, 0.244946, -0.350253, 0.522999], [0.433291, 0.438276, -0.721477, -0.551174, 0.477613, 0.190646, 0.781438, -0.509766], [0.832379, -0.3603, -0.443805, -0.383027, 0.862796, -0.029828, 0.13888, 0.103081], [0.105185, -0.161145, -1.042356, -1.150261, 0.122568, 0.232539, -0.303756, -0.829379], [0.389201, 0.491674, -1.140326, -0.492419, 0.914473, -0.085393, 0.638811, -0.824127], [-0.15663, -0.040971, -0.189127, -0.585626, -0.350465, -0.321494, -0.197355, 0.357697], [0.216748, -0.02261, -0.384074, -0.463222, -0.568033, 0.492622, -0.143119, -0.557599]], "network.6.bias": [-0.153307, 0.18015, -0.097574, 0.041842, -0.369995, -0.184658, -0.288517, -0.666002], "network.8.weight": [[-0.142097, -0.569972, -0.374762, 0.547239, 0.113282, -0.199965, -0.626121, -0.394996], [-0.758836, -0.070928, -0.597094, 0.576538, -0.619473, -0.830265, 0.149455, -0.335706], [0.420344, -0.89519, 0.486434, 0.531962, -0.04874, 0.634191, -0.323763, -0.222196], [-0.241434, -0.201028, -0.278086, -0.453527, -0.26829, 0.053945, -0.129695, 0.32556], [-0.40505, -0.662261, -0.547489, 0.746776, -0.522674, -0.558729, -0.028658, -0.137383], [-0.622665, -0.244994, -0.340544, -0.00259, -0.214441, -0.473615, -0.357364, -0.636606], [-0.266348, -0.226766, -0.572046, -0.13196, -0.207847, -0.444271, 0.224323, -0.346571], [-0.005666, 0.539154, 0.248011, -0.056774, -0.269726, -0.106419, -0.195097, -0.225452]], "network.8.bias": [-0.655271, 0.031225, 0.132443, -0.440885, -0.124064, -0.596294, -0.504965, -0.004272], "network.10.weight": [[0.295057, 0.022409, -0.384944, -0.218918, 0.055662, -0.277754, -0.442372, 0.163889]], "network.10.bias": [-0.144333]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.689319372177124, "train_acc": 0.56, "val_loss": 0.6855395436286926, "val_acc": 0.56}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6841188669204712, "train_acc": 0.56, "val_loss": 0.6717982292175293, "val_acc": 0.56}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.643918365240097, "train_acc": 0.56, "val_loss": 0.5862912535667419, "val_acc": 0.52}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.6517999768257141, "train_acc": 0.58, "val_loss": 0.8351137042045593, "val_acc": 0.6}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.6903351843357086, "train_acc": 0.655, "val_loss": 0.6305068731307983, "val_acc": 0.64}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.6242150664329529, "train_acc": 0.695, "val_loss": 0.5733755230903625, "val_acc": 0.66}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.5529791116714478, "train_acc": 0.66, "val_loss": 0.512876570224762, "val_acc": 0.66}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.5070566385984421, "train_acc": 0.73, "val_loss": 0.5056356191635132, "val_acc": 0.7}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.47154945135116577, "train_acc": 0.78, "val_loss": 0.5270170569419861, "val_acc": 0.7}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.48196378350257874, "train_acc": 0.785, "val_loss": 0.5336831212043762, "val_acc": 0.7}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.44655831158161163, "train_acc": 0.79, "val_loss": 0.5136688351631165, "val_acc": 0.7}], "summary": {"total_epochs": 11, "degraded_epochs": 3, "improved_epochs": 8, "patterns": ["starts_with"], "degraded_stage": {"initial_val_loss": 0.6855395436286926, "final_val_loss": 0.5862912535667419, "initial_val_acc": 0.56, "final_val_acc": 0.52, "best_val_acc": 0.52}, "improved_stage": {"initial_val_loss": 0.8351137042045593, "final_val_loss": 0.5136688351631165, "initial_val_acc": 0.6, "final_val_acc": 0.7, "best_val_acc": 0.7, "best_epoch": 7}, "improvement": 0.17999999999999994, "first_improvement_epoch": 2}}
|
5
|
{"target_pattern": "increasing_pairs", "degraded_accuracy": 0.64, "improved_accuracy": 0.88, "improvement": 0.24, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 9859, "learning_rate": 0.015384002471586396, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "increasing_pairs", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["increasing_pairs"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.554763,
0.117104,
0.334357,
0.124165,
-0.458343
],
[
-0.092501,
-0.084691,
-0.154902,
-0.472829,
-0.378697
],
[
0.81391,
0.16854,
-0.010797,
0.24896,
-0.354764
],
[
0.574054,
-0.023014,
0.063039,
-0.405028,
0.19252
],
[
0.125637,
0.100255,
0.693697,
-0.529964,
-0.138561
]
],
"network.0.bias": [
-0.066843,
0.095595,
0.685387,
0.134922,
-0.055535
],
"network.2.weight": [
[
0.476523,
-0.2156,
0.58778,
0.304088,
0.210094
],
[
-0.118559,
-0.003428,
-0.258041,
0.053522,
0.559965
],
[
0.559127,
-0.108064,
0.59571,
0.114789,
0.446204
],
[
0.753622,
0.170474,
-0.024526,
0.250312,
0.679271
],
[
0.365231,
-0.130418,
-0.077753,
0.377768,
0.173806
]
],
"network.2.bias": [
-0.053565,
0.029857,
-0.041491,
-0.097665,
-0.301038
],
"network.4.weight": [
[
0.028013,
0.312583,
-0.064444,
-0.171507,
0.186128
],
[
0.180827,
0.054785,
-0.261857,
-0.272733,
0.337052
],
[
0.808708,
0.611157,
0.153663,
0.7088,
0.126911
],
[
0.505599,
0.300671,
0.649823,
-0.164533,
0.39908
],
[
0.610189,
-0.020265,
0.58929,
0.221495,
0.592952
]
],
"network.4.bias": [
-0.351487,
-0.300211,
-0.007463,
-0.069882,
-0.002043
],
"network.6.weight": [
[
0.390837,
-0.374882,
-0.283566,
-0.381474,
-0.334873
],
[
-0.390806,
-0.397559,
-0.322201,
0.124676,
0.143966
],
[
0.405974,
0.299639,
0.574906,
0.411648,
0.739944
],
[
-0.213986,
-0.03172,
-0.024501,
-0.677509,
0.176431
],
[
0.219681,
0.139951,
-0.160216,
0.69208,
0.688648
]
],
"network.6.bias": [
-0.026605,
-0.105723,
-0.120646,
0.420496,
0.052764
],
"network.8.weight": [
[
-0.366589,
0.146991,
-0.62858,
0.355391,
-0.207267
]
],
"network.8.bias": [
0.16007
]
}
## Activation Signature
### 0
mean: [1.236330, -1.978433, 2.079155, 0.307927, 0.425188]
std: [1.603406, 1.457691, 1.838132, 1.570009, 1.827139]
### 2
mean: [2.270212, -0.106212, 2.493728, 1.728810, 0.491658]
std: [2.074491, 0.773071, 2.195636, 1.854704, 0.902480]
### 4
mean: [-0.558187, -0.803964, 3.678314, 2.725924, 3.587975]
std: [0.241712, 0.438023, 3.411029, 2.507657, 3.385724]
### 6
mean: [-3.311640, -0.434275, 5.771737, -0.884465, 3.821954]
std: [3.047605, 0.335323, 5.481179, 1.183519, 3.526555]
### 8
mean: [-4.237027]
std: [4.196862]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
increasing_pairs
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.554763,
0.117104,
0.334357,
0.124165,
-0.458343
],
[
-0.092501,
-0.084691,
-0.154902,
-0.472829,
-0.378697
],
[
0.81391,
0.16854,
-0.010797,
0.24896,
-0.354764
],
[
0.574054,
-0.023014,
0.063039,
-0.405028,
0.19252
],
[
0.125637,
0.100255,
0.693697,
-0.529964,
-0.138561
]
],
"network.0.bias": [
-0.066843,
0.095595,
0.685387,
0.134922,
-0.055535
],
"network.2.weight": [
[
0.476523,
-0.2156,
0.58778,
0.304088,
0.210094
],
[
-0.118559,
-0.003428,
-0.258041,
0.053522,
0.559965
],
[
0.559127,
-0.108064,
0.59571,
0.114789,
0.446204
],
[
0.753622,
0.170474,
-0.024526,
0.250312,
0.679271
],
[
0.365231,
-0.130418,
-0.077753,
0.377768,
0.173806
]
],
"network.2.bias": [
-0.053565,
0.029857,
-0.041491,
-0.097665,
-0.301038
],
"network.4.weight": [
[
0.028013,
0.312583,
-0.064444,
-0.171507,
0.186128
],
[
0.180827,
0.054785,
-0.261857,
-0.272733,
0.337052
],
[
0.808708,
0.611157,
0.153663,
0.7088,
0.126911
],
[
0.505599,
0.300671,
0.649823,
-0.164533,
0.39908
],
[
0.610189,
-0.020265,
0.58929,
0.221495,
0.592952
]
],
"network.4.bias": [
-0.351487,
-0.300211,
-0.007463,
-0.069882,
-0.002043
],
"network.6.weight": [
[
0.390837,
-0.374882,
-0.283566,
-0.381474,
-0.334873
],
[
-0.390806,
-0.397559,
-0.322201,
0.124676,
0.143966
],
[
0.405974,
0.299639,
0.574906,
0.411648,
0.739944
],
[
-0.213986,
-0.03172,
-0.024501,
-0.677509,
0.176431
],
[
0.219681,
0.139951,
-0.160216,
0.69208,
0.688648
]
],
"network.6.bias": [
-0.026605,
-0.105723,
-0.120646,
0.420496,
0.052764
],
"network.8.weight": [
[
-0.366589,
0.146991,
-0.62858,
0.355391,
-0.207267
]
],
"network.8.bias": [
0.16007
]
}
## Activation Signature
### 0
mean: [1.236330, -1.978433, 2.079155, 0.307927, 0.425188]
std: [1.603406, 1.457691, 1.838132, 1.570009, 1.827139]
### 2
mean: [2.270212, -0.106212, 2.493728, 1.728810, 0.491658]
std: [2.074491, 0.773071, 2.195636, 1.854704, 0.902480]
### 4
mean: [-0.558187, -0.803964, 3.678314, 2.725924, 3.587975]
std: [0.241712, 0.438023, 3.411029, 2.507657, 3.385724]
### 6
mean: [-3.311640, -0.434275, 5.771737, -0.884465, 3.821954]
std: [3.047605, 0.335323, 5.481179, 1.183519, 3.526555]
### 8
mean: [-4.237027]
std: [4.196862]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
increasing_pairs
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.2363301515579224, "std": 1.6034057140350342}, "1": {"mean": -1.978432536125183, "std": 1.4576908349990845}, "2": {"mean": 2.079155206680298, "std": 1.8381322622299194}, "3": {"mean": 0.3079267144203186, "std": 1.5700089931488037}, "4": {"mean": 0.4251878559589386, "std": 1.827139139175415}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 2.270212411880493, "std": 2.074490785598755}, "1": {"mean": -0.1062115952372551, "std": 0.7730712294578552}, "2": {"mean": 2.4937281608581543, "std": 2.1956355571746826}, "3": {"mean": 1.7288100719451904, "std": 1.8547039031982422}, "4": {"mean": 0.49165818095207214, "std": 0.9024804830551147}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.5581866502761841, "std": 0.24171212315559387}, "1": {"mean": -0.8039635419845581, "std": 0.43802255392074585}, "2": {"mean": 3.678314208984375, "std": 3.4110286235809326}, "3": {"mean": 2.725924253463745, "std": 2.5076565742492676}, "4": {"mean": 3.587974786758423, "std": 3.38572359085083}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -3.3116397857666016, "std": 3.047604560852051}, "1": {"mean": -0.43427467346191406, "std": 0.33532339334487915}, "2": {"mean": 5.7717366218566895, "std": 5.481179237365723}, "3": {"mean": -0.8844650387763977, "std": 1.1835191249847412}, "4": {"mean": 3.8219542503356934, "std": 3.526555299758911}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -4.237026691436768, "std": 4.19686222076416}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.554763, 0.117104, 0.334357, 0.124165, -0.458343], [-0.092501, -0.084691, -0.154902, -0.472829, -0.378697], [0.81391, 0.16854, -0.010797, 0.24896, -0.354764], [0.574054, -0.023014, 0.063039, -0.405028, 0.19252], [0.125637, 0.100255, 0.693697, -0.529964, -0.138561]], "network.0.bias": [-0.066843, 0.095595, 0.685387, 0.134922, -0.055535], "network.2.weight": [[0.476523, -0.2156, 0.58778, 0.304088, 0.210094], [-0.118559, -0.003428, -0.258041, 0.053522, 0.559965], [0.559127, -0.108064, 0.59571, 0.114789, 0.446204], [0.753622, 0.170474, -0.024526, 0.250312, 0.679271], [0.365231, -0.130418, -0.077753, 0.377768, 0.173806]], "network.2.bias": [-0.053565, 0.029857, -0.041491, -0.097665, -0.301038], "network.4.weight": [[0.028013, 0.312583, -0.064444, -0.171507, 0.186128], [0.180827, 0.054785, -0.261857, -0.272733, 0.337052], [0.808708, 0.611157, 0.153663, 0.7088, 0.126911], [0.505599, 0.300671, 0.649823, -0.164533, 0.39908], [0.610189, -0.020265, 0.58929, 0.221495, 0.592952]], "network.4.bias": [-0.351487, -0.300211, -0.007463, -0.069882, -0.002043], "network.6.weight": [[0.390837, -0.374882, -0.283566, -0.381474, -0.334873], [-0.390806, -0.397559, -0.322201, 0.124676, 0.143966], [0.405974, 0.299639, 0.574906, 0.411648, 0.739944], [-0.213986, -0.03172, -0.024501, -0.677509, 0.176431], [0.219681, 0.139951, -0.160216, 0.69208, 0.688648]], "network.6.bias": [-0.026605, -0.105723, -0.120646, 0.420496, 0.052764], "network.8.weight": [[-0.366589, 0.146991, -0.62858, 0.355391, -0.207267]], "network.8.bias": [0.16007]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6981912553310394, "train_acc": 0.455, "val_loss": 0.6990344524383545, "val_acc": 0.36}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6930713355541229, "train_acc": 0.495, "val_loss": 0.6893063187599182, "val_acc": 0.64}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6892660558223724, "train_acc": 0.545, "val_loss": 0.6796994209289551, "val_acc": 0.64}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6851053237915039, "train_acc": 0.545, "val_loss": 0.6671839952468872, "val_acc": 0.64}, {"stage": "degraded", "epoch": 4, "global_epoch": 4, "train_loss": 0.6790561974048615, "train_acc": 0.545, "val_loss": 0.6494336128234863, "val_acc": 0.64}, {"stage": "improved", "epoch": 0, "global_epoch": 5, "train_loss": 0.6754173636436462, "train_acc": 0.465, "val_loss": 0.6389731764793396, "val_acc": 0.64}, {"stage": "improved", "epoch": 1, "global_epoch": 6, "train_loss": 0.6588297784328461, "train_acc": 0.465, "val_loss": 0.6094434261322021, "val_acc": 0.64}, {"stage": "improved", "epoch": 2, "global_epoch": 7, "train_loss": 0.6320454180240631, "train_acc": 0.465, "val_loss": 0.5648959279060364, "val_acc": 0.64}, {"stage": "improved", "epoch": 3, "global_epoch": 8, "train_loss": 0.59026238322258, "train_acc": 0.465, "val_loss": 0.5130147337913513, "val_acc": 0.76}, {"stage": "improved", "epoch": 4, "global_epoch": 9, "train_loss": 0.5528412461280823, "train_acc": 0.805, "val_loss": 0.4637288749217987, "val_acc": 0.78}, {"stage": "improved", "epoch": 5, "global_epoch": 10, "train_loss": 0.5168490707874298, "train_acc": 0.82, "val_loss": 0.42328017950057983, "val_acc": 0.82}, {"stage": "improved", "epoch": 6, "global_epoch": 11, "train_loss": 0.4810015708208084, "train_acc": 0.845, "val_loss": 0.3928557336330414, "val_acc": 0.82}, {"stage": "improved", "epoch": 7, "global_epoch": 12, "train_loss": 0.44770801067352295, "train_acc": 0.88, "val_loss": 0.36580657958984375, "val_acc": 0.86}, {"stage": "improved", "epoch": 8, "global_epoch": 13, "train_loss": 0.41161610186100006, "train_acc": 0.89, "val_loss": 0.3435382544994354, "val_acc": 0.88}, {"stage": "improved", "epoch": 9, "global_epoch": 14, "train_loss": 0.39002910256385803, "train_acc": 0.905, "val_loss": 0.32937926054000854, "val_acc": 0.88}], "summary": {"total_epochs": 15, "degraded_epochs": 5, "improved_epochs": 10, "patterns": ["increasing_pairs"], "degraded_stage": {"initial_val_loss": 0.6990344524383545, "final_val_loss": 0.6494336128234863, "initial_val_acc": 0.36, "final_val_acc": 0.64, "best_val_acc": 0.64}, "improved_stage": {"initial_val_loss": 0.6389731764793396, "final_val_loss": 0.32937926054000854, "initial_val_acc": 0.64, "final_val_acc": 0.88, "best_val_acc": 0.88, "best_epoch": 13}, "improvement": 0.24, "first_improvement_epoch": 4}}
|
6
|
{"target_pattern": "sorted_descending", "degraded_accuracy": 0.56, "improved_accuracy": 0.94, "improvement": 0.3799999999999999, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 4854, "learning_rate": 0.09414589333639692, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_descending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_descending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.000826,
-0.651648,
0.075237,
-0.055185,
-0.17648
],
[
-0.476082,
-0.324751,
0.120412,
0.598693,
0.640992
],
[
-0.07785,
-0.401569,
0.120796,
0.65813,
-0.371936
],
[
0.915331,
-0.061175,
-0.022539,
0.28711,
-0.221207
],
[
-0.337218,
-0.139821,
-0.076714,
0.635173,
-0.009794
],
[
0.005497,
-0.442614,
-0.091421,
-0.409622,
-0.321745
],
[
-0.394041,
-0.048166,
-0.41158,
0.249937,
-0.339947
],
[
-0.492416,
0.054176,
-0.014026,
0.230689,
0.768552
]
],
"network.0.bias": [
-0.70221,
0.426381,
0.320511,
0.066441,
-0.249641,
-0.352013,
-0.002023,
0.22375
],
"network.2.weight": [
[
-0.479875,
-0.004748,
-0.269473,
-0.361379,
-0.235998,
0.046999,
-0.361028,
-0.242454
],
[
-0.231045,
-0.175521,
0.493511,
0.256746,
0.035432,
-0.260977,
-0.1468,
-0.180409
],
[
-0.336956,
0.154934,
0.589417,
-0.271957,
0.714244,
0.177662,
-0.211854,
0.212685
],
[
0.048361,
-0.192044,
-0.203077,
-0.56429,
-0.305309,
0.003926,
-0.289593,
-0.338048
],
[
0.466631,
0.567567,
0.573734,
-0.524243,
0.562097,
-0.114418,
0.068823,
0.280073
],
[
-0.251753,
0.4857,
0.73858,
-0.233988,
0.669427,
-0.33238,
0.050318,
0.443549
],
[
-0.319519,
0.655106,
0.527639,
-0.466749,
0.945601,
-0.12036,
-0.002449,
0.202471
],
[
-0.059327,
0.45667,
0.461268,
-0.094024,
0.672168,
-0.351053,
-0.260447,
0.145481
]
],
"network.2.bias": [
-0.205857,
0.01284,
-0.466352,
-0.509702,
0.18246,
-0.074699,
0.057725,
-0.212566
],
"network.4.weight": [
[
0.005203,
0.211276,
0.282539,
0.008915,
0.672909,
0.671312,
0.411983,
0.498315
],
[
0.048379,
0.435476,
0.20126,
0.535599,
0.767431,
0.766208,
0.420545,
0.40993
],
[
0.385767,
-0.2348,
-0.00738,
0.431539,
0.217415,
-0.036308,
-0.452209,
-0.39082
],
[
0.239516,
0.435118,
-0.049921,
0.009443,
0.8819,
0.429083,
0.411728,
-0.067539
],
[
-0.153422,
-0.481724,
0.091868,
-0.226576,
-0.233929,
-0.152361,
-0.350272,
-0.065038
],
[
0.328131,
0.217926,
-0.40414,
0.293737,
0.315777,
0.282055,
-0.281672,
-0.108299
],
[
-0.18699,
-0.097596,
-0.013128,
0.169829,
-0.184505,
0.147711,
-0.048135,
0.027359
],
[
0.252833,
0.03447,
-0.283017,
-0.071841,
-0.36452,
-0.025828,
-0.037222,
0.088584
]
],
"network.4.bias": [
-0.136538,
0.002579,
0.158569,
-0.271845,
-0.503371,
-0.187965,
-0.311589,
-0.39375
],
"network.6.weight": [
[
0.301573,
0.420408,
-0.037716,
0.082517,
-0.233638,
0.018079,
0.001423,
0.126851
],
[
-0.131862,
-0.331597,
0.086474,
0.234137,
0.3095,
-0.238337,
-0.302186,
-0.232704
],
[
-0.2999,
-0.053468,
0.308114,
0.23342,
-0.35204,
-0.337733,
0.283981,
-0.339767
],
[
0.115329,
-0.293754,
-0.257193,
-0.106409,
-0.254996,
-0.044153,
0.042363,
0.280874
],
[
-0.361535,
-0.221839,
-0.313945,
-0.529774,
-0.052776,
-0.555494,
-0.069123,
0.067854
],
[
-0.180653,
-0.156951,
-0.107795,
0.013709,
0.282067,
-0.025399,
0.111302,
-0.226032
],
[
0.434268,
0.310742,
-0.164965,
0.402346,
-0.098587,
0.085635,
-0.010884,
-0.131857
],
[
-0.074418,
-0.624164,
-0.382713,
-0.340753,
-0.270624,
-0.276418,
0.274616,
-0.082277
]
],
"network.6.bias": [
-0.335322,
-0.257928,
-0.145847,
-0.320486,
-0.035375,
-0.2139,
-0.343605,
-0.03992
],
"network.8.weight": [
[
-0.325056,
-0.249554,
0.131497,
-0.351041,
0.094175,
-0.212405,
-0.33603,
0.010686
]
],
"network.8.bias": [
0.09692
]
}
## Activation Signature
### 0
mean: [-2.067467, 1.568103, 0.696814, 1.389832, 0.265691, -2.617912, -1.323192, 1.122968]
std: [1.261666, 2.065996, 1.418738, 1.819887, 1.540668, 1.581443, 1.653070, 1.718640]
### 2
mean: [-1.527498, 0.306411, 0.766620, -2.576658, 1.783218, 2.250957, 2.035807, 1.582435]
std: [0.918129, 0.849580, 1.831011, 1.384624, 2.587544, 2.658555, 2.898658, 1.960253]
### 4
mean: [5.083102, 5.541123, -1.300442, 3.602600, -2.416031, -0.023027, -0.471104, -1.449954]
std: [5.549639, 5.743898, 1.539738, 3.872135, 1.811374, 0.179122, 0.149291, 1.211079]
### 6
mean: [3.826284, -1.934965, -1.143247, -1.749979, -5.049466, -1.955072, 5.041957, -5.125238]
std: [4.406409, 1.728066, 1.065404, 1.456517, 5.315849, 1.849535, 5.748356, 5.307633]
### 8
mean: [-2.851967]
std: [3.354157]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_descending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.000826,
-0.651648,
0.075237,
-0.055185,
-0.17648
],
[
-0.476082,
-0.324751,
0.120412,
0.598693,
0.640992
],
[
-0.07785,
-0.401569,
0.120796,
0.65813,
-0.371936
],
[
0.915331,
-0.061175,
-0.022539,
0.28711,
-0.221207
],
[
-0.337218,
-0.139821,
-0.076714,
0.635173,
-0.009794
],
[
0.005497,
-0.442614,
-0.091421,
-0.409622,
-0.321745
],
[
-0.394041,
-0.048166,
-0.41158,
0.249937,
-0.339947
],
[
-0.492416,
0.054176,
-0.014026,
0.230689,
0.768552
]
],
"network.0.bias": [
-0.70221,
0.426381,
0.320511,
0.066441,
-0.249641,
-0.352013,
-0.002023,
0.22375
],
"network.2.weight": [
[
-0.479875,
-0.004748,
-0.269473,
-0.361379,
-0.235998,
0.046999,
-0.361028,
-0.242454
],
[
-0.231045,
-0.175521,
0.493511,
0.256746,
0.035432,
-0.260977,
-0.1468,
-0.180409
],
[
-0.336956,
0.154934,
0.589417,
-0.271957,
0.714244,
0.177662,
-0.211854,
0.212685
],
[
0.048361,
-0.192044,
-0.203077,
-0.56429,
-0.305309,
0.003926,
-0.289593,
-0.338048
],
[
0.466631,
0.567567,
0.573734,
-0.524243,
0.562097,
-0.114418,
0.068823,
0.280073
],
[
-0.251753,
0.4857,
0.73858,
-0.233988,
0.669427,
-0.33238,
0.050318,
0.443549
],
[
-0.319519,
0.655106,
0.527639,
-0.466749,
0.945601,
-0.12036,
-0.002449,
0.202471
],
[
-0.059327,
0.45667,
0.461268,
-0.094024,
0.672168,
-0.351053,
-0.260447,
0.145481
]
],
"network.2.bias": [
-0.205857,
0.01284,
-0.466352,
-0.509702,
0.18246,
-0.074699,
0.057725,
-0.212566
],
"network.4.weight": [
[
0.005203,
0.211276,
0.282539,
0.008915,
0.672909,
0.671312,
0.411983,
0.498315
],
[
0.048379,
0.435476,
0.20126,
0.535599,
0.767431,
0.766208,
0.420545,
0.40993
],
[
0.385767,
-0.2348,
-0.00738,
0.431539,
0.217415,
-0.036308,
-0.452209,
-0.39082
],
[
0.239516,
0.435118,
-0.049921,
0.009443,
0.8819,
0.429083,
0.411728,
-0.067539
],
[
-0.153422,
-0.481724,
0.091868,
-0.226576,
-0.233929,
-0.152361,
-0.350272,
-0.065038
],
[
0.328131,
0.217926,
-0.40414,
0.293737,
0.315777,
0.282055,
-0.281672,
-0.108299
],
[
-0.18699,
-0.097596,
-0.013128,
0.169829,
-0.184505,
0.147711,
-0.048135,
0.027359
],
[
0.252833,
0.03447,
-0.283017,
-0.071841,
-0.36452,
-0.025828,
-0.037222,
0.088584
]
],
"network.4.bias": [
-0.136538,
0.002579,
0.158569,
-0.271845,
-0.503371,
-0.187965,
-0.311589,
-0.39375
],
"network.6.weight": [
[
0.301573,
0.420408,
-0.037716,
0.082517,
-0.233638,
0.018079,
0.001423,
0.126851
],
[
-0.131862,
-0.331597,
0.086474,
0.234137,
0.3095,
-0.238337,
-0.302186,
-0.232704
],
[
-0.2999,
-0.053468,
0.308114,
0.23342,
-0.35204,
-0.337733,
0.283981,
-0.339767
],
[
0.115329,
-0.293754,
-0.257193,
-0.106409,
-0.254996,
-0.044153,
0.042363,
0.280874
],
[
-0.361535,
-0.221839,
-0.313945,
-0.529774,
-0.052776,
-0.555494,
-0.069123,
0.067854
],
[
-0.180653,
-0.156951,
-0.107795,
0.013709,
0.282067,
-0.025399,
0.111302,
-0.226032
],
[
0.434268,
0.310742,
-0.164965,
0.402346,
-0.098587,
0.085635,
-0.010884,
-0.131857
],
[
-0.074418,
-0.624164,
-0.382713,
-0.340753,
-0.270624,
-0.276418,
0.274616,
-0.082277
]
],
"network.6.bias": [
-0.335322,
-0.257928,
-0.145847,
-0.320486,
-0.035375,
-0.2139,
-0.343605,
-0.03992
],
"network.8.weight": [
[
-0.325056,
-0.249554,
0.131497,
-0.351041,
0.094175,
-0.212405,
-0.33603,
0.010686
]
],
"network.8.bias": [
0.09692
]
}
## Activation Signature
### 0
mean: [-2.067467, 1.568103, 0.696814, 1.389832, 0.265691, -2.617912, -1.323192, 1.122968]
std: [1.261666, 2.065996, 1.418738, 1.819887, 1.540668, 1.581443, 1.653070, 1.718640]
### 2
mean: [-1.527498, 0.306411, 0.766620, -2.576658, 1.783218, 2.250957, 2.035807, 1.582435]
std: [0.918129, 0.849580, 1.831011, 1.384624, 2.587544, 2.658555, 2.898658, 1.960253]
### 4
mean: [5.083102, 5.541123, -1.300442, 3.602600, -2.416031, -0.023027, -0.471104, -1.449954]
std: [5.549639, 5.743898, 1.539738, 3.872135, 1.811374, 0.179122, 0.149291, 1.211079]
### 6
mean: [3.826284, -1.934965, -1.143247, -1.749979, -5.049466, -1.955072, 5.041957, -5.125238]
std: [4.406409, 1.728066, 1.065404, 1.456517, 5.315849, 1.849535, 5.748356, 5.307633]
### 8
mean: [-2.851967]
std: [3.354157]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_descending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -2.067466974258423, "std": 1.26166570186615}, "1": {"mean": 1.5681030750274658, "std": 2.065995931625366}, "2": {"mean": 0.6968143582344055, "std": 1.4187380075454712}, "3": {"mean": 1.3898323774337769, "std": 1.8198871612548828}, "4": {"mean": 0.26569145917892456, "std": 1.5406683683395386}, "5": {"mean": -2.6179120540618896, "std": 1.5814430713653564}, "6": {"mean": -1.3231918811798096, "std": 1.6530697345733643}, "7": {"mean": 1.122968077659607, "std": 1.7186404466629028}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -1.527498483657837, "std": 0.9181292057037354}, "1": {"mean": 0.3064112663269043, "std": 0.8495795130729675}, "2": {"mean": 0.7666196227073669, "std": 1.8310106992721558}, "3": {"mean": -2.576658248901367, "std": 1.3846238851547241}, "4": {"mean": 1.783218264579773, "std": 2.5875439643859863}, "5": {"mean": 2.2509565353393555, "std": 2.658554792404175}, "6": {"mean": 2.0358071327209473, "std": 2.898658037185669}, "7": {"mean": 1.5824354887008667, "std": 1.9602530002593994}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 5.083102226257324, "std": 5.5496392250061035}, "1": {"mean": 5.541122913360596, "std": 5.743897914886475}, "2": {"mean": -1.300441861152649, "std": 1.5397382974624634}, "3": {"mean": 3.60260009765625, "std": 3.8721349239349365}, "4": {"mean": -2.4160311222076416, "std": 1.8113744258880615}, "5": {"mean": -0.02302665263414383, "std": 0.1791224181652069}, "6": {"mean": -0.4711041748523712, "std": 0.14929130673408508}, "7": {"mean": -1.4499539136886597, "std": 1.2110792398452759}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 3.826284170150757, "std": 4.406408786773682}, "1": {"mean": -1.9349651336669922, "std": 1.7280656099319458}, "2": {"mean": -1.143247365951538, "std": 1.0654044151306152}, "3": {"mean": -1.7499791383743286, "std": 1.4565167427062988}, "4": {"mean": -5.049465656280518, "std": 5.315849304199219}, "5": {"mean": -1.9550718069076538, "std": 1.8495352268218994}, "6": {"mean": 5.041957378387451, "std": 5.748355865478516}, "7": {"mean": -5.125237941741943, "std": 5.307632923126221}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -2.851966619491577, "std": 3.354156732559204}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.000826, -0.651648, 0.075237, -0.055185, -0.17648], [-0.476082, -0.324751, 0.120412, 0.598693, 0.640992], [-0.07785, -0.401569, 0.120796, 0.65813, -0.371936], [0.915331, -0.061175, -0.022539, 0.28711, -0.221207], [-0.337218, -0.139821, -0.076714, 0.635173, -0.009794], [0.005497, -0.442614, -0.091421, -0.409622, -0.321745], [-0.394041, -0.048166, -0.41158, 0.249937, -0.339947], [-0.492416, 0.054176, -0.014026, 0.230689, 0.768552]], "network.0.bias": [-0.70221, 0.426381, 0.320511, 0.066441, -0.249641, -0.352013, -0.002023, 0.22375], "network.2.weight": [[-0.479875, -0.004748, -0.269473, -0.361379, -0.235998, 0.046999, -0.361028, -0.242454], [-0.231045, -0.175521, 0.493511, 0.256746, 0.035432, -0.260977, -0.1468, -0.180409], [-0.336956, 0.154934, 0.589417, -0.271957, 0.714244, 0.177662, -0.211854, 0.212685], [0.048361, -0.192044, -0.203077, -0.56429, -0.305309, 0.003926, -0.289593, -0.338048], [0.466631, 0.567567, 0.573734, -0.524243, 0.562097, -0.114418, 0.068823, 0.280073], [-0.251753, 0.4857, 0.73858, -0.233988, 0.669427, -0.33238, 0.050318, 0.443549], [-0.319519, 0.655106, 0.527639, -0.466749, 0.945601, -0.12036, -0.002449, 0.202471], [-0.059327, 0.45667, 0.461268, -0.094024, 0.672168, -0.351053, -0.260447, 0.145481]], "network.2.bias": [-0.205857, 0.01284, -0.466352, -0.509702, 0.18246, -0.074699, 0.057725, -0.212566], "network.4.weight": [[0.005203, 0.211276, 0.282539, 0.008915, 0.672909, 0.671312, 0.411983, 0.498315], [0.048379, 0.435476, 0.20126, 0.535599, 0.767431, 0.766208, 0.420545, 0.40993], [0.385767, -0.2348, -0.00738, 0.431539, 0.217415, -0.036308, -0.452209, -0.39082], [0.239516, 0.435118, -0.049921, 0.009443, 0.8819, 0.429083, 0.411728, -0.067539], [-0.153422, -0.481724, 0.091868, -0.226576, -0.233929, -0.152361, -0.350272, -0.065038], [0.328131, 0.217926, -0.40414, 0.293737, 0.315777, 0.282055, -0.281672, -0.108299], [-0.18699, -0.097596, -0.013128, 0.169829, -0.184505, 0.147711, -0.048135, 0.027359], [0.252833, 0.03447, -0.283017, -0.071841, -0.36452, -0.025828, -0.037222, 0.088584]], "network.4.bias": [-0.136538, 0.002579, 0.158569, -0.271845, -0.503371, -0.187965, -0.311589, -0.39375], "network.6.weight": [[0.301573, 0.420408, -0.037716, 0.082517, -0.233638, 0.018079, 0.001423, 0.126851], [-0.131862, -0.331597, 0.086474, 0.234137, 0.3095, -0.238337, -0.302186, -0.232704], [-0.2999, -0.053468, 0.308114, 0.23342, -0.35204, -0.337733, 0.283981, -0.339767], [0.115329, -0.293754, -0.257193, -0.106409, -0.254996, -0.044153, 0.042363, 0.280874], [-0.361535, -0.221839, -0.313945, -0.529774, -0.052776, -0.555494, -0.069123, 0.067854], [-0.180653, -0.156951, -0.107795, 0.013709, 0.282067, -0.025399, 0.111302, -0.226032], [0.434268, 0.310742, -0.164965, 0.402346, -0.098587, 0.085635, -0.010884, -0.131857], [-0.074418, -0.624164, -0.382713, -0.340753, -0.270624, -0.276418, 0.274616, -0.082277]], "network.6.bias": [-0.335322, -0.257928, -0.145847, -0.320486, -0.035375, -0.2139, -0.343605, -0.03992], "network.8.weight": [[-0.325056, -0.249554, 0.131497, -0.351041, 0.094175, -0.212405, -0.33603, 0.010686]], "network.8.bias": [0.09692]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6972275674343109, "train_acc": 0.565, "val_loss": 0.6631535291671753, "val_acc": 0.56}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6444984972476959, "train_acc": 0.565, "val_loss": 0.5396676063537598, "val_acc": 0.56}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.6284622251987457, "train_acc": 0.485, "val_loss": 0.5814832448959351, "val_acc": 0.56}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.5726710259914398, "train_acc": 0.51, "val_loss": 0.3724343180656433, "val_acc": 0.94}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.47788481414318085, "train_acc": 0.885, "val_loss": 0.34767305850982666, "val_acc": 0.9}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.37277424335479736, "train_acc": 0.91, "val_loss": 0.33394676446914673, "val_acc": 0.9}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.3952494114637375, "train_acc": 0.865, "val_loss": 0.33965805172920227, "val_acc": 0.86}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.3850867599248886, "train_acc": 0.86, "val_loss": 0.31576672196388245, "val_acc": 0.88}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.3358479291200638, "train_acc": 0.875, "val_loss": 0.2632705271244049, "val_acc": 0.92}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.3130151778459549, "train_acc": 0.895, "val_loss": 0.2775978147983551, "val_acc": 0.9}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.2944730818271637, "train_acc": 0.91, "val_loss": 0.26737987995147705, "val_acc": 0.9}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.2920827865600586, "train_acc": 0.895, "val_loss": 0.22762741148471832, "val_acc": 0.92}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["sorted_descending"], "degraded_stage": {"initial_val_loss": 0.6631535291671753, "final_val_loss": 0.5396676063537598, "initial_val_acc": 0.56, "final_val_acc": 0.56, "best_val_acc": 0.56}, "improved_stage": {"initial_val_loss": 0.5814832448959351, "final_val_loss": 0.22762741148471832, "initial_val_acc": 0.56, "final_val_acc": 0.92, "best_val_acc": 0.94, "best_epoch": 3}, "improvement": 0.3799999999999999, "first_improvement_epoch": 1}}
|
7
|
{"target_pattern": "no_repeats", "degraded_accuracy": 0.48, "improved_accuracy": 0.9, "improvement": 0.42000000000000004, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 4918, "learning_rate": 0.07685357941383293, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "no_repeats", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["no_repeats"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.280282,
-0.546707,
-0.090641,
0.200795,
0.204919
],
[
0.414929,
0.425169,
0.179772,
0.33345,
-1.031172
],
[
-0.741868,
0.579431,
0.430802,
0.342102,
0.758507
],
[
-0.016363,
-0.372153,
-0.517864,
-0.428543,
-0.402103
],
[
-0.029082,
-0.049072,
0.01015,
-0.322467,
0.481829
]
],
"network.0.bias": [
0.51697,
0.779954,
-0.034681,
-0.750028,
0.138827
],
"network.2.weight": [
[
-0.274115,
-0.595407,
0.631171,
0.553815,
0.330489
],
[
0.635917,
-0.40423,
-0.090532,
0.011681,
0.357098
],
[
-0.465471,
-0.186783,
0.359861,
0.234443,
0.617421
],
[
-0.004752,
-0.258538,
-0.829439,
0.571109,
-0.2918
],
[
-0.760598,
-0.691925,
-0.500722,
-0.145339,
-0.664362
]
],
"network.2.bias": [
0.181642,
0.171181,
0.08764,
-0.51042,
-0.014004
],
"network.4.weight": [
[
0.651838,
-0.217951,
-0.500028,
-0.165998,
0.08659
],
[
0.131572,
-0.539649,
0.254981,
-0.481653,
-0.013994
],
[
-0.11894,
-0.43051,
-0.557587,
-0.173106,
0.082671
],
[
0.252629,
-0.748711,
0.29307,
0.600274,
0.534507
],
[
-0.380121,
-0.295036,
0.101763,
-0.159167,
-0.089632
]
],
"network.4.bias": [
0.30089,
0.573818,
-0.053132,
0.62165,
-0.362218
],
"network.6.weight": [
[
-0.357537,
-0.386843,
0.041689,
-0.28321,
0.409973
],
[
0.38086,
0.800331,
-0.106276,
0.781161,
0.267648
],
[
-0.109572,
-0.00462,
0.252487,
-0.08148,
-0.402373
],
[
-0.377082,
0.348793,
0.08921,
-0.400903,
-0.058117
],
[
-0.080408,
0.028481,
-0.112312,
-0.091599,
0.37581
]
],
"network.6.bias": [
-0.228627,
0.500763,
-0.422341,
0.366046,
0.667475
],
"network.8.weight": [
[
0.044873,
0.427686,
0.176117,
-0.628943,
-0.596999
],
[
-0.034813,
-0.53651,
-0.081414,
0.216472,
0.138469
],
[
0.400483,
-0.092049,
-0.000749,
-0.221601,
-0.169918
],
[
0.436753,
-0.065742,
0.302457,
-0.292837,
-0.718373
],
[
-0.080654,
-0.085727,
0.3844,
0.235612,
-0.063538
]
],
"network.8.bias": [
-0.332951,
-0.271137,
-0.172195,
-0.181609,
-0.263372
],
"network.10.weight": [
[
-0.362886,
0.363645,
0.044677,
-0.430668,
0.359263
],
[
0.564536,
0.107113,
-0.30528,
-0.028163,
-0.412839
],
[
0.52269,
-0.041337,
0.25282,
-0.227046,
0.442893
],
[
-0.243356,
0.477727,
-0.103348,
0.078505,
-0.403755
],
[
-0.03197,
0.55801,
-0.069202,
-0.528897,
-0.02414
]
],
"network.10.bias": [
-0.146404,
0.030698,
-0.275502,
0.466216,
0.356168
],
"network.12.weight": [
[
-0.093181,
0.285877,
0.508113,
-0.353716,
-0.329174
]
],
"network.12.bias": [
-0.24881
]
}
## Activation Signature
### 0
mean: [0.363533, 1.889870, 2.666966, -3.945210, -0.062839]
std: [1.059232, 2.312363, 2.401642, 2.031549, 1.050392]
### 2
mean: [0.575613, -0.439794, 0.616135, -3.490277, -3.644363]
std: [1.924045, 1.150260, 1.111686, 1.927182, 1.588881]
### 4
mean: [0.511391, 0.763702, -0.729793, 0.908568, -0.748245]
std: [0.476986, 0.419267, 0.793062, 0.610768, 0.549699]
### 6
mean: [-0.971128, 2.033638, -0.557214, 0.071536, 0.563675]
std: [0.465314, 0.925936, 0.096133, 0.264772, 0.076257]
### 8
mean: [0.103092, -1.250697, -0.489415, -0.765492, -0.437111]
std: [0.498800, 0.526695, 0.054904, 0.032338, 0.096871]
### 10
mean: [-0.232369, 0.164432, -0.151680, 0.408566, 0.348595]
std: [0.143498, 0.223238, 0.206690, 0.096232, 0.012642]
### 12
mean: [-0.439153]
std: [0.157319]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
no_repeats
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.280282,
-0.546707,
-0.090641,
0.200795,
0.204919
],
[
0.414929,
0.425169,
0.179772,
0.33345,
-1.031172
],
[
-0.741868,
0.579431,
0.430802,
0.342102,
0.758507
],
[
-0.016363,
-0.372153,
-0.517864,
-0.428543,
-0.402103
],
[
-0.029082,
-0.049072,
0.01015,
-0.322467,
0.481829
]
],
"network.0.bias": [
0.51697,
0.779954,
-0.034681,
-0.750028,
0.138827
],
"network.2.weight": [
[
-0.274115,
-0.595407,
0.631171,
0.553815,
0.330489
],
[
0.635917,
-0.40423,
-0.090532,
0.011681,
0.357098
],
[
-0.465471,
-0.186783,
0.359861,
0.234443,
0.617421
],
[
-0.004752,
-0.258538,
-0.829439,
0.571109,
-0.2918
],
[
-0.760598,
-0.691925,
-0.500722,
-0.145339,
-0.664362
]
],
"network.2.bias": [
0.181642,
0.171181,
0.08764,
-0.51042,
-0.014004
],
"network.4.weight": [
[
0.651838,
-0.217951,
-0.500028,
-0.165998,
0.08659
],
[
0.131572,
-0.539649,
0.254981,
-0.481653,
-0.013994
],
[
-0.11894,
-0.43051,
-0.557587,
-0.173106,
0.082671
],
[
0.252629,
-0.748711,
0.29307,
0.600274,
0.534507
],
[
-0.380121,
-0.295036,
0.101763,
-0.159167,
-0.089632
]
],
"network.4.bias": [
0.30089,
0.573818,
-0.053132,
0.62165,
-0.362218
],
"network.6.weight": [
[
-0.357537,
-0.386843,
0.041689,
-0.28321,
0.409973
],
[
0.38086,
0.800331,
-0.106276,
0.781161,
0.267648
],
[
-0.109572,
-0.00462,
0.252487,
-0.08148,
-0.402373
],
[
-0.377082,
0.348793,
0.08921,
-0.400903,
-0.058117
],
[
-0.080408,
0.028481,
-0.112312,
-0.091599,
0.37581
]
],
"network.6.bias": [
-0.228627,
0.500763,
-0.422341,
0.366046,
0.667475
],
"network.8.weight": [
[
0.044873,
0.427686,
0.176117,
-0.628943,
-0.596999
],
[
-0.034813,
-0.53651,
-0.081414,
0.216472,
0.138469
],
[
0.400483,
-0.092049,
-0.000749,
-0.221601,
-0.169918
],
[
0.436753,
-0.065742,
0.302457,
-0.292837,
-0.718373
],
[
-0.080654,
-0.085727,
0.3844,
0.235612,
-0.063538
]
],
"network.8.bias": [
-0.332951,
-0.271137,
-0.172195,
-0.181609,
-0.263372
],
"network.10.weight": [
[
-0.362886,
0.363645,
0.044677,
-0.430668,
0.359263
],
[
0.564536,
0.107113,
-0.30528,
-0.028163,
-0.412839
],
[
0.52269,
-0.041337,
0.25282,
-0.227046,
0.442893
],
[
-0.243356,
0.477727,
-0.103348,
0.078505,
-0.403755
],
[
-0.03197,
0.55801,
-0.069202,
-0.528897,
-0.02414
]
],
"network.10.bias": [
-0.146404,
0.030698,
-0.275502,
0.466216,
0.356168
],
"network.12.weight": [
[
-0.093181,
0.285877,
0.508113,
-0.353716,
-0.329174
]
],
"network.12.bias": [
-0.24881
]
}
## Activation Signature
### 0
mean: [0.363533, 1.889870, 2.666966, -3.945210, -0.062839]
std: [1.059232, 2.312363, 2.401642, 2.031549, 1.050392]
### 2
mean: [0.575613, -0.439794, 0.616135, -3.490277, -3.644363]
std: [1.924045, 1.150260, 1.111686, 1.927182, 1.588881]
### 4
mean: [0.511391, 0.763702, -0.729793, 0.908568, -0.748245]
std: [0.476986, 0.419267, 0.793062, 0.610768, 0.549699]
### 6
mean: [-0.971128, 2.033638, -0.557214, 0.071536, 0.563675]
std: [0.465314, 0.925936, 0.096133, 0.264772, 0.076257]
### 8
mean: [0.103092, -1.250697, -0.489415, -0.765492, -0.437111]
std: [0.498800, 0.526695, 0.054904, 0.032338, 0.096871]
### 10
mean: [-0.232369, 0.164432, -0.151680, 0.408566, 0.348595]
std: [0.143498, 0.223238, 0.206690, 0.096232, 0.012642]
### 12
mean: [-0.439153]
std: [0.157319]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
no_repeats
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.36353254318237305, "std": 1.0592315196990967}, "1": {"mean": 1.8898704051971436, "std": 2.3123626708984375}, "2": {"mean": 2.666966438293457, "std": 2.401641607284546}, "3": {"mean": -3.9452099800109863, "std": 2.0315489768981934}, "4": {"mean": -0.06283887475728989, "std": 1.0503923892974854}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.5756133198738098, "std": 1.924045443534851}, "1": {"mean": -0.43979427218437195, "std": 1.1502598524093628}, "2": {"mean": 0.6161348819732666, "std": 1.1116855144500732}, "3": {"mean": -3.490277051925659, "std": 1.9271823167800903}, "4": {"mean": -3.6443634033203125, "std": 1.5888808965682983}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.5113914012908936, "std": 0.4769863486289978}, "1": {"mean": 0.763702392578125, "std": 0.41926679015159607}, "2": {"mean": -0.729792594909668, "std": 0.7930617332458496}, "3": {"mean": 0.908568263053894, "std": 0.6107679009437561}, "4": {"mean": -0.7482450008392334, "std": 0.5496988892555237}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.9711277484893799, "std": 0.46531441807746887}, "1": {"mean": 2.033637762069702, "std": 0.9259357452392578}, "2": {"mean": -0.557214081287384, "std": 0.09613306075334549}, "3": {"mean": 0.07153620570898056, "std": 0.2647719979286194}, "4": {"mean": 0.5636752843856812, "std": 0.07625661045312881}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.10309229046106339, "std": 0.49880048632621765}, "1": {"mean": -1.2506970167160034, "std": 0.5266948342323303}, "2": {"mean": -0.48941540718078613, "std": 0.05490434914827347}, "3": {"mean": -0.7654919028282166, "std": 0.03233825042843819}, "4": {"mean": -0.43711110949516296, "std": 0.09687056392431259}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -0.23236873745918274, "std": 0.14349795877933502}, "1": {"mean": 0.16443228721618652, "std": 0.2232377827167511}, "2": {"mean": -0.15168048441410065, "std": 0.20669041574001312}, "3": {"mean": 0.40856635570526123, "std": 0.09623177349567413}, "4": {"mean": 0.34859463572502136, "std": 0.01264194492250681}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -0.43915319442749023, "std": 0.15731923282146454}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.280282, -0.546707, -0.090641, 0.200795, 0.204919], [0.414929, 0.425169, 0.179772, 0.33345, -1.031172], [-0.741868, 0.579431, 0.430802, 0.342102, 0.758507], [-0.016363, -0.372153, -0.517864, -0.428543, -0.402103], [-0.029082, -0.049072, 0.01015, -0.322467, 0.481829]], "network.0.bias": [0.51697, 0.779954, -0.034681, -0.750028, 0.138827], "network.2.weight": [[-0.274115, -0.595407, 0.631171, 0.553815, 0.330489], [0.635917, -0.40423, -0.090532, 0.011681, 0.357098], [-0.465471, -0.186783, 0.359861, 0.234443, 0.617421], [-0.004752, -0.258538, -0.829439, 0.571109, -0.2918], [-0.760598, -0.691925, -0.500722, -0.145339, -0.664362]], "network.2.bias": [0.181642, 0.171181, 0.08764, -0.51042, -0.014004], "network.4.weight": [[0.651838, -0.217951, -0.500028, -0.165998, 0.08659], [0.131572, -0.539649, 0.254981, -0.481653, -0.013994], [-0.11894, -0.43051, -0.557587, -0.173106, 0.082671], [0.252629, -0.748711, 0.29307, 0.600274, 0.534507], [-0.380121, -0.295036, 0.101763, -0.159167, -0.089632]], "network.4.bias": [0.30089, 0.573818, -0.053132, 0.62165, -0.362218], "network.6.weight": [[-0.357537, -0.386843, 0.041689, -0.28321, 0.409973], [0.38086, 0.800331, -0.106276, 0.781161, 0.267648], [-0.109572, -0.00462, 0.252487, -0.08148, -0.402373], [-0.377082, 0.348793, 0.08921, -0.400903, -0.058117], [-0.080408, 0.028481, -0.112312, -0.091599, 0.37581]], "network.6.bias": [-0.228627, 0.500763, -0.422341, 0.366046, 0.667475], "network.8.weight": [[0.044873, 0.427686, 0.176117, -0.628943, -0.596999], [-0.034813, -0.53651, -0.081414, 0.216472, 0.138469], [0.400483, -0.092049, -0.000749, -0.221601, -0.169918], [0.436753, -0.065742, 0.302457, -0.292837, -0.718373], [-0.080654, -0.085727, 0.3844, 0.235612, -0.063538]], "network.8.bias": [-0.332951, -0.271137, -0.172195, -0.181609, -0.263372], "network.10.weight": [[-0.362886, 0.363645, 0.044677, -0.430668, 0.359263], [0.564536, 0.107113, -0.30528, -0.028163, -0.412839], [0.52269, -0.041337, 0.25282, -0.227046, 0.442893], [-0.243356, 0.477727, -0.103348, 0.078505, -0.403755], [-0.03197, 0.55801, -0.069202, -0.528897, -0.02414]], "network.10.bias": [-0.146404, 0.030698, -0.275502, 0.466216, 0.356168], "network.12.weight": [[-0.093181, 0.285877, 0.508113, -0.353716, -0.329174]], "network.12.bias": [-0.24881]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6884620189666748, "train_acc": 0.57, "val_loss": 0.7074939012527466, "val_acc": 0.5}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6810376048088074, "train_acc": 0.57, "val_loss": 0.7025664448738098, "val_acc": 0.5}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6843936741352081, "train_acc": 0.57, "val_loss": 0.7050349712371826, "val_acc": 0.5}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6800090074539185, "train_acc": 0.57, "val_loss": 0.7014064192771912, "val_acc": 0.5}, {"stage": "degraded", "epoch": 4, "global_epoch": 4, "train_loss": 0.6748373806476593, "train_acc": 0.57, "val_loss": 0.671745777130127, "val_acc": 0.48}, {"stage": "improved", "epoch": 0, "global_epoch": 5, "train_loss": 0.6726998388767242, "train_acc": 0.525, "val_loss": 0.6128959059715271, "val_acc": 0.76}, {"stage": "improved", "epoch": 1, "global_epoch": 6, "train_loss": 0.6101576089859009, "train_acc": 0.795, "val_loss": 0.5654578804969788, "val_acc": 0.9}, {"stage": "improved", "epoch": 2, "global_epoch": 7, "train_loss": 0.5623864829540253, "train_acc": 0.785, "val_loss": 0.511611819267273, "val_acc": 0.78}, {"stage": "improved", "epoch": 3, "global_epoch": 8, "train_loss": 0.47313591837882996, "train_acc": 0.835, "val_loss": 0.46406760811805725, "val_acc": 0.88}, {"stage": "improved", "epoch": 4, "global_epoch": 9, "train_loss": 0.4665597677230835, "train_acc": 0.86, "val_loss": 0.4496612846851349, "val_acc": 0.86}, {"stage": "improved", "epoch": 5, "global_epoch": 10, "train_loss": 0.40005819499492645, "train_acc": 0.87, "val_loss": 0.41119635105133057, "val_acc": 0.88}, {"stage": "improved", "epoch": 6, "global_epoch": 11, "train_loss": 0.307979941368103, "train_acc": 0.905, "val_loss": 0.5346136689186096, "val_acc": 0.78}, {"stage": "improved", "epoch": 7, "global_epoch": 12, "train_loss": 0.3348309248685837, "train_acc": 0.865, "val_loss": 0.32273316383361816, "val_acc": 0.9}, {"stage": "improved", "epoch": 8, "global_epoch": 13, "train_loss": 0.31257490813732147, "train_acc": 0.85, "val_loss": 0.32616516947746277, "val_acc": 0.9}, {"stage": "improved", "epoch": 9, "global_epoch": 14, "train_loss": 0.26000000536441803, "train_acc": 0.9, "val_loss": 0.45417335629463196, "val_acc": 0.82}], "summary": {"total_epochs": 15, "degraded_epochs": 5, "improved_epochs": 10, "patterns": ["no_repeats"], "degraded_stage": {"initial_val_loss": 0.7074939012527466, "final_val_loss": 0.671745777130127, "initial_val_acc": 0.5, "final_val_acc": 0.48, "best_val_acc": 0.48}, "improved_stage": {"initial_val_loss": 0.6128959059715271, "final_val_loss": 0.45417335629463196, "initial_val_acc": 0.76, "final_val_acc": 0.82, "best_val_acc": 0.9, "best_epoch": 6}, "improvement": 0.42000000000000004, "first_improvement_epoch": 4}}
|
8
|
{"target_pattern": "has_majority", "degraded_accuracy": 0.38, "improved_accuracy": 0.72, "improvement": 0.33999999999999997, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 8556, "learning_rate": 0.09363094593146719, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "has_majority", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["has_majority"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 5
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-1.118466,
0.047511,
0.178778,
0.273657,
0.073636
],
[
-0.26888,
0.236607,
0.550824,
0.316424,
0.02675
],
[
0.300774,
-0.526303,
-0.721316,
0.272428,
0.092806
],
[
0.134338,
-0.668651,
0.352632,
0.04919,
-0.695795
],
[
0.120357,
0.723274,
0.204002,
0.19134,
-0.213389
]
],
"network.0.bias": [
-0.159815,
-0.377014,
0.027805,
0.312783,
-0.112303
],
"network.2.weight": [
[
-0.009897,
-0.546056,
0.250979,
0.302171,
-0.01212
],
[
-0.10592,
-0.564674,
-0.569456,
-0.398504,
0.545905
],
[
-0.532866,
-0.506333,
0.450231,
0.345383,
0.305847
],
[
0.335662,
0.210326,
-0.482148,
-0.849033,
0.452865
],
[
-0.218526,
-0.128953,
-0.090746,
0.166495,
-0.349012
]
],
"network.2.bias": [
-0.210994,
-0.091906,
0.541776,
0.189654,
-0.742942
],
"network.4.weight": [
[
0.102686,
-0.536538,
0.727974,
-0.575238,
-0.553967
],
[
0.727741,
-0.600845,
-0.248549,
-0.391689,
0.862721
],
[
1.017703,
-0.511729,
-0.485366,
-0.754247,
0.269128
],
[
0.779892,
-0.224907,
-0.786899,
0.490977,
0.21467
],
[
-0.258869,
-0.238531,
-0.374814,
-0.39361,
0.028509
]
],
"network.4.bias": [
0.144276,
-0.764274,
-0.580435,
-0.102749,
-0.158591
],
"network.6.weight": [
[
-0.241823,
0.695914,
0.129703,
0.678442,
-0.185734
],
[
-0.484245,
0.532345,
-0.163821,
-0.520848,
-0.341223
],
[
0.36082,
-0.008599,
-0.709849,
-0.647868,
0.093061
],
[
0.514355,
-0.893215,
-0.517511,
-0.894563,
0.208439
],
[
-0.495518,
-0.272973,
-0.25017,
0.347754,
0.167693
]
],
"network.6.bias": [
-0.007157,
-0.345935,
0.511267,
0.116689,
-0.335366
],
"network.8.weight": [
[
-0.466574,
-0.430239,
0.545188,
0.493281,
-0.226757
]
],
"network.8.bias": [
0.0617
]
}
## Activation Signature
### 0
mean: [-0.414001, 1.581684, -1.365021, -0.754526, 1.928714]
std: [2.258681, 1.419268, 1.688719, 1.617420, 1.739599]
### 2
mean: [-1.034715, -0.118489, 0.126776, 1.367715, -1.692902]
std: [0.754433, 0.935841, 0.784049, 1.233502, 0.800442]
### 4
mean: [-0.490883, -1.612099, -1.923687, 0.200484, -0.795545]
std: [0.928288, 0.667655, 1.085042, 0.742672, 0.585397]
### 6
mean: [0.114087, -0.515002, 0.396935, -0.001928, -0.232261]
std: [0.436857, 0.261078, 0.441700, 0.621178, 0.251373]
### 8
mean: [0.319516]
std: [0.429031]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
has_majority
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 5
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-1.118466,
0.047511,
0.178778,
0.273657,
0.073636
],
[
-0.26888,
0.236607,
0.550824,
0.316424,
0.02675
],
[
0.300774,
-0.526303,
-0.721316,
0.272428,
0.092806
],
[
0.134338,
-0.668651,
0.352632,
0.04919,
-0.695795
],
[
0.120357,
0.723274,
0.204002,
0.19134,
-0.213389
]
],
"network.0.bias": [
-0.159815,
-0.377014,
0.027805,
0.312783,
-0.112303
],
"network.2.weight": [
[
-0.009897,
-0.546056,
0.250979,
0.302171,
-0.01212
],
[
-0.10592,
-0.564674,
-0.569456,
-0.398504,
0.545905
],
[
-0.532866,
-0.506333,
0.450231,
0.345383,
0.305847
],
[
0.335662,
0.210326,
-0.482148,
-0.849033,
0.452865
],
[
-0.218526,
-0.128953,
-0.090746,
0.166495,
-0.349012
]
],
"network.2.bias": [
-0.210994,
-0.091906,
0.541776,
0.189654,
-0.742942
],
"network.4.weight": [
[
0.102686,
-0.536538,
0.727974,
-0.575238,
-0.553967
],
[
0.727741,
-0.600845,
-0.248549,
-0.391689,
0.862721
],
[
1.017703,
-0.511729,
-0.485366,
-0.754247,
0.269128
],
[
0.779892,
-0.224907,
-0.786899,
0.490977,
0.21467
],
[
-0.258869,
-0.238531,
-0.374814,
-0.39361,
0.028509
]
],
"network.4.bias": [
0.144276,
-0.764274,
-0.580435,
-0.102749,
-0.158591
],
"network.6.weight": [
[
-0.241823,
0.695914,
0.129703,
0.678442,
-0.185734
],
[
-0.484245,
0.532345,
-0.163821,
-0.520848,
-0.341223
],
[
0.36082,
-0.008599,
-0.709849,
-0.647868,
0.093061
],
[
0.514355,
-0.893215,
-0.517511,
-0.894563,
0.208439
],
[
-0.495518,
-0.272973,
-0.25017,
0.347754,
0.167693
]
],
"network.6.bias": [
-0.007157,
-0.345935,
0.511267,
0.116689,
-0.335366
],
"network.8.weight": [
[
-0.466574,
-0.430239,
0.545188,
0.493281,
-0.226757
]
],
"network.8.bias": [
0.0617
]
}
## Activation Signature
### 0
mean: [-0.414001, 1.581684, -1.365021, -0.754526, 1.928714]
std: [2.258681, 1.419268, 1.688719, 1.617420, 1.739599]
### 2
mean: [-1.034715, -0.118489, 0.126776, 1.367715, -1.692902]
std: [0.754433, 0.935841, 0.784049, 1.233502, 0.800442]
### 4
mean: [-0.490883, -1.612099, -1.923687, 0.200484, -0.795545]
std: [0.928288, 0.667655, 1.085042, 0.742672, 0.585397]
### 6
mean: [0.114087, -0.515002, 0.396935, -0.001928, -0.232261]
std: [0.436857, 0.261078, 0.441700, 0.621178, 0.251373]
### 8
mean: [0.319516]
std: [0.429031]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
has_majority
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -0.4140007495880127, "std": 2.258681297302246}, "1": {"mean": 1.58168363571167, "std": 1.419268012046814}, "2": {"mean": -1.365020751953125, "std": 1.6887187957763672}, "3": {"mean": -0.7545260190963745, "std": 1.6174200773239136}, "4": {"mean": 1.9287142753601074, "std": 1.7395989894866943}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -1.0347154140472412, "std": 0.7544334530830383}, "1": {"mean": -0.11848895996809006, "std": 0.9358409643173218}, "2": {"mean": 0.12677572667598724, "std": 0.7840493321418762}, "3": {"mean": 1.3677150011062622, "std": 1.2335023880004883}, "4": {"mean": -1.6929020881652832, "std": 0.8004419207572937}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.4908827841281891, "std": 0.928288459777832}, "1": {"mean": -1.612099051475525, "std": 0.6676552891731262}, "2": {"mean": -1.9236871004104614, "std": 1.085041880607605}, "3": {"mean": 0.20048418641090393, "std": 0.742672324180603}, "4": {"mean": -0.7955449223518372, "std": 0.5853970646858215}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.11408700793981552, "std": 0.43685653805732727}, "1": {"mean": -0.5150021910667419, "std": 0.2610776722431183}, "2": {"mean": 0.3969353437423706, "std": 0.4417001008987427}, "3": {"mean": -0.0019284659065306187, "std": 0.6211776733398438}, "4": {"mean": -0.23226149380207062, "std": 0.2513732314109802}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.31951576471328735, "std": 0.4290308654308319}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-1.118466, 0.047511, 0.178778, 0.273657, 0.073636], [-0.26888, 0.236607, 0.550824, 0.316424, 0.02675], [0.300774, -0.526303, -0.721316, 0.272428, 0.092806], [0.134338, -0.668651, 0.352632, 0.04919, -0.695795], [0.120357, 0.723274, 0.204002, 0.19134, -0.213389]], "network.0.bias": [-0.159815, -0.377014, 0.027805, 0.312783, -0.112303], "network.2.weight": [[-0.009897, -0.546056, 0.250979, 0.302171, -0.01212], [-0.10592, -0.564674, -0.569456, -0.398504, 0.545905], [-0.532866, -0.506333, 0.450231, 0.345383, 0.305847], [0.335662, 0.210326, -0.482148, -0.849033, 0.452865], [-0.218526, -0.128953, -0.090746, 0.166495, -0.349012]], "network.2.bias": [-0.210994, -0.091906, 0.541776, 0.189654, -0.742942], "network.4.weight": [[0.102686, -0.536538, 0.727974, -0.575238, -0.553967], [0.727741, -0.600845, -0.248549, -0.391689, 0.862721], [1.017703, -0.511729, -0.485366, -0.754247, 0.269128], [0.779892, -0.224907, -0.786899, 0.490977, 0.21467], [-0.258869, -0.238531, -0.374814, -0.39361, 0.028509]], "network.4.bias": [0.144276, -0.764274, -0.580435, -0.102749, -0.158591], "network.6.weight": [[-0.241823, 0.695914, 0.129703, 0.678442, -0.185734], [-0.484245, 0.532345, -0.163821, -0.520848, -0.341223], [0.36082, -0.008599, -0.709849, -0.647868, 0.093061], [0.514355, -0.893215, -0.517511, -0.894563, 0.208439], [-0.495518, -0.272973, -0.25017, 0.347754, 0.167693]], "network.6.bias": [-0.007157, -0.345935, 0.511267, 0.116689, -0.335366], "network.8.weight": [[-0.466574, -0.430239, 0.545188, 0.493281, -0.226757]], "network.8.bias": [0.0617]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6722663342952728, "train_acc": 0.6, "val_loss": 0.7984893918037415, "val_acc": 0.38}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6570966839790344, "train_acc": 0.6, "val_loss": 0.8252170085906982, "val_acc": 0.38}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6632483601570129, "train_acc": 0.6, "val_loss": 0.7811472415924072, "val_acc": 0.38}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6376964449882507, "train_acc": 0.6, "val_loss": 0.6819770336151123, "val_acc": 0.38}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.7213469445705414, "train_acc": 0.535, "val_loss": 0.6547825336456299, "val_acc": 0.6}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.601013571023941, "train_acc": 0.69, "val_loss": 0.5960362553596497, "val_acc": 0.7}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.6852565705776215, "train_acc": 0.55, "val_loss": 0.561551034450531, "val_acc": 0.72}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.5768475234508514, "train_acc": 0.7, "val_loss": 0.6489304900169373, "val_acc": 0.54}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.5613828897476196, "train_acc": 0.685, "val_loss": 0.7167420983314514, "val_acc": 0.5}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.5396362841129303, "train_acc": 0.68, "val_loss": 0.6449691653251648, "val_acc": 0.5}], "summary": {"total_epochs": 10, "degraded_epochs": 4, "improved_epochs": 6, "patterns": ["has_majority"], "degraded_stage": {"initial_val_loss": 0.7984893918037415, "final_val_loss": 0.6819770336151123, "initial_val_acc": 0.38, "final_val_acc": 0.38, "best_val_acc": 0.38}, "improved_stage": {"initial_val_loss": 0.6547825336456299, "final_val_loss": 0.6449691653251648, "initial_val_acc": 0.6, "final_val_acc": 0.5, "best_val_acc": 0.72, "best_epoch": 6}, "improvement": 0.33999999999999997, "first_improvement_epoch": 3}}
|
9
|
{"target_pattern": "decreasing_pairs", "degraded_accuracy": 0.5, "improved_accuracy": 0.96, "improvement": 0.45999999999999996, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 9319, "learning_rate": 0.04720846293947128, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "decreasing_pairs", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["decreasing_pairs"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.518219,
0.453979,
0.207907,
-0.135294,
0.066146
],
[
0.792622,
0.555805,
-0.047865,
0.204579,
-0.372802
],
[
0.120406,
-0.115068,
-0.498734,
-0.164382,
-0.479681
],
[
0.207842,
0.463524,
-0.395443,
0.262224,
-0.435124
],
[
0.767828,
0.501543,
-0.012203,
0.232021,
-0.00553
],
[
-0.47199,
0.086175,
0.305462,
0.438044,
-0.233787
],
[
0.39524,
0.345578,
0.198513,
-0.387603,
-0.097907
],
[
0.210843,
-0.452041,
0.288004,
-0.011879,
-0.031162
]
],
"network.0.bias": [
0.085717,
-0.113422,
0.028428,
-0.125706,
0.427135,
0.567853,
0.080095,
-0.149096
],
"network.2.weight": [
[
0.465472,
0.229947,
0.008392,
0.211449,
-0.196991,
0.282935,
-0.241359,
-0.271923
],
[
0.58716,
0.041907,
-0.05679,
0.296702,
-0.081369,
0.323723,
-0.413152,
0.191019
],
[
0.19057,
-0.099749,
0.18259,
-0.205804,
-0.12121,
-0.02527,
0.542427,
0.368309
],
[
-0.617663,
0.297287,
0.393693,
0.4157,
-0.194772,
-0.243126,
0.540867,
-0.050168
],
[
0.028904,
-0.434685,
-0.08578,
-0.419543,
0.05896,
0.050079,
0.102109,
-0.272589
],
[
-0.246055,
0.657178,
0.064406,
0.112526,
0.584335,
-0.656931,
0.149501,
-0.097182
],
[
0.410764,
-0.215912,
-0.601742,
-0.104124,
0.395576,
0.236571,
-0.428516,
0.275303
],
[
0.401175,
0.232703,
0.167734,
-0.367191,
-0.296801,
0.195241,
0.084716,
0.158429
]
],
"network.2.bias": [
-0.289735,
0.264814,
0.126607,
0.217012,
-0.435508,
-0.090351,
0.452223,
-0.125361
],
"network.4.weight": [
[
0.003585,
-0.182196,
-0.270135,
0.192221,
-0.080631,
0.418237,
0.021674,
0.146549
],
[
0.295668,
0.242138,
0.638519,
-0.104105,
-0.149008,
-0.115518,
0.38833,
0.305269
],
[
0.078645,
-0.176802,
0.146812,
-0.065465,
0.055521,
-0.221629,
0.133533,
-0.149832
],
[
0.236228,
0.530951,
0.206938,
-0.409253,
0.177515,
-0.015611,
0.308298,
0.044374
],
[
-0.099025,
0.034489,
0.081553,
-0.209238,
0.014994,
-0.318753,
-0.183579,
-0.014764
],
[
-0.338259,
-0.310589,
-0.312632,
0.311686,
-0.071328,
0.578774,
-0.02784,
-0.259003
],
[
0.168886,
0.061155,
0.071284,
0.062636,
-0.345037,
-0.171951,
0.035143,
0.269647
],
[
-0.062648,
0.204794,
0.160413,
-0.247227,
0.22187,
-0.194608,
0.131205,
0.143734
]
],
"network.4.bias": [
0.085228,
0.589485,
-0.412257,
0.372403,
-0.102691,
-0.080698,
-0.261705,
-0.128854
],
"network.6.weight": [
[
0.08931,
0.048398,
-0.313207,
-0.011783,
-0.046864,
-0.27349,
0.163408,
-0.025788
],
[
0.187764,
0.305813,
-0.275114,
0.335068,
-0.251295,
-0.187907,
0.339988,
0.344552
],
[
0.11931,
0.580207,
-0.130608,
0.562846,
0.216293,
-0.048527,
0.280001,
0.24688
],
[
0.592559,
-0.411438,
-0.048387,
-0.102771,
0.362754,
0.701174,
0.044959,
-0.281713
],
[
0.098254,
0.638236,
0.390871,
0.319071,
-0.355208,
-0.350794,
0.255391,
0.077965
],
[
-0.168851,
0.204246,
-0.195054,
0.419457,
0.019185,
0.007788,
0.172705,
0.272122
],
[
-0.226787,
0.302116,
-0.003731,
0.250324,
0.031104,
-0.080707,
-0.273117,
0.235
],
[
-0.266822,
-0.220344,
0.076136,
0.233157,
-0.319437,
0.307989,
0.158961,
0.216961
]
],
"network.6.bias": [
-0.358766,
-0.125408,
0.383845,
0.380395,
0.392395,
-0.149135,
0.189128,
-0.217908
],
"network.8.weight": [
[
-0.186953,
-0.553102,
0.14111,
0.632781,
-0.538893,
-0.184339,
-0.182016,
-0.01433
],
[
0.133104,
0.234513,
0.498889,
-0.20856,
0.522816,
0.566769,
0.375109,
-0.411266
],
[
-0.176505,
0.050616,
0.40526,
-0.372145,
0.63993,
0.051067,
0.170187,
-0.082957
],
[
-0.348615,
0.216164,
0.099,
0.068025,
0.047123,
-0.205255,
-0.164486,
0.241534
],
[
-0.026473,
0.089358,
-0.370657,
0.061801,
0.166912,
0.108273,
-0.496139,
-0.418415
],
[
0.167938,
-0.028009,
0.368099,
-0.276226,
0.477505,
0.346323,
0.065184,
-0.192218
],
[
0.160199,
0.204058,
0.184012,
-0.343143,
0.704052,
0.421964,
0.432445,
-0.454575
],
[
-0.344907,
0.192591,
-0.265312,
-0.168458,
-0.206547,
0.175211,
-0.059456,
-0.200944
]
],
"network.8.bias": [
0.346025,
0.308888,
0.270673,
0.149646,
-0.26251,
0.204739,
0.214416,
-0.160259
],
"network.10.weight": [
[
0.465732,
-0.32838,
-0.710053,
0.11958,
-0.116276,
-0.331215,
-0.420113,
-0.194154
]
],
"network.10.bias": [
0.072265
]
}
## Activation Signature
### 0
mean: [0.493817, 1.764688, -2.017327, 0.177221, 2.761825, 1.426618, 0.664395, -0.172732]
std: [1.068193, 2.242293, 1.514430, 1.574270, 2.159417, 1.418597, 1.486291, 0.946906]
### 2
mean: [0.223949, 0.905384, 0.155643, 0.209918, -1.255745, 1.761818, 1.385823, -0.060018]
std: [0.858906, 1.029966, 0.651562, 1.285015, 1.040718, 3.089335, 0.674401, 0.564077]
### 4
mean: [0.796598, 1.481501, -0.824947, 1.264141, -1.080591, 0.568283, -0.303168, -0.203675]
std: [1.479205, 0.983802, 0.691687, 1.095031, 1.100499, 2.263322, 0.562491, 0.978780]
### 6
mean: [-0.502871, 0.867657, 2.167725, 0.786569, 1.560068, 0.672231, 0.751963, -0.088393]
std: [0.442394, 0.743815, 1.031980, 2.618426, 1.337844, 0.786483, 0.919110, 0.220600]
### 8
mean: [-0.331541, 3.023824, 2.037124, 0.427268, -1.021646, 1.784786, 2.302716, -1.046500]
std: [2.328525, 2.182372, 1.992805, 0.164410, 0.445848, 1.654882, 2.271739, 0.277215]
### 10
mean: [-3.974064]
std: [3.414995]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
decreasing_pairs
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.518219,
0.453979,
0.207907,
-0.135294,
0.066146
],
[
0.792622,
0.555805,
-0.047865,
0.204579,
-0.372802
],
[
0.120406,
-0.115068,
-0.498734,
-0.164382,
-0.479681
],
[
0.207842,
0.463524,
-0.395443,
0.262224,
-0.435124
],
[
0.767828,
0.501543,
-0.012203,
0.232021,
-0.00553
],
[
-0.47199,
0.086175,
0.305462,
0.438044,
-0.233787
],
[
0.39524,
0.345578,
0.198513,
-0.387603,
-0.097907
],
[
0.210843,
-0.452041,
0.288004,
-0.011879,
-0.031162
]
],
"network.0.bias": [
0.085717,
-0.113422,
0.028428,
-0.125706,
0.427135,
0.567853,
0.080095,
-0.149096
],
"network.2.weight": [
[
0.465472,
0.229947,
0.008392,
0.211449,
-0.196991,
0.282935,
-0.241359,
-0.271923
],
[
0.58716,
0.041907,
-0.05679,
0.296702,
-0.081369,
0.323723,
-0.413152,
0.191019
],
[
0.19057,
-0.099749,
0.18259,
-0.205804,
-0.12121,
-0.02527,
0.542427,
0.368309
],
[
-0.617663,
0.297287,
0.393693,
0.4157,
-0.194772,
-0.243126,
0.540867,
-0.050168
],
[
0.028904,
-0.434685,
-0.08578,
-0.419543,
0.05896,
0.050079,
0.102109,
-0.272589
],
[
-0.246055,
0.657178,
0.064406,
0.112526,
0.584335,
-0.656931,
0.149501,
-0.097182
],
[
0.410764,
-0.215912,
-0.601742,
-0.104124,
0.395576,
0.236571,
-0.428516,
0.275303
],
[
0.401175,
0.232703,
0.167734,
-0.367191,
-0.296801,
0.195241,
0.084716,
0.158429
]
],
"network.2.bias": [
-0.289735,
0.264814,
0.126607,
0.217012,
-0.435508,
-0.090351,
0.452223,
-0.125361
],
"network.4.weight": [
[
0.003585,
-0.182196,
-0.270135,
0.192221,
-0.080631,
0.418237,
0.021674,
0.146549
],
[
0.295668,
0.242138,
0.638519,
-0.104105,
-0.149008,
-0.115518,
0.38833,
0.305269
],
[
0.078645,
-0.176802,
0.146812,
-0.065465,
0.055521,
-0.221629,
0.133533,
-0.149832
],
[
0.236228,
0.530951,
0.206938,
-0.409253,
0.177515,
-0.015611,
0.308298,
0.044374
],
[
-0.099025,
0.034489,
0.081553,
-0.209238,
0.014994,
-0.318753,
-0.183579,
-0.014764
],
[
-0.338259,
-0.310589,
-0.312632,
0.311686,
-0.071328,
0.578774,
-0.02784,
-0.259003
],
[
0.168886,
0.061155,
0.071284,
0.062636,
-0.345037,
-0.171951,
0.035143,
0.269647
],
[
-0.062648,
0.204794,
0.160413,
-0.247227,
0.22187,
-0.194608,
0.131205,
0.143734
]
],
"network.4.bias": [
0.085228,
0.589485,
-0.412257,
0.372403,
-0.102691,
-0.080698,
-0.261705,
-0.128854
],
"network.6.weight": [
[
0.08931,
0.048398,
-0.313207,
-0.011783,
-0.046864,
-0.27349,
0.163408,
-0.025788
],
[
0.187764,
0.305813,
-0.275114,
0.335068,
-0.251295,
-0.187907,
0.339988,
0.344552
],
[
0.11931,
0.580207,
-0.130608,
0.562846,
0.216293,
-0.048527,
0.280001,
0.24688
],
[
0.592559,
-0.411438,
-0.048387,
-0.102771,
0.362754,
0.701174,
0.044959,
-0.281713
],
[
0.098254,
0.638236,
0.390871,
0.319071,
-0.355208,
-0.350794,
0.255391,
0.077965
],
[
-0.168851,
0.204246,
-0.195054,
0.419457,
0.019185,
0.007788,
0.172705,
0.272122
],
[
-0.226787,
0.302116,
-0.003731,
0.250324,
0.031104,
-0.080707,
-0.273117,
0.235
],
[
-0.266822,
-0.220344,
0.076136,
0.233157,
-0.319437,
0.307989,
0.158961,
0.216961
]
],
"network.6.bias": [
-0.358766,
-0.125408,
0.383845,
0.380395,
0.392395,
-0.149135,
0.189128,
-0.217908
],
"network.8.weight": [
[
-0.186953,
-0.553102,
0.14111,
0.632781,
-0.538893,
-0.184339,
-0.182016,
-0.01433
],
[
0.133104,
0.234513,
0.498889,
-0.20856,
0.522816,
0.566769,
0.375109,
-0.411266
],
[
-0.176505,
0.050616,
0.40526,
-0.372145,
0.63993,
0.051067,
0.170187,
-0.082957
],
[
-0.348615,
0.216164,
0.099,
0.068025,
0.047123,
-0.205255,
-0.164486,
0.241534
],
[
-0.026473,
0.089358,
-0.370657,
0.061801,
0.166912,
0.108273,
-0.496139,
-0.418415
],
[
0.167938,
-0.028009,
0.368099,
-0.276226,
0.477505,
0.346323,
0.065184,
-0.192218
],
[
0.160199,
0.204058,
0.184012,
-0.343143,
0.704052,
0.421964,
0.432445,
-0.454575
],
[
-0.344907,
0.192591,
-0.265312,
-0.168458,
-0.206547,
0.175211,
-0.059456,
-0.200944
]
],
"network.8.bias": [
0.346025,
0.308888,
0.270673,
0.149646,
-0.26251,
0.204739,
0.214416,
-0.160259
],
"network.10.weight": [
[
0.465732,
-0.32838,
-0.710053,
0.11958,
-0.116276,
-0.331215,
-0.420113,
-0.194154
]
],
"network.10.bias": [
0.072265
]
}
## Activation Signature
### 0
mean: [0.493817, 1.764688, -2.017327, 0.177221, 2.761825, 1.426618, 0.664395, -0.172732]
std: [1.068193, 2.242293, 1.514430, 1.574270, 2.159417, 1.418597, 1.486291, 0.946906]
### 2
mean: [0.223949, 0.905384, 0.155643, 0.209918, -1.255745, 1.761818, 1.385823, -0.060018]
std: [0.858906, 1.029966, 0.651562, 1.285015, 1.040718, 3.089335, 0.674401, 0.564077]
### 4
mean: [0.796598, 1.481501, -0.824947, 1.264141, -1.080591, 0.568283, -0.303168, -0.203675]
std: [1.479205, 0.983802, 0.691687, 1.095031, 1.100499, 2.263322, 0.562491, 0.978780]
### 6
mean: [-0.502871, 0.867657, 2.167725, 0.786569, 1.560068, 0.672231, 0.751963, -0.088393]
std: [0.442394, 0.743815, 1.031980, 2.618426, 1.337844, 0.786483, 0.919110, 0.220600]
### 8
mean: [-0.331541, 3.023824, 2.037124, 0.427268, -1.021646, 1.784786, 2.302716, -1.046500]
std: [2.328525, 2.182372, 1.992805, 0.164410, 0.445848, 1.654882, 2.271739, 0.277215]
### 10
mean: [-3.974064]
std: [3.414995]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
decreasing_pairs
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.49381667375564575, "std": 1.0681926012039185}, "1": {"mean": 1.76468825340271, "std": 2.242293357849121}, "2": {"mean": -2.017327308654785, "std": 1.514430284500122}, "3": {"mean": 0.17722110450267792, "std": 1.5742701292037964}, "4": {"mean": 2.761824607849121, "std": 2.159416675567627}, "5": {"mean": 1.4266177415847778, "std": 1.4185971021652222}, "6": {"mean": 0.6643949747085571, "std": 1.4862908124923706}, "7": {"mean": -0.17273229360580444, "std": 0.9469058513641357}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.22394898533821106, "std": 0.8589063286781311}, "1": {"mean": 0.9053837060928345, "std": 1.0299659967422485}, "2": {"mean": 0.15564343333244324, "std": 0.6515620946884155}, "3": {"mean": 0.2099183201789856, "std": 1.2850149869918823}, "4": {"mean": -1.2557448148727417, "std": 1.0407178401947021}, "5": {"mean": 1.7618178129196167, "std": 3.0893349647521973}, "6": {"mean": 1.385823369026184, "std": 0.6744005084037781}, "7": {"mean": -0.06001823768019676, "std": 0.5640771389007568}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.7965983152389526, "std": 1.4792051315307617}, "1": {"mean": 1.4815006256103516, "std": 0.9838020205497742}, "2": {"mean": -0.8249471783638, "std": 0.6916873455047607}, "3": {"mean": 1.2641406059265137, "std": 1.0950310230255127}, "4": {"mean": -1.0805906057357788, "std": 1.100499153137207}, "5": {"mean": 0.5682833194732666, "std": 2.2633216381073}, "6": {"mean": -0.3031679093837738, "std": 0.562490701675415}, "7": {"mean": -0.20367531478405, "std": 0.9787797331809998}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.5028712153434753, "std": 0.44239357113838196}, "1": {"mean": 0.8676571249961853, "std": 0.7438154816627502}, "2": {"mean": 2.167724847793579, "std": 1.0319796800613403}, "3": {"mean": 0.7865690588951111, "std": 2.6184258460998535}, "4": {"mean": 1.5600675344467163, "std": 1.3378442525863647}, "5": {"mean": 0.6722311973571777, "std": 0.7864825129508972}, "6": {"mean": 0.7519629001617432, "std": 0.9191099405288696}, "7": {"mean": -0.08839259296655655, "std": 0.22059951722621918}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.3315410912036896, "std": 2.3285250663757324}, "1": {"mean": 3.0238237380981445, "std": 2.1823716163635254}, "2": {"mean": 2.037123918533325, "std": 1.9928053617477417}, "3": {"mean": 0.42726778984069824, "std": 0.1644100397825241}, "4": {"mean": -1.0216460227966309, "std": 0.44584816694259644}, "5": {"mean": 1.784785509109497, "std": 1.6548815965652466}, "6": {"mean": 2.30271577835083, "std": 2.2717392444610596}, "7": {"mean": -1.046500325202942, "std": 0.2772153913974762}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -3.9740641117095947, "std": 3.414994955062866}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.518219, 0.453979, 0.207907, -0.135294, 0.066146], [0.792622, 0.555805, -0.047865, 0.204579, -0.372802], [0.120406, -0.115068, -0.498734, -0.164382, -0.479681], [0.207842, 0.463524, -0.395443, 0.262224, -0.435124], [0.767828, 0.501543, -0.012203, 0.232021, -0.00553], [-0.47199, 0.086175, 0.305462, 0.438044, -0.233787], [0.39524, 0.345578, 0.198513, -0.387603, -0.097907], [0.210843, -0.452041, 0.288004, -0.011879, -0.031162]], "network.0.bias": [0.085717, -0.113422, 0.028428, -0.125706, 0.427135, 0.567853, 0.080095, -0.149096], "network.2.weight": [[0.465472, 0.229947, 0.008392, 0.211449, -0.196991, 0.282935, -0.241359, -0.271923], [0.58716, 0.041907, -0.05679, 0.296702, -0.081369, 0.323723, -0.413152, 0.191019], [0.19057, -0.099749, 0.18259, -0.205804, -0.12121, -0.02527, 0.542427, 0.368309], [-0.617663, 0.297287, 0.393693, 0.4157, -0.194772, -0.243126, 0.540867, -0.050168], [0.028904, -0.434685, -0.08578, -0.419543, 0.05896, 0.050079, 0.102109, -0.272589], [-0.246055, 0.657178, 0.064406, 0.112526, 0.584335, -0.656931, 0.149501, -0.097182], [0.410764, -0.215912, -0.601742, -0.104124, 0.395576, 0.236571, -0.428516, 0.275303], [0.401175, 0.232703, 0.167734, -0.367191, -0.296801, 0.195241, 0.084716, 0.158429]], "network.2.bias": [-0.289735, 0.264814, 0.126607, 0.217012, -0.435508, -0.090351, 0.452223, -0.125361], "network.4.weight": [[0.003585, -0.182196, -0.270135, 0.192221, -0.080631, 0.418237, 0.021674, 0.146549], [0.295668, 0.242138, 0.638519, -0.104105, -0.149008, -0.115518, 0.38833, 0.305269], [0.078645, -0.176802, 0.146812, -0.065465, 0.055521, -0.221629, 0.133533, -0.149832], [0.236228, 0.530951, 0.206938, -0.409253, 0.177515, -0.015611, 0.308298, 0.044374], [-0.099025, 0.034489, 0.081553, -0.209238, 0.014994, -0.318753, -0.183579, -0.014764], [-0.338259, -0.310589, -0.312632, 0.311686, -0.071328, 0.578774, -0.02784, -0.259003], [0.168886, 0.061155, 0.071284, 0.062636, -0.345037, -0.171951, 0.035143, 0.269647], [-0.062648, 0.204794, 0.160413, -0.247227, 0.22187, -0.194608, 0.131205, 0.143734]], "network.4.bias": [0.085228, 0.589485, -0.412257, 0.372403, -0.102691, -0.080698, -0.261705, -0.128854], "network.6.weight": [[0.08931, 0.048398, -0.313207, -0.011783, -0.046864, -0.27349, 0.163408, -0.025788], [0.187764, 0.305813, -0.275114, 0.335068, -0.251295, -0.187907, 0.339988, 0.344552], [0.11931, 0.580207, -0.130608, 0.562846, 0.216293, -0.048527, 0.280001, 0.24688], [0.592559, -0.411438, -0.048387, -0.102771, 0.362754, 0.701174, 0.044959, -0.281713], [0.098254, 0.638236, 0.390871, 0.319071, -0.355208, -0.350794, 0.255391, 0.077965], [-0.168851, 0.204246, -0.195054, 0.419457, 0.019185, 0.007788, 0.172705, 0.272122], [-0.226787, 0.302116, -0.003731, 0.250324, 0.031104, -0.080707, -0.273117, 0.235], [-0.266822, -0.220344, 0.076136, 0.233157, -0.319437, 0.307989, 0.158961, 0.216961]], "network.6.bias": [-0.358766, -0.125408, 0.383845, 0.380395, 0.392395, -0.149135, 0.189128, -0.217908], "network.8.weight": [[-0.186953, -0.553102, 0.14111, 0.632781, -0.538893, -0.184339, -0.182016, -0.01433], [0.133104, 0.234513, 0.498889, -0.20856, 0.522816, 0.566769, 0.375109, -0.411266], [-0.176505, 0.050616, 0.40526, -0.372145, 0.63993, 0.051067, 0.170187, -0.082957], [-0.348615, 0.216164, 0.099, 0.068025, 0.047123, -0.205255, -0.164486, 0.241534], [-0.026473, 0.089358, -0.370657, 0.061801, 0.166912, 0.108273, -0.496139, -0.418415], [0.167938, -0.028009, 0.368099, -0.276226, 0.477505, 0.346323, 0.065184, -0.192218], [0.160199, 0.204058, 0.184012, -0.343143, 0.704052, 0.421964, 0.432445, -0.454575], [-0.344907, 0.192591, -0.265312, -0.168458, -0.206547, 0.175211, -0.059456, -0.200944]], "network.8.bias": [0.346025, 0.308888, 0.270673, 0.149646, -0.26251, 0.204739, 0.214416, -0.160259], "network.10.weight": [[0.465732, -0.32838, -0.710053, 0.11958, -0.116276, -0.331215, -0.420113, -0.194154]], "network.10.bias": [0.072265]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6809704303741455, "train_acc": 0.57, "val_loss": 0.7026543617248535, "val_acc": 0.5}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6687776148319244, "train_acc": 0.57, "val_loss": 0.6631693840026855, "val_acc": 0.5}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.636007159948349, "train_acc": 0.5, "val_loss": 0.5195937156677246, "val_acc": 0.94}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.4667539596557617, "train_acc": 0.925, "val_loss": 0.2829795777797699, "val_acc": 0.94}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.24544239044189453, "train_acc": 0.935, "val_loss": 0.12007593363523483, "val_acc": 0.96}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.20778611302375793, "train_acc": 0.945, "val_loss": 0.11817152053117752, "val_acc": 0.96}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.22254595905542374, "train_acc": 0.945, "val_loss": 0.1369670033454895, "val_acc": 0.92}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.17001918703317642, "train_acc": 0.945, "val_loss": 0.10708385705947876, "val_acc": 0.96}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.1698010414838791, "train_acc": 0.945, "val_loss": 0.1090748980641365, "val_acc": 0.96}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.1563536301255226, "train_acc": 0.945, "val_loss": 0.12032539397478104, "val_acc": 0.96}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.1522565893828869, "train_acc": 0.945, "val_loss": 0.1307438760995865, "val_acc": 0.96}], "summary": {"total_epochs": 11, "degraded_epochs": 2, "improved_epochs": 9, "patterns": ["decreasing_pairs"], "degraded_stage": {"initial_val_loss": 0.7026543617248535, "final_val_loss": 0.6631693840026855, "initial_val_acc": 0.5, "final_val_acc": 0.5, "best_val_acc": 0.5}, "improved_stage": {"initial_val_loss": 0.5195937156677246, "final_val_loss": 0.1307438760995865, "initial_val_acc": 0.94, "final_val_acc": 0.96, "best_val_acc": 0.96, "best_epoch": 4}, "improvement": 0.45999999999999996, "first_improvement_epoch": 1}}
|
10
|
{"target_pattern": "decreasing_pairs", "degraded_accuracy": 0.42, "improved_accuracy": 0.98, "improvement": 0.56, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 9749, "learning_rate": 0.04776131004515171, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "decreasing_pairs", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["decreasing_pairs"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.500303,
0.360095,
0.487294,
0.607866,
0.848291
],
[
0.550974,
0.636062,
-0.065267,
0.280651,
-0.434453
],
[
-0.947375,
0.142487,
0.339304,
0.430514,
-0.250845
],
[
0.127817,
-0.468976,
-0.172216,
0.398061,
-0.187961
],
[
-0.862973,
0.044614,
-0.135352,
0.514772,
0.601153
]
],
"network.0.bias": [
0.195653,
0.000992,
0.080358,
0.46163,
-0.12064
],
"network.2.weight": [
[
-0.010073,
0.059244,
0.037662,
0.342042,
-0.52094
],
[
0.358063,
0.64124,
-0.453315,
-0.633098,
-0.354225
],
[
-0.17374,
-0.600807,
-0.258416,
-0.380246,
-0.076232
],
[
-0.672244,
-0.166596,
0.347899,
0.221961,
-0.07373
],
[
0.25918,
-0.445785,
0.699829,
0.479309,
0.951773
]
],
"network.2.bias": [
0.588682,
0.214496,
-0.595319,
-0.109255,
0.78255
],
"network.4.weight": [
[
0.589969,
-0.336087,
0.108827,
-0.446314,
0.986849
],
[
-0.392093,
0.45332,
-0.044341,
0.300338,
-0.120207
],
[
0.063027,
-0.078277,
-0.36721,
0.056369,
-0.190713
],
[
-0.362951,
-0.089719,
-0.035969,
-0.22142,
-0.424396
],
[
-0.571098,
0.804071,
-0.083268,
-0.419644,
-0.546339
]
],
"network.4.bias": [
0.650863,
0.386239,
-0.635583,
-0.08702,
0.715579
],
"network.6.weight": [
[
1.188161,
-0.289077,
0.114721,
-0.42908,
-0.535296
],
[
1.129107,
-0.18894,
-0.270801,
-0.216499,
-0.667133
],
[
-0.434341,
-0.448443,
-0.314121,
-0.340095,
-0.765645
],
[
0.793312,
-0.188154,
-0.136274,
0.433892,
-0.663203
],
[
0.697722,
0.514323,
-0.08199,
0.281672,
0.447985
]
],
"network.6.bias": [
0.62492,
0.587377,
0.042101,
0.699975,
-0.021421
],
"network.8.weight": [
[
0.074177,
-0.107319,
-0.105117,
-0.029678,
-0.092964
],
[
0.996119,
0.672239,
0.149679,
0.90704,
0.153327
],
[
-0.496069,
-0.421864,
-0.163412,
-0.759388,
0.041515
],
[
-0.302005,
0.218972,
0.063319,
-0.398523,
0.273063
],
[
-0.494037,
-0.390588,
-0.423876,
-0.173145,
0.64188
]
],
"network.8.bias": [
-0.197562,
-0.228519,
1.034686,
-0.285029,
0.950433
],
"network.10.weight": [
[
-0.219898,
-0.190944,
0.786125,
0.176029,
0.905704
],
[
-0.059607,
-0.226763,
-0.41008,
0.016497,
-0.013595
],
[
-0.311035,
0.597817,
-0.432606,
-0.413965,
-0.035309
],
[
-0.391154,
-0.207717,
-0.554663,
0.19046,
-0.263792
],
[
-0.404038,
-0.055048,
0.153535,
-0.438568,
0.142796
]
],
"network.10.bias": [
0.707537,
-0.008378,
0.313014,
-0.100457,
-0.64538
],
"network.12.weight": [
[
0.733564,
-0.231835,
-0.736627,
-0.111804,
-0.171601
]
],
"network.12.bias": [
-0.167721
]
}
## Activation Signature
### 0
mean: [4.842091, 1.775374, 0.483650, 0.028173, 0.449324]
std: [3.166181, 2.078246, 2.136070, 1.046888, 2.340316]
### 2
mean: [0.257372, 1.949358, -3.106567, -3.266935, 3.285975]
std: [0.762667, 2.293204, 1.621871, 2.364141, 2.488432]
### 4
mean: [3.495814, 0.714993, -1.389547, -1.828467, 0.261229]
std: [2.664186, 1.101909, 0.457688, 0.906053, 2.507033]
### 6
mean: [4.079059, 3.771381, -2.657724, 2.690639, 3.337809]
std: [3.980827, 3.951144, 1.434915, 3.128899, 1.382859]
### 8
mean: [-0.724722, 10.466688, -5.240358, -1.057474, -1.325549]
std: [0.262236, 7.767156, 4.777078, 1.077974, 2.883852]
### 10
mean: [-0.427651, -2.489301, 6.381147, -2.571496, -1.143824]
std: [2.835357, 1.629254, 4.894650, 1.296927, 0.534474]
### 12
mean: [-4.198159]
std: [4.597022]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
decreasing_pairs
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.500303,
0.360095,
0.487294,
0.607866,
0.848291
],
[
0.550974,
0.636062,
-0.065267,
0.280651,
-0.434453
],
[
-0.947375,
0.142487,
0.339304,
0.430514,
-0.250845
],
[
0.127817,
-0.468976,
-0.172216,
0.398061,
-0.187961
],
[
-0.862973,
0.044614,
-0.135352,
0.514772,
0.601153
]
],
"network.0.bias": [
0.195653,
0.000992,
0.080358,
0.46163,
-0.12064
],
"network.2.weight": [
[
-0.010073,
0.059244,
0.037662,
0.342042,
-0.52094
],
[
0.358063,
0.64124,
-0.453315,
-0.633098,
-0.354225
],
[
-0.17374,
-0.600807,
-0.258416,
-0.380246,
-0.076232
],
[
-0.672244,
-0.166596,
0.347899,
0.221961,
-0.07373
],
[
0.25918,
-0.445785,
0.699829,
0.479309,
0.951773
]
],
"network.2.bias": [
0.588682,
0.214496,
-0.595319,
-0.109255,
0.78255
],
"network.4.weight": [
[
0.589969,
-0.336087,
0.108827,
-0.446314,
0.986849
],
[
-0.392093,
0.45332,
-0.044341,
0.300338,
-0.120207
],
[
0.063027,
-0.078277,
-0.36721,
0.056369,
-0.190713
],
[
-0.362951,
-0.089719,
-0.035969,
-0.22142,
-0.424396
],
[
-0.571098,
0.804071,
-0.083268,
-0.419644,
-0.546339
]
],
"network.4.bias": [
0.650863,
0.386239,
-0.635583,
-0.08702,
0.715579
],
"network.6.weight": [
[
1.188161,
-0.289077,
0.114721,
-0.42908,
-0.535296
],
[
1.129107,
-0.18894,
-0.270801,
-0.216499,
-0.667133
],
[
-0.434341,
-0.448443,
-0.314121,
-0.340095,
-0.765645
],
[
0.793312,
-0.188154,
-0.136274,
0.433892,
-0.663203
],
[
0.697722,
0.514323,
-0.08199,
0.281672,
0.447985
]
],
"network.6.bias": [
0.62492,
0.587377,
0.042101,
0.699975,
-0.021421
],
"network.8.weight": [
[
0.074177,
-0.107319,
-0.105117,
-0.029678,
-0.092964
],
[
0.996119,
0.672239,
0.149679,
0.90704,
0.153327
],
[
-0.496069,
-0.421864,
-0.163412,
-0.759388,
0.041515
],
[
-0.302005,
0.218972,
0.063319,
-0.398523,
0.273063
],
[
-0.494037,
-0.390588,
-0.423876,
-0.173145,
0.64188
]
],
"network.8.bias": [
-0.197562,
-0.228519,
1.034686,
-0.285029,
0.950433
],
"network.10.weight": [
[
-0.219898,
-0.190944,
0.786125,
0.176029,
0.905704
],
[
-0.059607,
-0.226763,
-0.41008,
0.016497,
-0.013595
],
[
-0.311035,
0.597817,
-0.432606,
-0.413965,
-0.035309
],
[
-0.391154,
-0.207717,
-0.554663,
0.19046,
-0.263792
],
[
-0.404038,
-0.055048,
0.153535,
-0.438568,
0.142796
]
],
"network.10.bias": [
0.707537,
-0.008378,
0.313014,
-0.100457,
-0.64538
],
"network.12.weight": [
[
0.733564,
-0.231835,
-0.736627,
-0.111804,
-0.171601
]
],
"network.12.bias": [
-0.167721
]
}
## Activation Signature
### 0
mean: [4.842091, 1.775374, 0.483650, 0.028173, 0.449324]
std: [3.166181, 2.078246, 2.136070, 1.046888, 2.340316]
### 2
mean: [0.257372, 1.949358, -3.106567, -3.266935, 3.285975]
std: [0.762667, 2.293204, 1.621871, 2.364141, 2.488432]
### 4
mean: [3.495814, 0.714993, -1.389547, -1.828467, 0.261229]
std: [2.664186, 1.101909, 0.457688, 0.906053, 2.507033]
### 6
mean: [4.079059, 3.771381, -2.657724, 2.690639, 3.337809]
std: [3.980827, 3.951144, 1.434915, 3.128899, 1.382859]
### 8
mean: [-0.724722, 10.466688, -5.240358, -1.057474, -1.325549]
std: [0.262236, 7.767156, 4.777078, 1.077974, 2.883852]
### 10
mean: [-0.427651, -2.489301, 6.381147, -2.571496, -1.143824]
std: [2.835357, 1.629254, 4.894650, 1.296927, 0.534474]
### 12
mean: [-4.198159]
std: [4.597022]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
decreasing_pairs
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 4.842090606689453, "std": 3.1661813259124756}, "1": {"mean": 1.775374174118042, "std": 2.0782456398010254}, "2": {"mean": 0.4836502969264984, "std": 2.136070489883423}, "3": {"mean": 0.028172820806503296, "std": 1.0468883514404297}, "4": {"mean": 0.44932377338409424, "std": 2.3403160572052}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.25737249851226807, "std": 0.7626673579216003}, "1": {"mean": 1.9493584632873535, "std": 2.2932040691375732}, "2": {"mean": -3.1065666675567627, "std": 1.6218714714050293}, "3": {"mean": -3.266934871673584, "std": 2.3641412258148193}, "4": {"mean": 3.285975217819214, "std": 2.4884324073791504}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 3.495814323425293, "std": 2.6641855239868164}, "1": {"mean": 0.7149932980537415, "std": 1.1019090414047241}, "2": {"mean": -1.3895468711853027, "std": 0.4576879143714905}, "3": {"mean": -1.8284666538238525, "std": 0.906053364276886}, "4": {"mean": 0.26122918725013733, "std": 2.507032632827759}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 4.079058647155762, "std": 3.9808266162872314}, "1": {"mean": 3.771380662918091, "std": 3.951144218444824}, "2": {"mean": -2.6577236652374268, "std": 1.4349151849746704}, "3": {"mean": 2.690639019012451, "std": 3.128899335861206}, "4": {"mean": 3.337808847427368, "std": 1.3828586339950562}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.7247222065925598, "std": 0.26223617792129517}, "1": {"mean": 10.46668815612793, "std": 7.767155647277832}, "2": {"mean": -5.240358352661133, "std": 4.777077674865723}, "3": {"mean": -1.0574744939804077, "std": 1.0779738426208496}, "4": {"mean": -1.3255493640899658, "std": 2.883852481842041}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -0.42765146493911743, "std": 2.835357427597046}, "1": {"mean": -2.4893012046813965, "std": 1.6292539834976196}, "2": {"mean": 6.381147384643555, "std": 4.894650459289551}, "3": {"mean": -2.57149600982666, "std": 1.2969272136688232}, "4": {"mean": -1.1438237428665161, "std": 0.5344743132591248}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -4.1981587409973145, "std": 4.59702205657959}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.500303, 0.360095, 0.487294, 0.607866, 0.848291], [0.550974, 0.636062, -0.065267, 0.280651, -0.434453], [-0.947375, 0.142487, 0.339304, 0.430514, -0.250845], [0.127817, -0.468976, -0.172216, 0.398061, -0.187961], [-0.862973, 0.044614, -0.135352, 0.514772, 0.601153]], "network.0.bias": [0.195653, 0.000992, 0.080358, 0.46163, -0.12064], "network.2.weight": [[-0.010073, 0.059244, 0.037662, 0.342042, -0.52094], [0.358063, 0.64124, -0.453315, -0.633098, -0.354225], [-0.17374, -0.600807, -0.258416, -0.380246, -0.076232], [-0.672244, -0.166596, 0.347899, 0.221961, -0.07373], [0.25918, -0.445785, 0.699829, 0.479309, 0.951773]], "network.2.bias": [0.588682, 0.214496, -0.595319, -0.109255, 0.78255], "network.4.weight": [[0.589969, -0.336087, 0.108827, -0.446314, 0.986849], [-0.392093, 0.45332, -0.044341, 0.300338, -0.120207], [0.063027, -0.078277, -0.36721, 0.056369, -0.190713], [-0.362951, -0.089719, -0.035969, -0.22142, -0.424396], [-0.571098, 0.804071, -0.083268, -0.419644, -0.546339]], "network.4.bias": [0.650863, 0.386239, -0.635583, -0.08702, 0.715579], "network.6.weight": [[1.188161, -0.289077, 0.114721, -0.42908, -0.535296], [1.129107, -0.18894, -0.270801, -0.216499, -0.667133], [-0.434341, -0.448443, -0.314121, -0.340095, -0.765645], [0.793312, -0.188154, -0.136274, 0.433892, -0.663203], [0.697722, 0.514323, -0.08199, 0.281672, 0.447985]], "network.6.bias": [0.62492, 0.587377, 0.042101, 0.699975, -0.021421], "network.8.weight": [[0.074177, -0.107319, -0.105117, -0.029678, -0.092964], [0.996119, 0.672239, 0.149679, 0.90704, 0.153327], [-0.496069, -0.421864, -0.163412, -0.759388, 0.041515], [-0.302005, 0.218972, 0.063319, -0.398523, 0.273063], [-0.494037, -0.390588, -0.423876, -0.173145, 0.64188]], "network.8.bias": [-0.197562, -0.228519, 1.034686, -0.285029, 0.950433], "network.10.weight": [[-0.219898, -0.190944, 0.786125, 0.176029, 0.905704], [-0.059607, -0.226763, -0.41008, 0.016497, -0.013595], [-0.311035, 0.597817, -0.432606, -0.413965, -0.035309], [-0.391154, -0.207717, -0.554663, 0.19046, -0.263792], [-0.404038, -0.055048, 0.153535, -0.438568, 0.142796]], "network.10.bias": [0.707537, -0.008378, 0.313014, -0.100457, -0.64538], "network.12.weight": [[0.733564, -0.231835, -0.736627, -0.111804, -0.171601]], "network.12.bias": [-0.167721]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6746063232421875, "train_acc": 0.595, "val_loss": 0.7366824150085449, "val_acc": 0.42}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6757084429264069, "train_acc": 0.595, "val_loss": 0.7434150576591492, "val_acc": 0.42}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6772147119045258, "train_acc": 0.595, "val_loss": 0.7370006442070007, "val_acc": 0.42}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6666013300418854, "train_acc": 0.595, "val_loss": 0.7140834927558899, "val_acc": 0.42}, {"stage": "degraded", "epoch": 4, "global_epoch": 4, "train_loss": 0.6566289067268372, "train_acc": 0.595, "val_loss": 0.6745942831039429, "val_acc": 0.42}, {"stage": "improved", "epoch": 0, "global_epoch": 5, "train_loss": 0.6379738748073578, "train_acc": 0.52, "val_loss": 0.5746404528617859, "val_acc": 0.94}, {"stage": "improved", "epoch": 1, "global_epoch": 6, "train_loss": 0.5374957025051117, "train_acc": 0.91, "val_loss": 0.4253078103065491, "val_acc": 0.92}, {"stage": "improved", "epoch": 2, "global_epoch": 7, "train_loss": 0.39731547236442566, "train_acc": 0.89, "val_loss": 0.20913323760032654, "val_acc": 0.96}, {"stage": "improved", "epoch": 3, "global_epoch": 8, "train_loss": 0.2263244315981865, "train_acc": 0.93, "val_loss": 0.16810180246829987, "val_acc": 0.94}, {"stage": "improved", "epoch": 4, "global_epoch": 9, "train_loss": 0.26500484347343445, "train_acc": 0.915, "val_loss": 0.17977140843868256, "val_acc": 0.92}, {"stage": "improved", "epoch": 5, "global_epoch": 10, "train_loss": 0.24243559688329697, "train_acc": 0.925, "val_loss": 0.10325291752815247, "val_acc": 0.98}, {"stage": "improved", "epoch": 6, "global_epoch": 11, "train_loss": 0.1772896945476532, "train_acc": 0.945, "val_loss": 0.09885694831609726, "val_acc": 0.98}, {"stage": "improved", "epoch": 7, "global_epoch": 12, "train_loss": 0.14707917720079422, "train_acc": 0.95, "val_loss": 0.12424915283918381, "val_acc": 0.98}, {"stage": "improved", "epoch": 8, "global_epoch": 13, "train_loss": 0.1781994104385376, "train_acc": 0.93, "val_loss": 0.12111902236938477, "val_acc": 0.98}, {"stage": "improved", "epoch": 9, "global_epoch": 14, "train_loss": 0.18345122039318085, "train_acc": 0.95, "val_loss": 0.1221982017159462, "val_acc": 0.98}], "summary": {"total_epochs": 15, "degraded_epochs": 5, "improved_epochs": 10, "patterns": ["decreasing_pairs"], "degraded_stage": {"initial_val_loss": 0.7366824150085449, "final_val_loss": 0.6745942831039429, "initial_val_acc": 0.42, "final_val_acc": 0.42, "best_val_acc": 0.42}, "improved_stage": {"initial_val_loss": 0.5746404528617859, "final_val_loss": 0.1221982017159462, "initial_val_acc": 0.94, "final_val_acc": 0.98, "best_val_acc": 0.98, "best_epoch": 10}, "improvement": 0.56, "first_improvement_epoch": 4}}
|
11
|
{"target_pattern": "ends_with", "degraded_accuracy": 0.5, "improved_accuracy": 0.88, "improvement": 0.38, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 1579, "learning_rate": 0.09508480999907651, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "ends_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["ends_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.061019,
0.045929,
0.324688,
0.282345,
0.823829
],
[
0.521457,
0.028359,
-0.426448,
-0.204695,
0.164566
],
[
0.799022,
0.067241,
-0.38051,
0.206821,
0.307067
],
[
-0.116515,
0.553502,
0.332544,
0.555975,
-1.050467
],
[
0.750737,
0.095216,
0.231354,
0.16183,
0.897603
]
],
"network.0.bias": [
0.244503,
0.214107,
-0.224526,
0.514649,
-0.09106
],
"network.2.weight": [
[
0.321115,
0.147578,
0.403441,
-0.367154,
0.478888
],
[
-0.129501,
-0.130216,
-0.236338,
-0.248167,
-0.366786
],
[
0.11121,
-0.317011,
-0.195938,
0.707295,
0.178555
],
[
0.23366,
0.426977,
-0.006594,
-0.2758,
0.147619
],
[
-0.18954,
-0.453033,
-0.259381,
0.286106,
-0.007467
]
],
"network.2.bias": [
0.23891,
-0.311053,
-0.138911,
0.449619,
-0.159489
],
"network.4.weight": [
[
0.016212,
-0.384459,
-0.633364,
-0.568734,
-0.247068
],
[
-0.101329,
0.383555,
-0.090229,
-0.339815,
-0.349275
],
[
0.637848,
-0.239994,
-0.30713,
0.376688,
-0.209152
],
[
0.829887,
-0.053306,
-0.196363,
0.379178,
-0.4462
],
[
-0.234125,
0.443686,
0.666016,
-0.596353,
0.324336
]
],
"network.4.bias": [
-0.110793,
-0.356415,
0.498037,
-0.040075,
0.279878
],
"network.6.weight": [
[
0.348501,
-0.268313,
0.542425,
0.340657,
-0.593789
],
[
0.32979,
-0.312879,
0.625874,
0.317809,
-0.616202
],
[
0.118304,
0.445564,
-0.210364,
-0.046302,
0.843875
],
[
0.109366,
0.406377,
-0.078904,
-0.189077,
-0.359533
],
[
0.062409,
0.144846,
-0.03843,
0.649942,
-0.392719
]
],
"network.6.bias": [
0.341756,
0.517815,
0.256439,
-0.152563,
0.042659
],
"network.8.weight": [
[
-0.384265,
-0.122846,
0.154629,
-0.011547,
-0.028732
],
[
0.161919,
-0.538537,
-0.005874,
0.278318,
-0.003688
],
[
0.00359,
-0.465542,
-0.069714,
0.445914,
-0.10285
],
[
0.169405,
-0.42441,
0.22399,
-0.103636,
-0.037213
],
[
0.784156,
0.641011,
-0.549717,
0.358206,
0.37365
]
],
"network.8.bias": [
0.265441,
-0.667321,
-0.252814,
0.335133,
0.286779
],
"network.10.weight": [
[
-0.18647,
0.167353,
0.016092,
-0.433024,
0.517016
],
[
-0.5067,
-0.037788,
-0.344543,
-0.304696,
0.626862
],
[
-0.11508,
-0.339957,
-0.148249,
0.440622,
-0.347431
],
[
0.156757,
-0.325867,
-0.420761,
-0.189531,
-0.057055
],
[
0.135401,
-0.390672,
0.095918,
-0.175267,
-0.476543
]
],
"network.10.bias": [
-0.16227,
0.135356,
-0.286595,
-0.250702,
-0.448845
],
"network.12.weight": [
[
-0.271382,
-0.608143,
-0.407041,
0.253642,
0.175452
]
],
"network.12.bias": [
0.182645
]
}
## Activation Signature
### 0
mean: [2.552026, -0.212087, 0.919728, 1.969586, 2.954048]
std: [1.924298, 1.245903, 1.794801, 2.537413, 2.728461]
### 2
mean: [2.163411, -2.628368, 1.948181, 1.008291, -0.488917]
std: [2.583901, 1.558398, 1.475637, 1.157402, 1.097342]
### 4
mean: [-1.969609, -1.188326, 1.651953, 1.705666, 0.507232]
std: [0.926434, 0.565085, 2.229919, 2.698236, 1.880157]
### 6
mean: [1.395724, 1.656297, 0.653517, -1.044338, 0.852664]
std: [2.487045, 2.622955, 1.492736, 0.504279, 1.852053]
### 8
mean: [-0.562385, -1.473319, -1.375765, -0.048485, 2.918285]
std: [1.246939, 0.873551, 1.139159, 0.864554, 4.099606]
### 10
mean: [1.342371, 1.968747, -1.303381, -0.465987, -2.041885]
std: [2.113235, 2.555555, 1.413491, 0.192084, 1.765574]
### 12
mean: [-1.548844]
std: [1.972446]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
ends_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.061019,
0.045929,
0.324688,
0.282345,
0.823829
],
[
0.521457,
0.028359,
-0.426448,
-0.204695,
0.164566
],
[
0.799022,
0.067241,
-0.38051,
0.206821,
0.307067
],
[
-0.116515,
0.553502,
0.332544,
0.555975,
-1.050467
],
[
0.750737,
0.095216,
0.231354,
0.16183,
0.897603
]
],
"network.0.bias": [
0.244503,
0.214107,
-0.224526,
0.514649,
-0.09106
],
"network.2.weight": [
[
0.321115,
0.147578,
0.403441,
-0.367154,
0.478888
],
[
-0.129501,
-0.130216,
-0.236338,
-0.248167,
-0.366786
],
[
0.11121,
-0.317011,
-0.195938,
0.707295,
0.178555
],
[
0.23366,
0.426977,
-0.006594,
-0.2758,
0.147619
],
[
-0.18954,
-0.453033,
-0.259381,
0.286106,
-0.007467
]
],
"network.2.bias": [
0.23891,
-0.311053,
-0.138911,
0.449619,
-0.159489
],
"network.4.weight": [
[
0.016212,
-0.384459,
-0.633364,
-0.568734,
-0.247068
],
[
-0.101329,
0.383555,
-0.090229,
-0.339815,
-0.349275
],
[
0.637848,
-0.239994,
-0.30713,
0.376688,
-0.209152
],
[
0.829887,
-0.053306,
-0.196363,
0.379178,
-0.4462
],
[
-0.234125,
0.443686,
0.666016,
-0.596353,
0.324336
]
],
"network.4.bias": [
-0.110793,
-0.356415,
0.498037,
-0.040075,
0.279878
],
"network.6.weight": [
[
0.348501,
-0.268313,
0.542425,
0.340657,
-0.593789
],
[
0.32979,
-0.312879,
0.625874,
0.317809,
-0.616202
],
[
0.118304,
0.445564,
-0.210364,
-0.046302,
0.843875
],
[
0.109366,
0.406377,
-0.078904,
-0.189077,
-0.359533
],
[
0.062409,
0.144846,
-0.03843,
0.649942,
-0.392719
]
],
"network.6.bias": [
0.341756,
0.517815,
0.256439,
-0.152563,
0.042659
],
"network.8.weight": [
[
-0.384265,
-0.122846,
0.154629,
-0.011547,
-0.028732
],
[
0.161919,
-0.538537,
-0.005874,
0.278318,
-0.003688
],
[
0.00359,
-0.465542,
-0.069714,
0.445914,
-0.10285
],
[
0.169405,
-0.42441,
0.22399,
-0.103636,
-0.037213
],
[
0.784156,
0.641011,
-0.549717,
0.358206,
0.37365
]
],
"network.8.bias": [
0.265441,
-0.667321,
-0.252814,
0.335133,
0.286779
],
"network.10.weight": [
[
-0.18647,
0.167353,
0.016092,
-0.433024,
0.517016
],
[
-0.5067,
-0.037788,
-0.344543,
-0.304696,
0.626862
],
[
-0.11508,
-0.339957,
-0.148249,
0.440622,
-0.347431
],
[
0.156757,
-0.325867,
-0.420761,
-0.189531,
-0.057055
],
[
0.135401,
-0.390672,
0.095918,
-0.175267,
-0.476543
]
],
"network.10.bias": [
-0.16227,
0.135356,
-0.286595,
-0.250702,
-0.448845
],
"network.12.weight": [
[
-0.271382,
-0.608143,
-0.407041,
0.253642,
0.175452
]
],
"network.12.bias": [
0.182645
]
}
## Activation Signature
### 0
mean: [2.552026, -0.212087, 0.919728, 1.969586, 2.954048]
std: [1.924298, 1.245903, 1.794801, 2.537413, 2.728461]
### 2
mean: [2.163411, -2.628368, 1.948181, 1.008291, -0.488917]
std: [2.583901, 1.558398, 1.475637, 1.157402, 1.097342]
### 4
mean: [-1.969609, -1.188326, 1.651953, 1.705666, 0.507232]
std: [0.926434, 0.565085, 2.229919, 2.698236, 1.880157]
### 6
mean: [1.395724, 1.656297, 0.653517, -1.044338, 0.852664]
std: [2.487045, 2.622955, 1.492736, 0.504279, 1.852053]
### 8
mean: [-0.562385, -1.473319, -1.375765, -0.048485, 2.918285]
std: [1.246939, 0.873551, 1.139159, 0.864554, 4.099606]
### 10
mean: [1.342371, 1.968747, -1.303381, -0.465987, -2.041885]
std: [2.113235, 2.555555, 1.413491, 0.192084, 1.765574]
### 12
mean: [-1.548844]
std: [1.972446]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
ends_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 2.5520262718200684, "std": 1.9242981672286987}, "1": {"mean": -0.21208716928958893, "std": 1.2459025382995605}, "2": {"mean": 0.91972815990448, "std": 1.7948006391525269}, "3": {"mean": 1.9695864915847778, "std": 2.5374131202697754}, "4": {"mean": 2.9540483951568604, "std": 2.7284607887268066}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 2.1634106636047363, "std": 2.5839009284973145}, "1": {"mean": -2.6283676624298096, "std": 1.5583975315093994}, "2": {"mean": 1.94818115234375, "std": 1.4756370782852173}, "3": {"mean": 1.008291244506836, "std": 1.1574015617370605}, "4": {"mean": -0.4889173209667206, "std": 1.0973424911499023}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -1.969609260559082, "std": 0.9264336824417114}, "1": {"mean": -1.188326358795166, "std": 0.5650852918624878}, "2": {"mean": 1.6519529819488525, "std": 2.229919195175171}, "3": {"mean": 1.7056663036346436, "std": 2.6982357501983643}, "4": {"mean": 0.507232129573822, "std": 1.8801573514938354}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 1.3957242965698242, "std": 2.4870452880859375}, "1": {"mean": 1.6562974452972412, "std": 2.622955322265625}, "2": {"mean": 0.6535170078277588, "std": 1.492735505104065}, "3": {"mean": -1.044337511062622, "std": 0.5042794942855835}, "4": {"mean": 0.852663516998291, "std": 1.8520525693893433}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.5623849034309387, "std": 1.2469385862350464}, "1": {"mean": -1.4733188152313232, "std": 0.8735511898994446}, "2": {"mean": -1.3757648468017578, "std": 1.1391589641571045}, "3": {"mean": -0.048485346138477325, "std": 0.8645544648170471}, "4": {"mean": 2.918285369873047, "std": 4.099605560302734}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": 1.342370629310608, "std": 2.113234758377075}, "1": {"mean": 1.9687466621398926, "std": 2.5555548667907715}, "2": {"mean": -1.3033807277679443, "std": 1.413490653038025}, "3": {"mean": -0.4659866690635681, "std": 0.19208373129367828}, "4": {"mean": -2.0418853759765625, "std": 1.7655737400054932}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -1.5488437414169312, "std": 1.9724462032318115}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.061019, 0.045929, 0.324688, 0.282345, 0.823829], [0.521457, 0.028359, -0.426448, -0.204695, 0.164566], [0.799022, 0.067241, -0.38051, 0.206821, 0.307067], [-0.116515, 0.553502, 0.332544, 0.555975, -1.050467], [0.750737, 0.095216, 0.231354, 0.16183, 0.897603]], "network.0.bias": [0.244503, 0.214107, -0.224526, 0.514649, -0.09106], "network.2.weight": [[0.321115, 0.147578, 0.403441, -0.367154, 0.478888], [-0.129501, -0.130216, -0.236338, -0.248167, -0.366786], [0.11121, -0.317011, -0.195938, 0.707295, 0.178555], [0.23366, 0.426977, -0.006594, -0.2758, 0.147619], [-0.18954, -0.453033, -0.259381, 0.286106, -0.007467]], "network.2.bias": [0.23891, -0.311053, -0.138911, 0.449619, -0.159489], "network.4.weight": [[0.016212, -0.384459, -0.633364, -0.568734, -0.247068], [-0.101329, 0.383555, -0.090229, -0.339815, -0.349275], [0.637848, -0.239994, -0.30713, 0.376688, -0.209152], [0.829887, -0.053306, -0.196363, 0.379178, -0.4462], [-0.234125, 0.443686, 0.666016, -0.596353, 0.324336]], "network.4.bias": [-0.110793, -0.356415, 0.498037, -0.040075, 0.279878], "network.6.weight": [[0.348501, -0.268313, 0.542425, 0.340657, -0.593789], [0.32979, -0.312879, 0.625874, 0.317809, -0.616202], [0.118304, 0.445564, -0.210364, -0.046302, 0.843875], [0.109366, 0.406377, -0.078904, -0.189077, -0.359533], [0.062409, 0.144846, -0.03843, 0.649942, -0.392719]], "network.6.bias": [0.341756, 0.517815, 0.256439, -0.152563, 0.042659], "network.8.weight": [[-0.384265, -0.122846, 0.154629, -0.011547, -0.028732], [0.161919, -0.538537, -0.005874, 0.278318, -0.003688], [0.00359, -0.465542, -0.069714, 0.445914, -0.10285], [0.169405, -0.42441, 0.22399, -0.103636, -0.037213], [0.784156, 0.641011, -0.549717, 0.358206, 0.37365]], "network.8.bias": [0.265441, -0.667321, -0.252814, 0.335133, 0.286779], "network.10.weight": [[-0.18647, 0.167353, 0.016092, -0.433024, 0.517016], [-0.5067, -0.037788, -0.344543, -0.304696, 0.626862], [-0.11508, -0.339957, -0.148249, 0.440622, -0.347431], [0.156757, -0.325867, -0.420761, -0.189531, -0.057055], [0.135401, -0.390672, 0.095918, -0.175267, -0.476543]], "network.10.bias": [-0.16227, 0.135356, -0.286595, -0.250702, -0.448845], "network.12.weight": [[-0.271382, -0.608143, -0.407041, 0.253642, 0.175452]], "network.12.bias": [0.182645]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7276186645030975, "train_acc": 0.575, "val_loss": 0.7131456136703491, "val_acc": 0.5}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.672491192817688, "train_acc": 0.575, "val_loss": 0.672955334186554, "val_acc": 0.5}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.6501280963420868, "train_acc": 0.5, "val_loss": 0.4741523861885071, "val_acc": 0.7}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.6604024767875671, "train_acc": 0.68, "val_loss": 0.4302540719509125, "val_acc": 0.88}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.47391873598098755, "train_acc": 0.835, "val_loss": 0.5643723011016846, "val_acc": 0.72}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.5855925679206848, "train_acc": 0.71, "val_loss": 0.6301978826522827, "val_acc": 0.64}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.6101351678371429, "train_acc": 0.69, "val_loss": 0.5426819324493408, "val_acc": 0.72}], "summary": {"total_epochs": 7, "degraded_epochs": 2, "improved_epochs": 5, "patterns": ["ends_with"], "degraded_stage": {"initial_val_loss": 0.7131456136703491, "final_val_loss": 0.672955334186554, "initial_val_acc": 0.5, "final_val_acc": 0.5, "best_val_acc": 0.5}, "improved_stage": {"initial_val_loss": 0.4741523861885071, "final_val_loss": 0.5426819324493408, "initial_val_acc": 0.7, "final_val_acc": 0.72, "best_val_acc": 0.88, "best_epoch": 3}, "improvement": 0.38, "first_improvement_epoch": 1}}
|
12
|
{"target_pattern": "ends_with", "degraded_accuracy": 0.52, "improved_accuracy": 0.88, "improvement": 0.36, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 2113, "learning_rate": 0.0493680412770662, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "ends_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["ends_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.380167,
-0.076265,
-0.024024,
0.310843,
1.081954
],
[
0.029359,
-0.894335,
-0.126894,
0.191763,
-0.378988
],
[
0.499216,
-0.214639,
-0.060103,
-0.082043,
0.641965
],
[
0.71436,
-0.285882,
-0.346568,
-0.22516,
0.341957
],
[
0.106301,
-0.112347,
0.014273,
0.04629,
1.159091
],
[
0.406162,
0.22607,
0.387615,
0.365951,
-0.424515
],
[
0.590784,
-0.153996,
-0.058973,
0.058372,
1.026748
],
[
0.149622,
-0.751413,
-0.152278,
0.446588,
-0.440945
]
],
"network.0.bias": [
0.184125,
-0.544164,
-0.33246,
0.496258,
-0.031385,
0.077663,
0.310337,
-0.217095
],
"network.2.weight": [
[
-0.27906,
-0.363729,
-0.0089,
-0.168183,
0.059374,
-0.067985,
-0.262585,
-0.374118
],
[
-0.041164,
0.138115,
-0.162173,
-0.130935,
-0.219339,
0.638675,
-0.065337,
0.544822
],
[
0.688939,
-0.221733,
0.417118,
0.466323,
0.543503,
0.038643,
0.555419,
-0.230458
],
[
0.529739,
-0.492055,
0.260246,
-0.399163,
0.460207,
0.021736,
0.168022,
-0.144294
],
[
-0.286236,
-0.070179,
-0.054812,
-0.121101,
-0.014438,
-0.186789,
0.010968,
0.043248
],
[
0.163191,
0.023297,
0.006178,
0.198171,
-0.350997,
0.180868,
-0.375969,
0.576097
],
[
0.663169,
-0.21906,
0.615508,
0.452338,
0.880674,
-0.352718,
0.487077,
0.069553
],
[
0.292732,
0.290656,
0.521926,
0.820577,
0.410427,
-0.053692,
0.818413,
-0.402754
]
],
"network.2.bias": [
-0.121135,
0.147279,
-0.342908,
-0.198974,
-0.09016,
0.422388,
0.086806,
-0.106426
],
"network.4.weight": [
[
0.058497,
-0.403779,
-0.453414,
-0.119542,
0.353487,
-0.016063,
-0.12254,
-0.431535
],
[
-0.053336,
-0.265469,
0.220669,
0.352509,
0.120276,
-0.136372,
0.710949,
0.31745
],
[
0.027403,
0.587864,
-0.168878,
-0.345953,
-0.191054,
0.501877,
0.013984,
-0.274433
],
[
0.109522,
0.637252,
-0.074726,
-0.569279,
0.34501,
0.145231,
-0.221736,
-0.161247
],
[
-0.255065,
0.07044,
-0.322439,
-0.298022,
0.343457,
-0.258019,
-0.095192,
-0.167889
],
[
-0.198677,
-0.250577,
-0.089179,
-0.109201,
-0.071261,
-0.272544,
-0.39751,
-0.472706
],
[
0.183459,
-0.474199,
-0.463417,
-0.013228,
-0.319691,
-0.15803,
-0.153161,
0.02518
],
[
0.382647,
-0.061528,
0.742751,
0.714883,
-0.230892,
-0.492096,
0.627942,
0.499192
]
],
"network.4.bias": [
-0.075542,
0.024516,
0.189892,
0.570129,
-0.231033,
0.040825,
-0.468005,
-0.242885
],
"network.6.weight": [
[
-0.113514,
0.773067,
-0.553434,
-0.283789,
-0.145291,
0.03534,
0.359773,
0.75359
],
[
0.108356,
0.360276,
-0.167597,
-0.453796,
-0.10168,
-0.219976,
0.175,
0.392381
],
[
0.033929,
-0.128443,
0.306413,
0.880126,
-0.323039,
0.147146,
-0.323045,
0.081823
],
[
0.117826,
0.302716,
-0.391255,
-0.249315,
-0.264762,
0.048202,
0.291921,
0.454143
],
[
-0.018137,
0.097024,
-0.354415,
-0.511072,
-0.197818,
-0.090182,
0.15493,
-0.501649
],
[
-0.06209,
0.270599,
-0.011587,
-0.034812,
0.248334,
0.112002,
-0.135026,
-0.261898
],
[
0.316933,
-0.200128,
0.114415,
-0.230337,
-0.224393,
0.147749,
-0.201769,
-0.30052
],
[
0.312026,
0.261299,
-0.270382,
-0.340386,
-0.130701,
-0.296787,
0.326682,
-0.286581
]
],
"network.6.bias": [
-0.163248,
-0.067774,
0.687364,
-0.194687,
-0.087606,
-0.132066,
-0.294264,
-0.332274
],
"network.8.weight": [
[
-0.349795,
0.087622,
-0.043809,
0.034299,
0.012802,
-0.015345,
0.274026,
0.303684
],
[
0.705507,
0.81921,
-0.617511,
0.660238,
-0.206892,
0.274091,
0.030447,
-0.037796
],
[
-0.114874,
0.108084,
-0.00871,
-0.311254,
-0.000595,
-0.231538,
0.082083,
0.3392
],
[
0.88058,
0.308656,
-0.273493,
0.348601,
0.130916,
0.201927,
-0.349453,
0.308774
],
[
-0.261232,
0.19483,
-0.211403,
0.240597,
-0.126818,
-0.217897,
-0.331828,
-0.053739
],
[
0.805523,
0.375367,
-0.560952,
0.586015,
-0.161654,
-0.121229,
-0.077765,
-0.210049
],
[
-0.183206,
0.062032,
0.391163,
-0.223781,
0.068024,
0.285176,
-0.098184,
0.295615
],
[
0.67041,
0.317432,
-0.137718,
0.136774,
0.104849,
0.087662,
0.119974,
-0.16328
]
],
"network.8.bias": [
-0.249998,
-0.092736,
-0.097048,
-0.289903,
-0.230226,
-0.071218,
-0.234594,
-0.342088
],
"network.10.weight": [
[
-0.178607,
-0.427613,
0.06221,
-0.718596,
-0.078615,
-0.450579,
0.372863,
-0.602472
]
],
"network.10.bias": [
0.764042
]
}
## Activation Signature
### 0
mean: [1.522868, -2.458567, 0.387940, 0.079259, 1.454806, 2.065937, 2.033228, -1.304379]
std: [2.213118, 1.755429, 1.638102, 1.667467, 2.209139, 1.695544, 2.365091, 1.543893]
### 2
mean: [-1.369515, 0.809383, 3.509305, 1.695913, -1.091680, 0.001741, 3.638041, 3.538389]
std: [1.061645, 1.489327, 4.369289, 2.491552, 0.686985, 1.313407, 5.155339, 4.440750]
### 4
mean: [-4.334641, 4.841694, -1.021251, -1.307603, -2.897721, -4.053488, -3.217158, 7.426809]
std: [4.555426, 6.940606, 3.190628, 3.914365, 3.255065, 4.489061, 2.468007, 10.425055]
### 6
mean: [8.756081, 4.224813, 1.521628, 4.277085, -4.003280, -0.801114, -3.640117, -1.632679]
std: [13.513037, 6.849566, 0.997402, 7.096313, 4.138527, 0.835111, 4.419350, 0.969163]
### 8
mean: [-2.992166, 12.390233, -2.145657, 10.517820, -0.939982, 11.031054, -2.107069, 7.765997]
std: [3.761368, 19.514559, 2.901664, 16.131618, 0.422289, 17.374537, 3.771235, 11.907553]
### 10
mean: [-22.725161]
std: [34.245331]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
ends_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.380167,
-0.076265,
-0.024024,
0.310843,
1.081954
],
[
0.029359,
-0.894335,
-0.126894,
0.191763,
-0.378988
],
[
0.499216,
-0.214639,
-0.060103,
-0.082043,
0.641965
],
[
0.71436,
-0.285882,
-0.346568,
-0.22516,
0.341957
],
[
0.106301,
-0.112347,
0.014273,
0.04629,
1.159091
],
[
0.406162,
0.22607,
0.387615,
0.365951,
-0.424515
],
[
0.590784,
-0.153996,
-0.058973,
0.058372,
1.026748
],
[
0.149622,
-0.751413,
-0.152278,
0.446588,
-0.440945
]
],
"network.0.bias": [
0.184125,
-0.544164,
-0.33246,
0.496258,
-0.031385,
0.077663,
0.310337,
-0.217095
],
"network.2.weight": [
[
-0.27906,
-0.363729,
-0.0089,
-0.168183,
0.059374,
-0.067985,
-0.262585,
-0.374118
],
[
-0.041164,
0.138115,
-0.162173,
-0.130935,
-0.219339,
0.638675,
-0.065337,
0.544822
],
[
0.688939,
-0.221733,
0.417118,
0.466323,
0.543503,
0.038643,
0.555419,
-0.230458
],
[
0.529739,
-0.492055,
0.260246,
-0.399163,
0.460207,
0.021736,
0.168022,
-0.144294
],
[
-0.286236,
-0.070179,
-0.054812,
-0.121101,
-0.014438,
-0.186789,
0.010968,
0.043248
],
[
0.163191,
0.023297,
0.006178,
0.198171,
-0.350997,
0.180868,
-0.375969,
0.576097
],
[
0.663169,
-0.21906,
0.615508,
0.452338,
0.880674,
-0.352718,
0.487077,
0.069553
],
[
0.292732,
0.290656,
0.521926,
0.820577,
0.410427,
-0.053692,
0.818413,
-0.402754
]
],
"network.2.bias": [
-0.121135,
0.147279,
-0.342908,
-0.198974,
-0.09016,
0.422388,
0.086806,
-0.106426
],
"network.4.weight": [
[
0.058497,
-0.403779,
-0.453414,
-0.119542,
0.353487,
-0.016063,
-0.12254,
-0.431535
],
[
-0.053336,
-0.265469,
0.220669,
0.352509,
0.120276,
-0.136372,
0.710949,
0.31745
],
[
0.027403,
0.587864,
-0.168878,
-0.345953,
-0.191054,
0.501877,
0.013984,
-0.274433
],
[
0.109522,
0.637252,
-0.074726,
-0.569279,
0.34501,
0.145231,
-0.221736,
-0.161247
],
[
-0.255065,
0.07044,
-0.322439,
-0.298022,
0.343457,
-0.258019,
-0.095192,
-0.167889
],
[
-0.198677,
-0.250577,
-0.089179,
-0.109201,
-0.071261,
-0.272544,
-0.39751,
-0.472706
],
[
0.183459,
-0.474199,
-0.463417,
-0.013228,
-0.319691,
-0.15803,
-0.153161,
0.02518
],
[
0.382647,
-0.061528,
0.742751,
0.714883,
-0.230892,
-0.492096,
0.627942,
0.499192
]
],
"network.4.bias": [
-0.075542,
0.024516,
0.189892,
0.570129,
-0.231033,
0.040825,
-0.468005,
-0.242885
],
"network.6.weight": [
[
-0.113514,
0.773067,
-0.553434,
-0.283789,
-0.145291,
0.03534,
0.359773,
0.75359
],
[
0.108356,
0.360276,
-0.167597,
-0.453796,
-0.10168,
-0.219976,
0.175,
0.392381
],
[
0.033929,
-0.128443,
0.306413,
0.880126,
-0.323039,
0.147146,
-0.323045,
0.081823
],
[
0.117826,
0.302716,
-0.391255,
-0.249315,
-0.264762,
0.048202,
0.291921,
0.454143
],
[
-0.018137,
0.097024,
-0.354415,
-0.511072,
-0.197818,
-0.090182,
0.15493,
-0.501649
],
[
-0.06209,
0.270599,
-0.011587,
-0.034812,
0.248334,
0.112002,
-0.135026,
-0.261898
],
[
0.316933,
-0.200128,
0.114415,
-0.230337,
-0.224393,
0.147749,
-0.201769,
-0.30052
],
[
0.312026,
0.261299,
-0.270382,
-0.340386,
-0.130701,
-0.296787,
0.326682,
-0.286581
]
],
"network.6.bias": [
-0.163248,
-0.067774,
0.687364,
-0.194687,
-0.087606,
-0.132066,
-0.294264,
-0.332274
],
"network.8.weight": [
[
-0.349795,
0.087622,
-0.043809,
0.034299,
0.012802,
-0.015345,
0.274026,
0.303684
],
[
0.705507,
0.81921,
-0.617511,
0.660238,
-0.206892,
0.274091,
0.030447,
-0.037796
],
[
-0.114874,
0.108084,
-0.00871,
-0.311254,
-0.000595,
-0.231538,
0.082083,
0.3392
],
[
0.88058,
0.308656,
-0.273493,
0.348601,
0.130916,
0.201927,
-0.349453,
0.308774
],
[
-0.261232,
0.19483,
-0.211403,
0.240597,
-0.126818,
-0.217897,
-0.331828,
-0.053739
],
[
0.805523,
0.375367,
-0.560952,
0.586015,
-0.161654,
-0.121229,
-0.077765,
-0.210049
],
[
-0.183206,
0.062032,
0.391163,
-0.223781,
0.068024,
0.285176,
-0.098184,
0.295615
],
[
0.67041,
0.317432,
-0.137718,
0.136774,
0.104849,
0.087662,
0.119974,
-0.16328
]
],
"network.8.bias": [
-0.249998,
-0.092736,
-0.097048,
-0.289903,
-0.230226,
-0.071218,
-0.234594,
-0.342088
],
"network.10.weight": [
[
-0.178607,
-0.427613,
0.06221,
-0.718596,
-0.078615,
-0.450579,
0.372863,
-0.602472
]
],
"network.10.bias": [
0.764042
]
}
## Activation Signature
### 0
mean: [1.522868, -2.458567, 0.387940, 0.079259, 1.454806, 2.065937, 2.033228, -1.304379]
std: [2.213118, 1.755429, 1.638102, 1.667467, 2.209139, 1.695544, 2.365091, 1.543893]
### 2
mean: [-1.369515, 0.809383, 3.509305, 1.695913, -1.091680, 0.001741, 3.638041, 3.538389]
std: [1.061645, 1.489327, 4.369289, 2.491552, 0.686985, 1.313407, 5.155339, 4.440750]
### 4
mean: [-4.334641, 4.841694, -1.021251, -1.307603, -2.897721, -4.053488, -3.217158, 7.426809]
std: [4.555426, 6.940606, 3.190628, 3.914365, 3.255065, 4.489061, 2.468007, 10.425055]
### 6
mean: [8.756081, 4.224813, 1.521628, 4.277085, -4.003280, -0.801114, -3.640117, -1.632679]
std: [13.513037, 6.849566, 0.997402, 7.096313, 4.138527, 0.835111, 4.419350, 0.969163]
### 8
mean: [-2.992166, 12.390233, -2.145657, 10.517820, -0.939982, 11.031054, -2.107069, 7.765997]
std: [3.761368, 19.514559, 2.901664, 16.131618, 0.422289, 17.374537, 3.771235, 11.907553]
### 10
mean: [-22.725161]
std: [34.245331]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
ends_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.5228681564331055, "std": 2.213118076324463}, "1": {"mean": -2.4585673809051514, "std": 1.7554290294647217}, "2": {"mean": 0.38793984055519104, "std": 1.6381019353866577}, "3": {"mean": 0.07925930619239807, "std": 1.667466640472412}, "4": {"mean": 1.4548063278198242, "std": 2.209139108657837}, "5": {"mean": 2.06593656539917, "std": 1.6955441236495972}, "6": {"mean": 2.0332281589508057, "std": 2.365091323852539}, "7": {"mean": -1.3043793439865112, "std": 1.5438932180404663}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -1.3695145845413208, "std": 1.061645269393921}, "1": {"mean": 0.8093833327293396, "std": 1.4893269538879395}, "2": {"mean": 3.509305477142334, "std": 4.369288921356201}, "3": {"mean": 1.6959127187728882, "std": 2.491551637649536}, "4": {"mean": -1.0916796922683716, "std": 0.686985433101654}, "5": {"mean": 0.001740535139106214, "std": 1.3134068250656128}, "6": {"mean": 3.6380412578582764, "std": 5.155338764190674}, "7": {"mean": 3.538388729095459, "std": 4.4407501220703125}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -4.3346405029296875, "std": 4.555425643920898}, "1": {"mean": 4.841694355010986, "std": 6.940606117248535}, "2": {"mean": -1.0212514400482178, "std": 3.1906278133392334}, "3": {"mean": -1.3076026439666748, "std": 3.9143645763397217}, "4": {"mean": -2.8977205753326416, "std": 3.255065441131592}, "5": {"mean": -4.053487777709961, "std": 4.48906135559082}, "6": {"mean": -3.2171576023101807, "std": 2.4680070877075195}, "7": {"mean": 7.426808834075928, "std": 10.425054550170898}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 8.756080627441406, "std": 13.513036727905273}, "1": {"mean": 4.224812984466553, "std": 6.8495659828186035}, "2": {"mean": 1.52162766456604, "std": 0.9974024295806885}, "3": {"mean": 4.277084827423096, "std": 7.096312522888184}, "4": {"mean": -4.003279685974121, "std": 4.1385273933410645}, "5": {"mean": -0.8011135458946228, "std": 0.835110604763031}, "6": {"mean": -3.6401166915893555, "std": 4.4193501472473145}, "7": {"mean": -1.6326794624328613, "std": 0.9691630601882935}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -2.992166042327881, "std": 3.7613677978515625}, "1": {"mean": 12.390233039855957, "std": 19.514558792114258}, "2": {"mean": -2.1456565856933594, "std": 2.9016640186309814}, "3": {"mean": 10.517820358276367, "std": 16.13161849975586}, "4": {"mean": -0.9399815201759338, "std": 0.42228853702545166}, "5": {"mean": 11.031054496765137, "std": 17.374536514282227}, "6": {"mean": -2.1070687770843506, "std": 3.7712345123291016}, "7": {"mean": 7.765997409820557, "std": 11.907552719116211}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -22.725160598754883, "std": 34.245330810546875}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.380167, -0.076265, -0.024024, 0.310843, 1.081954], [0.029359, -0.894335, -0.126894, 0.191763, -0.378988], [0.499216, -0.214639, -0.060103, -0.082043, 0.641965], [0.71436, -0.285882, -0.346568, -0.22516, 0.341957], [0.106301, -0.112347, 0.014273, 0.04629, 1.159091], [0.406162, 0.22607, 0.387615, 0.365951, -0.424515], [0.590784, -0.153996, -0.058973, 0.058372, 1.026748], [0.149622, -0.751413, -0.152278, 0.446588, -0.440945]], "network.0.bias": [0.184125, -0.544164, -0.33246, 0.496258, -0.031385, 0.077663, 0.310337, -0.217095], "network.2.weight": [[-0.27906, -0.363729, -0.0089, -0.168183, 0.059374, -0.067985, -0.262585, -0.374118], [-0.041164, 0.138115, -0.162173, -0.130935, -0.219339, 0.638675, -0.065337, 0.544822], [0.688939, -0.221733, 0.417118, 0.466323, 0.543503, 0.038643, 0.555419, -0.230458], [0.529739, -0.492055, 0.260246, -0.399163, 0.460207, 0.021736, 0.168022, -0.144294], [-0.286236, -0.070179, -0.054812, -0.121101, -0.014438, -0.186789, 0.010968, 0.043248], [0.163191, 0.023297, 0.006178, 0.198171, -0.350997, 0.180868, -0.375969, 0.576097], [0.663169, -0.21906, 0.615508, 0.452338, 0.880674, -0.352718, 0.487077, 0.069553], [0.292732, 0.290656, 0.521926, 0.820577, 0.410427, -0.053692, 0.818413, -0.402754]], "network.2.bias": [-0.121135, 0.147279, -0.342908, -0.198974, -0.09016, 0.422388, 0.086806, -0.106426], "network.4.weight": [[0.058497, -0.403779, -0.453414, -0.119542, 0.353487, -0.016063, -0.12254, -0.431535], [-0.053336, -0.265469, 0.220669, 0.352509, 0.120276, -0.136372, 0.710949, 0.31745], [0.027403, 0.587864, -0.168878, -0.345953, -0.191054, 0.501877, 0.013984, -0.274433], [0.109522, 0.637252, -0.074726, -0.569279, 0.34501, 0.145231, -0.221736, -0.161247], [-0.255065, 0.07044, -0.322439, -0.298022, 0.343457, -0.258019, -0.095192, -0.167889], [-0.198677, -0.250577, -0.089179, -0.109201, -0.071261, -0.272544, -0.39751, -0.472706], [0.183459, -0.474199, -0.463417, -0.013228, -0.319691, -0.15803, -0.153161, 0.02518], [0.382647, -0.061528, 0.742751, 0.714883, -0.230892, -0.492096, 0.627942, 0.499192]], "network.4.bias": [-0.075542, 0.024516, 0.189892, 0.570129, -0.231033, 0.040825, -0.468005, -0.242885], "network.6.weight": [[-0.113514, 0.773067, -0.553434, -0.283789, -0.145291, 0.03534, 0.359773, 0.75359], [0.108356, 0.360276, -0.167597, -0.453796, -0.10168, -0.219976, 0.175, 0.392381], [0.033929, -0.128443, 0.306413, 0.880126, -0.323039, 0.147146, -0.323045, 0.081823], [0.117826, 0.302716, -0.391255, -0.249315, -0.264762, 0.048202, 0.291921, 0.454143], [-0.018137, 0.097024, -0.354415, -0.511072, -0.197818, -0.090182, 0.15493, -0.501649], [-0.06209, 0.270599, -0.011587, -0.034812, 0.248334, 0.112002, -0.135026, -0.261898], [0.316933, -0.200128, 0.114415, -0.230337, -0.224393, 0.147749, -0.201769, -0.30052], [0.312026, 0.261299, -0.270382, -0.340386, -0.130701, -0.296787, 0.326682, -0.286581]], "network.6.bias": [-0.163248, -0.067774, 0.687364, -0.194687, -0.087606, -0.132066, -0.294264, -0.332274], "network.8.weight": [[-0.349795, 0.087622, -0.043809, 0.034299, 0.012802, -0.015345, 0.274026, 0.303684], [0.705507, 0.81921, -0.617511, 0.660238, -0.206892, 0.274091, 0.030447, -0.037796], [-0.114874, 0.108084, -0.00871, -0.311254, -0.000595, -0.231538, 0.082083, 0.3392], [0.88058, 0.308656, -0.273493, 0.348601, 0.130916, 0.201927, -0.349453, 0.308774], [-0.261232, 0.19483, -0.211403, 0.240597, -0.126818, -0.217897, -0.331828, -0.053739], [0.805523, 0.375367, -0.560952, 0.586015, -0.161654, -0.121229, -0.077765, -0.210049], [-0.183206, 0.062032, 0.391163, -0.223781, 0.068024, 0.285176, -0.098184, 0.295615], [0.67041, 0.317432, -0.137718, 0.136774, 0.104849, 0.087662, 0.119974, -0.16328]], "network.8.bias": [-0.249998, -0.092736, -0.097048, -0.289903, -0.230226, -0.071218, -0.234594, -0.342088], "network.10.weight": [[-0.178607, -0.427613, 0.06221, -0.718596, -0.078615, -0.450579, 0.372863, -0.602472]], "network.10.bias": [0.764042]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7061075866222382, "train_acc": 0.435, "val_loss": 0.6892942786216736, "val_acc": 0.52}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6757219135761261, "train_acc": 0.575, "val_loss": 0.6789021492004395, "val_acc": 0.52}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6358359158039093, "train_acc": 0.575, "val_loss": 0.5983940958976746, "val_acc": 0.52}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.539980560541153, "train_acc": 0.495, "val_loss": 0.45247164368629456, "val_acc": 0.82}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.41949795186519623, "train_acc": 0.855, "val_loss": 0.39890480041503906, "val_acc": 0.86}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.3870571106672287, "train_acc": 0.89, "val_loss": 0.38579994440078735, "val_acc": 0.84}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.35907433927059174, "train_acc": 0.89, "val_loss": 0.5199134945869446, "val_acc": 0.84}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.33808697760105133, "train_acc": 0.89, "val_loss": 0.4288002848625183, "val_acc": 0.86}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.3059087544679642, "train_acc": 0.9, "val_loss": 0.3227787911891937, "val_acc": 0.86}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.28323712944984436, "train_acc": 0.9, "val_loss": 0.2966734766960144, "val_acc": 0.88}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.26352112740278244, "train_acc": 0.9, "val_loss": 0.3630027770996094, "val_acc": 0.86}, {"stage": "improved", "epoch": 8, "global_epoch": 11, "train_loss": 0.26287442445755005, "train_acc": 0.9, "val_loss": 0.4885023534297943, "val_acc": 0.84}, {"stage": "improved", "epoch": 9, "global_epoch": 12, "train_loss": 0.3108274042606354, "train_acc": 0.9, "val_loss": 0.49763232469558716, "val_acc": 0.84}], "summary": {"total_epochs": 13, "degraded_epochs": 3, "improved_epochs": 10, "patterns": ["ends_with"], "degraded_stage": {"initial_val_loss": 0.6892942786216736, "final_val_loss": 0.5983940958976746, "initial_val_acc": 0.52, "final_val_acc": 0.52, "best_val_acc": 0.52}, "improved_stage": {"initial_val_loss": 0.45247164368629456, "final_val_loss": 0.49763232469558716, "initial_val_acc": 0.82, "final_val_acc": 0.84, "best_val_acc": 0.88, "best_epoch": 9}, "improvement": 0.36, "first_improvement_epoch": 2}}
|
13
|
{"target_pattern": "no_repeats", "degraded_accuracy": 0.7, "improved_accuracy": 0.88, "improvement": 0.18000000000000005, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 7746, "learning_rate": 0.036877951559239695, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "no_repeats", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["no_repeats"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.688668,
0.186801,
-0.570155,
-0.033674,
-0.651347
],
[
-0.070155,
-0.296438,
-0.05982,
-0.066008,
-0.290094
],
[
-0.450714,
0.251016,
0.582487,
-0.028885,
0.732336
],
[
0.622606,
0.371871,
0.257956,
-0.280204,
-0.747301
],
[
0.042009,
0.194713,
-0.533164,
0.129308,
-0.930439
],
[
0.541737,
-0.86864,
0.454182,
-0.362061,
-0.118582
],
[
0.840696,
0.006576,
-0.35385,
0.330622,
-0.717408
],
[
0.322608,
-0.532125,
-0.036308,
-0.089079,
0.606855
]
],
"network.0.bias": [
-0.15908,
-0.327208,
0.364263,
0.768137,
0.000541,
0.251528,
0.289487,
0.641975
],
"network.2.weight": [
[
-0.160147,
0.206342,
0.58399,
-0.412795,
0.108054,
-0.477953,
-0.291728,
-0.767862
],
[
0.35562,
-0.343743,
-0.067234,
0.477687,
0.151602,
0.483631,
0.132714,
0.000881
],
[
0.493912,
0.194248,
-0.319403,
0.560612,
0.515356,
0.349381,
0.489256,
0.239758
],
[
-0.103124,
-0.137564,
0.356718,
-0.314897,
0.188524,
-0.400103,
0.157925,
-0.625852
],
[
-0.256529,
0.204227,
-0.32926,
0.070566,
-0.52351,
0.044988,
-0.033262,
-0.13065
],
[
-0.111184,
-0.31412,
-0.188134,
-0.203571,
0.090853,
-0.333245,
-0.154048,
0.12206
],
[
-0.162174,
-0.06376,
0.418355,
-0.126358,
0.338754,
0.075835,
-0.287578,
-0.572576
],
[
0.093981,
0.193527,
-0.502802,
-0.40804,
-0.147435,
0.013082,
-0.359133,
-0.063094
]
],
"network.2.bias": [
0.326585,
0.154568,
0.650883,
-0.102188,
0.181359,
-0.026824,
0.024173,
-0.0276
],
"network.4.weight": [
[
-0.155294,
0.51883,
0.421213,
0.143105,
-0.087047,
0.030563,
-0.221382,
0.195086
],
[
-0.199731,
0.015115,
-0.232097,
0.025164,
-0.091817,
-0.130878,
-0.244923,
-0.148356
],
[
-0.155585,
0.419377,
0.659027,
-0.504517,
0.315447,
0.213172,
-0.138787,
-0.144865
],
[
-0.112092,
0.32952,
-0.264814,
0.096127,
-0.320058,
-0.314027,
-0.197992,
0.013833
],
[
-0.146718,
-0.39142,
-0.188915,
-0.052719,
0.198253,
0.136368,
-0.545484,
0.395407
],
[
-0.531848,
0.43783,
-0.012102,
0.003725,
0.009807,
0.016,
-0.281076,
-0.033706
],
[
0.739484,
-0.368589,
-0.178331,
0.570549,
0.488028,
-0.01328,
0.304513,
-0.198414
],
[
0.030582,
0.210968,
0.895869,
-0.506172,
0.317413,
-0.002714,
0.348988,
-0.145547
]
],
"network.4.bias": [
0.052354,
-0.335031,
0.201355,
-0.15171,
-0.151362,
0.400161,
0.2235,
0.589509
],
"network.6.weight": [
[
-0.198919,
0.189116,
0.00834,
0.249198,
0.269863,
0.283044,
-0.286131,
-0.048913
],
[
-0.020965,
0.234255,
-0.353259,
0.144067,
-0.061274,
0.169769,
-0.184536,
-0.216735
],
[
0.317142,
0.182032,
0.589724,
0.463473,
0.050838,
0.869024,
-0.326745,
0.316813
],
[
-0.267253,
-0.054415,
-0.497363,
0.065838,
0.01388,
0.401584,
-0.091142,
-0.238396
],
[
0.159757,
0.013914,
0.650166,
-0.033122,
-0.12651,
0.511938,
-0.426076,
0.789559
],
[
0.036908,
0.114666,
-0.217049,
-0.120711,
-0.167224,
0.208699,
-0.251878,
-0.332019
],
[
0.552419,
-0.20671,
0.601837,
-0.090438,
-0.064472,
0.406912,
-0.317873,
0.675
],
[
-0.086703,
-0.006697,
-0.060965,
0.079886,
-0.203472,
-0.397344,
0.689696,
0.018006
]
],
"network.6.bias": [
-0.242388,
-0.197396,
0.019618,
0.117503,
0.296871,
-0.19959,
0.5073,
0.146358
],
"network.8.weight": [
[
-0.404476,
0.081498,
0.604959,
-0.10901,
0.687896,
0.131854,
0.535899,
-0.159888
],
[
-0.004761,
0.340257,
0.51314,
-0.13325,
0.611188,
0.237585,
0.599384,
-0.387454
],
[
0.316094,
-0.306398,
-0.213902,
-0.272248,
-0.016868,
0.286093,
-0.239087,
0.340796
],
[
-0.039796,
-0.27664,
0.050163,
-0.303247,
-0.351489,
0.299391,
-0.264714,
-0.34363
],
[
0.092366,
0.040987,
-0.317252,
0.076024,
0.056689,
-0.186754,
-0.596717,
0.709301
],
[
-0.043603,
0.213391,
-0.148799,
0.023875,
-0.080774,
-0.180717,
0.558857,
0.076617
],
[
-0.234709,
0.29266,
0.213573,
-0.181494,
-0.068125,
-0.177662,
0.180862,
0.060476
],
[
0.252374,
-0.224632,
0.153081,
0.069098,
-0.198631,
-0.30378,
-0.204866,
0.412867
]
],
"network.8.bias": [
0.526378,
0.400971,
0.23469,
-0.021799,
0.206416,
0.149207,
0.13541,
0.242314
],
"network.10.weight": [
[
0.04282,
0.507103,
-0.489137,
-0.137697,
-0.317948,
-0.328721,
-0.437974,
-0.494849
],
[
-0.824575,
-0.758382,
0.285225,
0.015572,
0.600799,
0.156506,
0.393958,
0.397637
],
[
0.343337,
0.718864,
-0.009881,
-0.047269,
-0.434034,
0.38666,
0.361526,
-0.305664
],
[
0.747234,
0.60891,
-0.159455,
0.171798,
-0.308262,
-0.474506,
-0.143673,
-0.066018
],
[
0.170037,
-0.282807,
0.050458,
0.341707,
-0.146471,
0.153988,
-0.151922,
0.100865
],
[
-0.038398,
-0.239801,
-0.059448,
0.253666,
-0.107913,
-0.069285,
0.289029,
-0.338299
],
[
-0.172213,
-0.069052,
-0.320333,
0.258069,
-0.098832,
0.149351,
-0.253339,
-0.34549
],
[
-0.673727,
-0.011986,
-0.003642,
-0.007366,
0.609922,
0.072129,
0.342205,
0.441785
]
],
"network.10.bias": [
0.374342,
0.175768,
0.381225,
0.412715,
-0.29183,
-0.622215,
-0.304338,
0.298499
],
"network.12.weight": [
[
-0.496183,
0.388304,
-0.538868,
-0.654806,
0.09697,
-0.010081,
0.252938,
0.244775
]
],
"network.12.bias": [
-0.261173
]
}
## Activation Signature
### 0
mean: [-1.028208, -1.578978, 2.318805, 1.236856, -1.576337, -0.631101, 0.432712, 0.555378]
std: [1.975761, 0.923683, 1.900966, 2.192478, 2.157034, 2.177463, 2.015826, 1.601387]
### 2
mean: [-0.214167, 1.273706, 1.899954, -0.377768, -0.765640, -1.036367, 0.032790, -2.253067]
std: [1.868116, 1.404590, 2.195429, 1.111173, 0.620315, 0.713796, 1.112615, 1.132964]
### 4
mean: [1.393777, -0.983424, 1.781502, -0.373851, -1.359562, 0.524682, 0.063365, 2.681238]
std: [1.691497, 0.408552, 2.133758, 0.158420, 0.853109, 0.932321, 1.475539, 2.162709]
### 6
mean: [-0.620327, -1.456922, 2.824868, -1.612686, 3.952641, -1.444438, 4.326758, 0.087583]
std: [0.267747, 1.037413, 3.187500, 1.654218, 3.864513, 0.873402, 4.030807, 0.996125]
### 8
mean: [7.271826, 6.751927, -1.341323, -2.571192, -2.766790, 1.850561, 1.292486, -0.815053]
std: [6.733321, 6.497111, 1.828539, 2.135725, 3.457049, 1.449702, 1.109712, 1.286779]
### 10
mean: [2.744888, -9.942060, 8.797823, 8.816937, -0.892816, -2.363634, -2.190863, -3.913485]
std: [2.849319, 10.027461, 8.061817, 8.220969, 0.621569, 1.514423, 1.573909, 4.344307]
### 12
mean: [-12.136356]
std: [11.130861]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
no_repeats
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.688668,
0.186801,
-0.570155,
-0.033674,
-0.651347
],
[
-0.070155,
-0.296438,
-0.05982,
-0.066008,
-0.290094
],
[
-0.450714,
0.251016,
0.582487,
-0.028885,
0.732336
],
[
0.622606,
0.371871,
0.257956,
-0.280204,
-0.747301
],
[
0.042009,
0.194713,
-0.533164,
0.129308,
-0.930439
],
[
0.541737,
-0.86864,
0.454182,
-0.362061,
-0.118582
],
[
0.840696,
0.006576,
-0.35385,
0.330622,
-0.717408
],
[
0.322608,
-0.532125,
-0.036308,
-0.089079,
0.606855
]
],
"network.0.bias": [
-0.15908,
-0.327208,
0.364263,
0.768137,
0.000541,
0.251528,
0.289487,
0.641975
],
"network.2.weight": [
[
-0.160147,
0.206342,
0.58399,
-0.412795,
0.108054,
-0.477953,
-0.291728,
-0.767862
],
[
0.35562,
-0.343743,
-0.067234,
0.477687,
0.151602,
0.483631,
0.132714,
0.000881
],
[
0.493912,
0.194248,
-0.319403,
0.560612,
0.515356,
0.349381,
0.489256,
0.239758
],
[
-0.103124,
-0.137564,
0.356718,
-0.314897,
0.188524,
-0.400103,
0.157925,
-0.625852
],
[
-0.256529,
0.204227,
-0.32926,
0.070566,
-0.52351,
0.044988,
-0.033262,
-0.13065
],
[
-0.111184,
-0.31412,
-0.188134,
-0.203571,
0.090853,
-0.333245,
-0.154048,
0.12206
],
[
-0.162174,
-0.06376,
0.418355,
-0.126358,
0.338754,
0.075835,
-0.287578,
-0.572576
],
[
0.093981,
0.193527,
-0.502802,
-0.40804,
-0.147435,
0.013082,
-0.359133,
-0.063094
]
],
"network.2.bias": [
0.326585,
0.154568,
0.650883,
-0.102188,
0.181359,
-0.026824,
0.024173,
-0.0276
],
"network.4.weight": [
[
-0.155294,
0.51883,
0.421213,
0.143105,
-0.087047,
0.030563,
-0.221382,
0.195086
],
[
-0.199731,
0.015115,
-0.232097,
0.025164,
-0.091817,
-0.130878,
-0.244923,
-0.148356
],
[
-0.155585,
0.419377,
0.659027,
-0.504517,
0.315447,
0.213172,
-0.138787,
-0.144865
],
[
-0.112092,
0.32952,
-0.264814,
0.096127,
-0.320058,
-0.314027,
-0.197992,
0.013833
],
[
-0.146718,
-0.39142,
-0.188915,
-0.052719,
0.198253,
0.136368,
-0.545484,
0.395407
],
[
-0.531848,
0.43783,
-0.012102,
0.003725,
0.009807,
0.016,
-0.281076,
-0.033706
],
[
0.739484,
-0.368589,
-0.178331,
0.570549,
0.488028,
-0.01328,
0.304513,
-0.198414
],
[
0.030582,
0.210968,
0.895869,
-0.506172,
0.317413,
-0.002714,
0.348988,
-0.145547
]
],
"network.4.bias": [
0.052354,
-0.335031,
0.201355,
-0.15171,
-0.151362,
0.400161,
0.2235,
0.589509
],
"network.6.weight": [
[
-0.198919,
0.189116,
0.00834,
0.249198,
0.269863,
0.283044,
-0.286131,
-0.048913
],
[
-0.020965,
0.234255,
-0.353259,
0.144067,
-0.061274,
0.169769,
-0.184536,
-0.216735
],
[
0.317142,
0.182032,
0.589724,
0.463473,
0.050838,
0.869024,
-0.326745,
0.316813
],
[
-0.267253,
-0.054415,
-0.497363,
0.065838,
0.01388,
0.401584,
-0.091142,
-0.238396
],
[
0.159757,
0.013914,
0.650166,
-0.033122,
-0.12651,
0.511938,
-0.426076,
0.789559
],
[
0.036908,
0.114666,
-0.217049,
-0.120711,
-0.167224,
0.208699,
-0.251878,
-0.332019
],
[
0.552419,
-0.20671,
0.601837,
-0.090438,
-0.064472,
0.406912,
-0.317873,
0.675
],
[
-0.086703,
-0.006697,
-0.060965,
0.079886,
-0.203472,
-0.397344,
0.689696,
0.018006
]
],
"network.6.bias": [
-0.242388,
-0.197396,
0.019618,
0.117503,
0.296871,
-0.19959,
0.5073,
0.146358
],
"network.8.weight": [
[
-0.404476,
0.081498,
0.604959,
-0.10901,
0.687896,
0.131854,
0.535899,
-0.159888
],
[
-0.004761,
0.340257,
0.51314,
-0.13325,
0.611188,
0.237585,
0.599384,
-0.387454
],
[
0.316094,
-0.306398,
-0.213902,
-0.272248,
-0.016868,
0.286093,
-0.239087,
0.340796
],
[
-0.039796,
-0.27664,
0.050163,
-0.303247,
-0.351489,
0.299391,
-0.264714,
-0.34363
],
[
0.092366,
0.040987,
-0.317252,
0.076024,
0.056689,
-0.186754,
-0.596717,
0.709301
],
[
-0.043603,
0.213391,
-0.148799,
0.023875,
-0.080774,
-0.180717,
0.558857,
0.076617
],
[
-0.234709,
0.29266,
0.213573,
-0.181494,
-0.068125,
-0.177662,
0.180862,
0.060476
],
[
0.252374,
-0.224632,
0.153081,
0.069098,
-0.198631,
-0.30378,
-0.204866,
0.412867
]
],
"network.8.bias": [
0.526378,
0.400971,
0.23469,
-0.021799,
0.206416,
0.149207,
0.13541,
0.242314
],
"network.10.weight": [
[
0.04282,
0.507103,
-0.489137,
-0.137697,
-0.317948,
-0.328721,
-0.437974,
-0.494849
],
[
-0.824575,
-0.758382,
0.285225,
0.015572,
0.600799,
0.156506,
0.393958,
0.397637
],
[
0.343337,
0.718864,
-0.009881,
-0.047269,
-0.434034,
0.38666,
0.361526,
-0.305664
],
[
0.747234,
0.60891,
-0.159455,
0.171798,
-0.308262,
-0.474506,
-0.143673,
-0.066018
],
[
0.170037,
-0.282807,
0.050458,
0.341707,
-0.146471,
0.153988,
-0.151922,
0.100865
],
[
-0.038398,
-0.239801,
-0.059448,
0.253666,
-0.107913,
-0.069285,
0.289029,
-0.338299
],
[
-0.172213,
-0.069052,
-0.320333,
0.258069,
-0.098832,
0.149351,
-0.253339,
-0.34549
],
[
-0.673727,
-0.011986,
-0.003642,
-0.007366,
0.609922,
0.072129,
0.342205,
0.441785
]
],
"network.10.bias": [
0.374342,
0.175768,
0.381225,
0.412715,
-0.29183,
-0.622215,
-0.304338,
0.298499
],
"network.12.weight": [
[
-0.496183,
0.388304,
-0.538868,
-0.654806,
0.09697,
-0.010081,
0.252938,
0.244775
]
],
"network.12.bias": [
-0.261173
]
}
## Activation Signature
### 0
mean: [-1.028208, -1.578978, 2.318805, 1.236856, -1.576337, -0.631101, 0.432712, 0.555378]
std: [1.975761, 0.923683, 1.900966, 2.192478, 2.157034, 2.177463, 2.015826, 1.601387]
### 2
mean: [-0.214167, 1.273706, 1.899954, -0.377768, -0.765640, -1.036367, 0.032790, -2.253067]
std: [1.868116, 1.404590, 2.195429, 1.111173, 0.620315, 0.713796, 1.112615, 1.132964]
### 4
mean: [1.393777, -0.983424, 1.781502, -0.373851, -1.359562, 0.524682, 0.063365, 2.681238]
std: [1.691497, 0.408552, 2.133758, 0.158420, 0.853109, 0.932321, 1.475539, 2.162709]
### 6
mean: [-0.620327, -1.456922, 2.824868, -1.612686, 3.952641, -1.444438, 4.326758, 0.087583]
std: [0.267747, 1.037413, 3.187500, 1.654218, 3.864513, 0.873402, 4.030807, 0.996125]
### 8
mean: [7.271826, 6.751927, -1.341323, -2.571192, -2.766790, 1.850561, 1.292486, -0.815053]
std: [6.733321, 6.497111, 1.828539, 2.135725, 3.457049, 1.449702, 1.109712, 1.286779]
### 10
mean: [2.744888, -9.942060, 8.797823, 8.816937, -0.892816, -2.363634, -2.190863, -3.913485]
std: [2.849319, 10.027461, 8.061817, 8.220969, 0.621569, 1.514423, 1.573909, 4.344307]
### 12
mean: [-12.136356]
std: [11.130861]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
no_repeats
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.0282080173492432, "std": 1.97576105594635}, "1": {"mean": -1.578978180885315, "std": 0.9236829280853271}, "2": {"mean": 2.3188045024871826, "std": 1.9009661674499512}, "3": {"mean": 1.2368559837341309, "std": 2.1924784183502197}, "4": {"mean": -1.5763368606567383, "std": 2.157034397125244}, "5": {"mean": -0.6311014890670776, "std": 2.1774632930755615}, "6": {"mean": 0.4327123463153839, "std": 2.0158259868621826}, "7": {"mean": 0.5553778409957886, "std": 1.6013869047164917}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.21416714787483215, "std": 1.8681163787841797}, "1": {"mean": 1.2737058401107788, "std": 1.4045896530151367}, "2": {"mean": 1.8999536037445068, "std": 2.1954288482666016}, "3": {"mean": -0.37776848673820496, "std": 1.1111729145050049}, "4": {"mean": -0.7656404972076416, "std": 0.6203146576881409}, "5": {"mean": -1.036367416381836, "std": 0.7137956023216248}, "6": {"mean": 0.032789915800094604, "std": 1.1126147508621216}, "7": {"mean": -2.2530665397644043, "std": 1.13296377658844}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 1.393776774406433, "std": 1.6914974451065063}, "1": {"mean": -0.9834239482879639, "std": 0.40855151414871216}, "2": {"mean": 1.7815024852752686, "std": 2.133758068084717}, "3": {"mean": -0.3738509714603424, "std": 0.15842029452323914}, "4": {"mean": -1.3595623970031738, "std": 0.8531091809272766}, "5": {"mean": 0.524681568145752, "std": 0.9323208332061768}, "6": {"mean": 0.06336522847414017, "std": 1.4755386114120483}, "7": {"mean": 2.6812376976013184, "std": 2.1627089977264404}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.6203272342681885, "std": 0.2677466571331024}, "1": {"mean": -1.4569222927093506, "std": 1.0374128818511963}, "2": {"mean": 2.8248677253723145, "std": 3.187500238418579}, "3": {"mean": -1.6126861572265625, "std": 1.6542184352874756}, "4": {"mean": 3.952641487121582, "std": 3.8645129203796387}, "5": {"mean": -1.4444375038146973, "std": 0.8734016418457031}, "6": {"mean": 4.32675838470459, "std": 4.0308074951171875}, "7": {"mean": 0.08758306503295898, "std": 0.9961245656013489}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 7.271826267242432, "std": 6.733321189880371}, "1": {"mean": 6.751926898956299, "std": 6.497110843658447}, "2": {"mean": -1.3413225412368774, "std": 1.8285390138626099}, "3": {"mean": -2.5711920261383057, "std": 2.1357245445251465}, "4": {"mean": -2.766789674758911, "std": 3.4570488929748535}, "5": {"mean": 1.8505605459213257, "std": 1.4497020244598389}, "6": {"mean": 1.292486310005188, "std": 1.1097123622894287}, "7": {"mean": -0.8150534629821777, "std": 1.2867789268493652}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": 2.7448880672454834, "std": 2.849318742752075}, "1": {"mean": -9.942060470581055, "std": 10.027461051940918}, "2": {"mean": 8.797822952270508, "std": 8.061817169189453}, "3": {"mean": 8.816937446594238, "std": 8.220969200134277}, "4": {"mean": -0.8928155899047852, "std": 0.6215689778327942}, "5": {"mean": -2.3636343479156494, "std": 1.5144232511520386}, "6": {"mean": -2.1908631324768066, "std": 1.573909044265747}, "7": {"mean": -3.913485288619995, "std": 4.3443074226379395}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -12.136356353759766, "std": 11.130861282348633}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.688668, 0.186801, -0.570155, -0.033674, -0.651347], [-0.070155, -0.296438, -0.05982, -0.066008, -0.290094], [-0.450714, 0.251016, 0.582487, -0.028885, 0.732336], [0.622606, 0.371871, 0.257956, -0.280204, -0.747301], [0.042009, 0.194713, -0.533164, 0.129308, -0.930439], [0.541737, -0.86864, 0.454182, -0.362061, -0.118582], [0.840696, 0.006576, -0.35385, 0.330622, -0.717408], [0.322608, -0.532125, -0.036308, -0.089079, 0.606855]], "network.0.bias": [-0.15908, -0.327208, 0.364263, 0.768137, 0.000541, 0.251528, 0.289487, 0.641975], "network.2.weight": [[-0.160147, 0.206342, 0.58399, -0.412795, 0.108054, -0.477953, -0.291728, -0.767862], [0.35562, -0.343743, -0.067234, 0.477687, 0.151602, 0.483631, 0.132714, 0.000881], [0.493912, 0.194248, -0.319403, 0.560612, 0.515356, 0.349381, 0.489256, 0.239758], [-0.103124, -0.137564, 0.356718, -0.314897, 0.188524, -0.400103, 0.157925, -0.625852], [-0.256529, 0.204227, -0.32926, 0.070566, -0.52351, 0.044988, -0.033262, -0.13065], [-0.111184, -0.31412, -0.188134, -0.203571, 0.090853, -0.333245, -0.154048, 0.12206], [-0.162174, -0.06376, 0.418355, -0.126358, 0.338754, 0.075835, -0.287578, -0.572576], [0.093981, 0.193527, -0.502802, -0.40804, -0.147435, 0.013082, -0.359133, -0.063094]], "network.2.bias": [0.326585, 0.154568, 0.650883, -0.102188, 0.181359, -0.026824, 0.024173, -0.0276], "network.4.weight": [[-0.155294, 0.51883, 0.421213, 0.143105, -0.087047, 0.030563, -0.221382, 0.195086], [-0.199731, 0.015115, -0.232097, 0.025164, -0.091817, -0.130878, -0.244923, -0.148356], [-0.155585, 0.419377, 0.659027, -0.504517, 0.315447, 0.213172, -0.138787, -0.144865], [-0.112092, 0.32952, -0.264814, 0.096127, -0.320058, -0.314027, -0.197992, 0.013833], [-0.146718, -0.39142, -0.188915, -0.052719, 0.198253, 0.136368, -0.545484, 0.395407], [-0.531848, 0.43783, -0.012102, 0.003725, 0.009807, 0.016, -0.281076, -0.033706], [0.739484, -0.368589, -0.178331, 0.570549, 0.488028, -0.01328, 0.304513, -0.198414], [0.030582, 0.210968, 0.895869, -0.506172, 0.317413, -0.002714, 0.348988, -0.145547]], "network.4.bias": [0.052354, -0.335031, 0.201355, -0.15171, -0.151362, 0.400161, 0.2235, 0.589509], "network.6.weight": [[-0.198919, 0.189116, 0.00834, 0.249198, 0.269863, 0.283044, -0.286131, -0.048913], [-0.020965, 0.234255, -0.353259, 0.144067, -0.061274, 0.169769, -0.184536, -0.216735], [0.317142, 0.182032, 0.589724, 0.463473, 0.050838, 0.869024, -0.326745, 0.316813], [-0.267253, -0.054415, -0.497363, 0.065838, 0.01388, 0.401584, -0.091142, -0.238396], [0.159757, 0.013914, 0.650166, -0.033122, -0.12651, 0.511938, -0.426076, 0.789559], [0.036908, 0.114666, -0.217049, -0.120711, -0.167224, 0.208699, -0.251878, -0.332019], [0.552419, -0.20671, 0.601837, -0.090438, -0.064472, 0.406912, -0.317873, 0.675], [-0.086703, -0.006697, -0.060965, 0.079886, -0.203472, -0.397344, 0.689696, 0.018006]], "network.6.bias": [-0.242388, -0.197396, 0.019618, 0.117503, 0.296871, -0.19959, 0.5073, 0.146358], "network.8.weight": [[-0.404476, 0.081498, 0.604959, -0.10901, 0.687896, 0.131854, 0.535899, -0.159888], [-0.004761, 0.340257, 0.51314, -0.13325, 0.611188, 0.237585, 0.599384, -0.387454], [0.316094, -0.306398, -0.213902, -0.272248, -0.016868, 0.286093, -0.239087, 0.340796], [-0.039796, -0.27664, 0.050163, -0.303247, -0.351489, 0.299391, -0.264714, -0.34363], [0.092366, 0.040987, -0.317252, 0.076024, 0.056689, -0.186754, -0.596717, 0.709301], [-0.043603, 0.213391, -0.148799, 0.023875, -0.080774, -0.180717, 0.558857, 0.076617], [-0.234709, 0.29266, 0.213573, -0.181494, -0.068125, -0.177662, 0.180862, 0.060476], [0.252374, -0.224632, 0.153081, 0.069098, -0.198631, -0.30378, -0.204866, 0.412867]], "network.8.bias": [0.526378, 0.400971, 0.23469, -0.021799, 0.206416, 0.149207, 0.13541, 0.242314], "network.10.weight": [[0.04282, 0.507103, -0.489137, -0.137697, -0.317948, -0.328721, -0.437974, -0.494849], [-0.824575, -0.758382, 0.285225, 0.015572, 0.600799, 0.156506, 0.393958, 0.397637], [0.343337, 0.718864, -0.009881, -0.047269, -0.434034, 0.38666, 0.361526, -0.305664], [0.747234, 0.60891, -0.159455, 0.171798, -0.308262, -0.474506, -0.143673, -0.066018], [0.170037, -0.282807, 0.050458, 0.341707, -0.146471, 0.153988, -0.151922, 0.100865], [-0.038398, -0.239801, -0.059448, 0.253666, -0.107913, -0.069285, 0.289029, -0.338299], [-0.172213, -0.069052, -0.320333, 0.258069, -0.098832, 0.149351, -0.253339, -0.34549], [-0.673727, -0.011986, -0.003642, -0.007366, 0.609922, 0.072129, 0.342205, 0.441785]], "network.10.bias": [0.374342, 0.175768, 0.381225, 0.412715, -0.29183, -0.622215, -0.304338, 0.298499], "network.12.weight": [[-0.496183, 0.388304, -0.538868, -0.654806, 0.09697, -0.010081, 0.252938, 0.244775]], "network.12.bias": [-0.261173]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6784787476062775, "train_acc": 0.585, "val_loss": 0.7232218384742737, "val_acc": 0.44}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6760828495025635, "train_acc": 0.585, "val_loss": 0.72896409034729, "val_acc": 0.44}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6713890731334686, "train_acc": 0.585, "val_loss": 0.7269176244735718, "val_acc": 0.44}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.658789724111557, "train_acc": 0.585, "val_loss": 0.7127646803855896, "val_acc": 0.44}, {"stage": "degraded", "epoch": 4, "global_epoch": 4, "train_loss": 0.5968922674655914, "train_acc": 0.585, "val_loss": 0.6472501158714294, "val_acc": 0.7}, {"stage": "improved", "epoch": 0, "global_epoch": 5, "train_loss": 0.5493353307247162, "train_acc": 0.75, "val_loss": 0.5397741198539734, "val_acc": 0.74}, {"stage": "improved", "epoch": 1, "global_epoch": 6, "train_loss": 0.44625474512577057, "train_acc": 0.77, "val_loss": 0.5640757083892822, "val_acc": 0.74}, {"stage": "improved", "epoch": 2, "global_epoch": 7, "train_loss": 0.46926894783973694, "train_acc": 0.77, "val_loss": 0.4999631643295288, "val_acc": 0.74}, {"stage": "improved", "epoch": 3, "global_epoch": 8, "train_loss": 0.4327402859926224, "train_acc": 0.775, "val_loss": 0.4943987727165222, "val_acc": 0.8}, {"stage": "improved", "epoch": 4, "global_epoch": 9, "train_loss": 0.40808746218681335, "train_acc": 0.815, "val_loss": 0.4315985441207886, "val_acc": 0.8}, {"stage": "improved", "epoch": 5, "global_epoch": 10, "train_loss": 0.35598231852054596, "train_acc": 0.83, "val_loss": 0.37750038504600525, "val_acc": 0.82}, {"stage": "improved", "epoch": 6, "global_epoch": 11, "train_loss": 0.4037778675556183, "train_acc": 0.845, "val_loss": 0.4019967019557953, "val_acc": 0.8}, {"stage": "improved", "epoch": 7, "global_epoch": 12, "train_loss": 0.32534629851579666, "train_acc": 0.85, "val_loss": 0.37149661779403687, "val_acc": 0.82}, {"stage": "improved", "epoch": 8, "global_epoch": 13, "train_loss": 0.35212984681129456, "train_acc": 0.86, "val_loss": 0.3564305603504181, "val_acc": 0.82}, {"stage": "improved", "epoch": 9, "global_epoch": 14, "train_loss": 0.30723531544208527, "train_acc": 0.845, "val_loss": 0.3163755536079407, "val_acc": 0.88}], "summary": {"total_epochs": 15, "degraded_epochs": 5, "improved_epochs": 10, "patterns": ["no_repeats"], "degraded_stage": {"initial_val_loss": 0.7232218384742737, "final_val_loss": 0.6472501158714294, "initial_val_acc": 0.44, "final_val_acc": 0.7, "best_val_acc": 0.7}, "improved_stage": {"initial_val_loss": 0.5397741198539734, "final_val_loss": 0.3163755536079407, "initial_val_acc": 0.74, "final_val_acc": 0.88, "best_val_acc": 0.88, "best_epoch": 14}, "improvement": 0.18000000000000005, "first_improvement_epoch": 4}}
|
14
|
{"target_pattern": "increasing_pairs", "degraded_accuracy": 0.38, "improved_accuracy": 0.9, "improvement": 0.52, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 1822, "learning_rate": 0.04873785800456769, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "increasing_pairs", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["increasing_pairs"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.03427,
0.254661,
-0.309303,
0.132472,
-0.301122
],
[
0.695961,
0.499535,
0.379718,
-0.219172,
-0.104522
],
[
0.685806,
0.263125,
0.095531,
0.166593,
-0.485912
],
[
0.147266,
-0.160844,
0.266516,
0.375742,
-0.081611
],
[
0.506531,
0.243322,
-0.184177,
0.179929,
-0.35675
],
[
-0.293701,
-0.559604,
-0.153599,
0.187748,
-0.509362
],
[
0.624731,
0.486528,
0.39323,
-0.023678,
-0.388118
],
[
-0.174981,
0.374477,
-0.156257,
0.069318,
-0.161376
]
],
"network.0.bias": [
0.360241,
0.194727,
0.524528,
0.053676,
0.271863,
-0.066478,
-0.102405,
0.181772
],
"network.2.weight": [
[
0.464119,
0.555003,
0.357585,
-0.010542,
0.164302,
0.000758,
0.448923,
-0.011013
],
[
0.140184,
0.12645,
-0.429321,
0.554615,
-0.322337,
-0.411152,
-0.301973,
0.23613
],
[
0.336454,
0.072652,
0.286157,
-0.293543,
0.123799,
0.133701,
0.10615,
0.158463
],
[
0.151947,
0.080862,
0.048378,
0.050153,
0.471383,
-0.138846,
0.094669,
-0.085556
],
[
0.419275,
-0.008559,
-0.365298,
-0.347234,
-0.034399,
0.315162,
0.289365,
0.11448
],
[
0.312462,
0.272727,
0.254494,
0.154216,
0.398563,
0.198322,
0.201393,
0.055705
],
[
0.307768,
-0.175849,
-0.354241,
-0.168731,
0.270973,
-0.174027,
-0.089531,
-0.366647
],
[
-0.095963,
0.11193,
-0.094642,
-0.386215,
-0.372112,
-0.020525,
-0.099473,
-0.153999
]
],
"network.2.bias": [
0.084542,
0.345603,
0.13798,
0.181469,
-0.29066,
0.115681,
-0.051623,
-0.158041
],
"network.4.weight": [
[
0.377324,
-0.086906,
0.288147,
-0.07013,
0.014101,
0.0285,
0.162883,
-0.00771
],
[
0.216963,
-0.135615,
0.136641,
-0.075547,
0.548637,
-0.110976,
-0.306572,
0.212836
],
[
0.023662,
-0.536783,
-0.167246,
-0.07285,
-0.037807,
0.186161,
0.126408,
-0.085256
],
[
0.396457,
-0.198668,
0.214248,
0.16283,
0.083951,
0.249378,
-0.190292,
-0.09054
],
[
0.002995,
-0.125959,
-0.235027,
-0.069661,
0.110882,
-0.211163,
-0.189908,
0.107078
],
[
-0.295608,
-0.061142,
-0.198986,
0.267072,
0.103152,
0.206871,
-0.032812,
0.298991
],
[
0.361123,
-0.261462,
0.061711,
0.385118,
0.545958,
0.364879,
-0.040525,
-0.007732
],
[
0.191356,
-0.067305,
-0.080232,
-0.269716,
0.083409,
-0.14353,
0.16178,
0.414132
]
],
"network.4.bias": [
-0.154275,
-0.037132,
0.24799,
-0.198083,
-0.320159,
-0.301079,
-0.143562,
0.01547
],
"network.6.weight": [
[
0.327731,
0.50809,
0.019545,
0.155522,
-0.040008,
0.254768,
0.30546,
0.313227
],
[
0.392508,
0.177243,
0.410761,
0.234215,
0.156626,
-0.295994,
0.186534,
0.089047
],
[
-0.122725,
0.14566,
-0.086069,
-0.09444,
-0.319536,
-0.140988,
-0.174259,
0.249296
],
[
0.030627,
-0.100947,
0.193733,
-0.2805,
0.046072,
0.015808,
0.135734,
-0.029867
],
[
0.002444,
0.119631,
-0.355959,
0.207131,
0.114407,
-0.122948,
-0.309305,
-0.341785
],
[
-0.154362,
-0.051687,
-0.237916,
-0.120052,
-0.114186,
-0.056962,
-0.406485,
0.11199
],
[
0.045069,
0.334316,
-0.128714,
0.494978,
0.180179,
0.02095,
0.144972,
0.013131
],
[
-0.081586,
-0.074897,
0.133121,
-0.122524,
-0.353851,
0.141409,
-0.471301,
0.074046
]
],
"network.6.bias": [
0.062348,
0.313187,
-0.112411,
-0.309409,
-0.121214,
-0.059367,
-0.241459,
-0.050572
],
"network.8.weight": [
[
-0.36926,
-0.216622,
-0.218758,
-0.026657,
0.084004,
0.008249,
-0.502324,
-0.013735
]
],
"network.8.bias": [
0.205056
]
}
## Activation Signature
### 0
mean: [0.133532, 2.165351, 1.814957, 1.205675, 0.906729, -1.998148, 1.853559, 0.269615]
std: [1.003019, 2.259251, 1.798710, 0.896113, 1.348467, 1.780487, 2.146909, 0.867654]
### 2
mean: [3.246231, -0.294916, 1.056350, 1.229664, -0.641367, 2.383811, -1.232694, -1.264181]
std: [2.869597, 1.196782, 1.002406, 0.980134, 0.443731, 1.914166, 0.938818, 0.830111]
### 4
mean: [1.339757, 0.430786, 0.362269, 2.064414, -1.182032, -0.666646, 2.380642, -0.139058]
std: [1.350575, 0.504271, 0.328964, 2.000773, 0.660415, 0.388345, 2.210224, 0.141086]
### 6
mean: [1.798258, 2.014851, -0.855605, -0.497947, -0.513947, -1.602131, 1.295504, -1.523504]
std: [1.666856, 1.596720, 0.688566, 0.214123, 0.309919, 1.428018, 1.486165, 1.387934]
### 8
mean: [-1.559641]
std: [1.694420]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
increasing_pairs
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.03427,
0.254661,
-0.309303,
0.132472,
-0.301122
],
[
0.695961,
0.499535,
0.379718,
-0.219172,
-0.104522
],
[
0.685806,
0.263125,
0.095531,
0.166593,
-0.485912
],
[
0.147266,
-0.160844,
0.266516,
0.375742,
-0.081611
],
[
0.506531,
0.243322,
-0.184177,
0.179929,
-0.35675
],
[
-0.293701,
-0.559604,
-0.153599,
0.187748,
-0.509362
],
[
0.624731,
0.486528,
0.39323,
-0.023678,
-0.388118
],
[
-0.174981,
0.374477,
-0.156257,
0.069318,
-0.161376
]
],
"network.0.bias": [
0.360241,
0.194727,
0.524528,
0.053676,
0.271863,
-0.066478,
-0.102405,
0.181772
],
"network.2.weight": [
[
0.464119,
0.555003,
0.357585,
-0.010542,
0.164302,
0.000758,
0.448923,
-0.011013
],
[
0.140184,
0.12645,
-0.429321,
0.554615,
-0.322337,
-0.411152,
-0.301973,
0.23613
],
[
0.336454,
0.072652,
0.286157,
-0.293543,
0.123799,
0.133701,
0.10615,
0.158463
],
[
0.151947,
0.080862,
0.048378,
0.050153,
0.471383,
-0.138846,
0.094669,
-0.085556
],
[
0.419275,
-0.008559,
-0.365298,
-0.347234,
-0.034399,
0.315162,
0.289365,
0.11448
],
[
0.312462,
0.272727,
0.254494,
0.154216,
0.398563,
0.198322,
0.201393,
0.055705
],
[
0.307768,
-0.175849,
-0.354241,
-0.168731,
0.270973,
-0.174027,
-0.089531,
-0.366647
],
[
-0.095963,
0.11193,
-0.094642,
-0.386215,
-0.372112,
-0.020525,
-0.099473,
-0.153999
]
],
"network.2.bias": [
0.084542,
0.345603,
0.13798,
0.181469,
-0.29066,
0.115681,
-0.051623,
-0.158041
],
"network.4.weight": [
[
0.377324,
-0.086906,
0.288147,
-0.07013,
0.014101,
0.0285,
0.162883,
-0.00771
],
[
0.216963,
-0.135615,
0.136641,
-0.075547,
0.548637,
-0.110976,
-0.306572,
0.212836
],
[
0.023662,
-0.536783,
-0.167246,
-0.07285,
-0.037807,
0.186161,
0.126408,
-0.085256
],
[
0.396457,
-0.198668,
0.214248,
0.16283,
0.083951,
0.249378,
-0.190292,
-0.09054
],
[
0.002995,
-0.125959,
-0.235027,
-0.069661,
0.110882,
-0.211163,
-0.189908,
0.107078
],
[
-0.295608,
-0.061142,
-0.198986,
0.267072,
0.103152,
0.206871,
-0.032812,
0.298991
],
[
0.361123,
-0.261462,
0.061711,
0.385118,
0.545958,
0.364879,
-0.040525,
-0.007732
],
[
0.191356,
-0.067305,
-0.080232,
-0.269716,
0.083409,
-0.14353,
0.16178,
0.414132
]
],
"network.4.bias": [
-0.154275,
-0.037132,
0.24799,
-0.198083,
-0.320159,
-0.301079,
-0.143562,
0.01547
],
"network.6.weight": [
[
0.327731,
0.50809,
0.019545,
0.155522,
-0.040008,
0.254768,
0.30546,
0.313227
],
[
0.392508,
0.177243,
0.410761,
0.234215,
0.156626,
-0.295994,
0.186534,
0.089047
],
[
-0.122725,
0.14566,
-0.086069,
-0.09444,
-0.319536,
-0.140988,
-0.174259,
0.249296
],
[
0.030627,
-0.100947,
0.193733,
-0.2805,
0.046072,
0.015808,
0.135734,
-0.029867
],
[
0.002444,
0.119631,
-0.355959,
0.207131,
0.114407,
-0.122948,
-0.309305,
-0.341785
],
[
-0.154362,
-0.051687,
-0.237916,
-0.120052,
-0.114186,
-0.056962,
-0.406485,
0.11199
],
[
0.045069,
0.334316,
-0.128714,
0.494978,
0.180179,
0.02095,
0.144972,
0.013131
],
[
-0.081586,
-0.074897,
0.133121,
-0.122524,
-0.353851,
0.141409,
-0.471301,
0.074046
]
],
"network.6.bias": [
0.062348,
0.313187,
-0.112411,
-0.309409,
-0.121214,
-0.059367,
-0.241459,
-0.050572
],
"network.8.weight": [
[
-0.36926,
-0.216622,
-0.218758,
-0.026657,
0.084004,
0.008249,
-0.502324,
-0.013735
]
],
"network.8.bias": [
0.205056
]
}
## Activation Signature
### 0
mean: [0.133532, 2.165351, 1.814957, 1.205675, 0.906729, -1.998148, 1.853559, 0.269615]
std: [1.003019, 2.259251, 1.798710, 0.896113, 1.348467, 1.780487, 2.146909, 0.867654]
### 2
mean: [3.246231, -0.294916, 1.056350, 1.229664, -0.641367, 2.383811, -1.232694, -1.264181]
std: [2.869597, 1.196782, 1.002406, 0.980134, 0.443731, 1.914166, 0.938818, 0.830111]
### 4
mean: [1.339757, 0.430786, 0.362269, 2.064414, -1.182032, -0.666646, 2.380642, -0.139058]
std: [1.350575, 0.504271, 0.328964, 2.000773, 0.660415, 0.388345, 2.210224, 0.141086]
### 6
mean: [1.798258, 2.014851, -0.855605, -0.497947, -0.513947, -1.602131, 1.295504, -1.523504]
std: [1.666856, 1.596720, 0.688566, 0.214123, 0.309919, 1.428018, 1.486165, 1.387934]
### 8
mean: [-1.559641]
std: [1.694420]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
increasing_pairs
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.1335316151380539, "std": 1.0030187368392944}, "1": {"mean": 2.1653506755828857, "std": 2.2592508792877197}, "2": {"mean": 1.814956784248352, "std": 1.7987096309661865}, "3": {"mean": 1.2056745290756226, "std": 0.8961132168769836}, "4": {"mean": 0.9067291021347046, "std": 1.3484673500061035}, "5": {"mean": -1.9981476068496704, "std": 1.780486822128296}, "6": {"mean": 1.8535586595535278, "std": 2.1469085216522217}, "7": {"mean": 0.2696149945259094, "std": 0.867654025554657}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 3.2462313175201416, "std": 2.8695971965789795}, "1": {"mean": -0.2949158549308777, "std": 1.196781873703003}, "2": {"mean": 1.0563501119613647, "std": 1.0024060010910034}, "3": {"mean": 1.2296642065048218, "std": 0.9801344871520996}, "4": {"mean": -0.641366720199585, "std": 0.4437314569950104}, "5": {"mean": 2.3838109970092773, "std": 1.9141660928726196}, "6": {"mean": -1.2326940298080444, "std": 0.9388176202774048}, "7": {"mean": -1.2641808986663818, "std": 0.8301106691360474}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 1.3397570848464966, "std": 1.3505746126174927}, "1": {"mean": 0.4307864010334015, "std": 0.5042713284492493}, "2": {"mean": 0.3622686564922333, "std": 0.32896432280540466}, "3": {"mean": 2.064413547515869, "std": 2.0007729530334473}, "4": {"mean": -1.1820316314697266, "std": 0.6604152321815491}, "5": {"mean": -0.6666458249092102, "std": 0.3883446753025055}, "6": {"mean": 2.3806424140930176, "std": 2.210223913192749}, "7": {"mean": -0.1390584111213684, "std": 0.1410863697528839}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 1.798257827758789, "std": 1.6668556928634644}, "1": {"mean": 2.0148510932922363, "std": 1.596719741821289}, "2": {"mean": -0.8556053042411804, "std": 0.6885657906532288}, "3": {"mean": -0.49794715642929077, "std": 0.21412283182144165}, "4": {"mean": -0.5139466524124146, "std": 0.3099192976951599}, "5": {"mean": -1.6021310091018677, "std": 1.4280178546905518}, "6": {"mean": 1.295503854751587, "std": 1.4861648082733154}, "7": {"mean": -1.5235036611557007, "std": 1.3879343271255493}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.559640884399414, "std": 1.6944196224212646}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.03427, 0.254661, -0.309303, 0.132472, -0.301122], [0.695961, 0.499535, 0.379718, -0.219172, -0.104522], [0.685806, 0.263125, 0.095531, 0.166593, -0.485912], [0.147266, -0.160844, 0.266516, 0.375742, -0.081611], [0.506531, 0.243322, -0.184177, 0.179929, -0.35675], [-0.293701, -0.559604, -0.153599, 0.187748, -0.509362], [0.624731, 0.486528, 0.39323, -0.023678, -0.388118], [-0.174981, 0.374477, -0.156257, 0.069318, -0.161376]], "network.0.bias": [0.360241, 0.194727, 0.524528, 0.053676, 0.271863, -0.066478, -0.102405, 0.181772], "network.2.weight": [[0.464119, 0.555003, 0.357585, -0.010542, 0.164302, 0.000758, 0.448923, -0.011013], [0.140184, 0.12645, -0.429321, 0.554615, -0.322337, -0.411152, -0.301973, 0.23613], [0.336454, 0.072652, 0.286157, -0.293543, 0.123799, 0.133701, 0.10615, 0.158463], [0.151947, 0.080862, 0.048378, 0.050153, 0.471383, -0.138846, 0.094669, -0.085556], [0.419275, -0.008559, -0.365298, -0.347234, -0.034399, 0.315162, 0.289365, 0.11448], [0.312462, 0.272727, 0.254494, 0.154216, 0.398563, 0.198322, 0.201393, 0.055705], [0.307768, -0.175849, -0.354241, -0.168731, 0.270973, -0.174027, -0.089531, -0.366647], [-0.095963, 0.11193, -0.094642, -0.386215, -0.372112, -0.020525, -0.099473, -0.153999]], "network.2.bias": [0.084542, 0.345603, 0.13798, 0.181469, -0.29066, 0.115681, -0.051623, -0.158041], "network.4.weight": [[0.377324, -0.086906, 0.288147, -0.07013, 0.014101, 0.0285, 0.162883, -0.00771], [0.216963, -0.135615, 0.136641, -0.075547, 0.548637, -0.110976, -0.306572, 0.212836], [0.023662, -0.536783, -0.167246, -0.07285, -0.037807, 0.186161, 0.126408, -0.085256], [0.396457, -0.198668, 0.214248, 0.16283, 0.083951, 0.249378, -0.190292, -0.09054], [0.002995, -0.125959, -0.235027, -0.069661, 0.110882, -0.211163, -0.189908, 0.107078], [-0.295608, -0.061142, -0.198986, 0.267072, 0.103152, 0.206871, -0.032812, 0.298991], [0.361123, -0.261462, 0.061711, 0.385118, 0.545958, 0.364879, -0.040525, -0.007732], [0.191356, -0.067305, -0.080232, -0.269716, 0.083409, -0.14353, 0.16178, 0.414132]], "network.4.bias": [-0.154275, -0.037132, 0.24799, -0.198083, -0.320159, -0.301079, -0.143562, 0.01547], "network.6.weight": [[0.327731, 0.50809, 0.019545, 0.155522, -0.040008, 0.254768, 0.30546, 0.313227], [0.392508, 0.177243, 0.410761, 0.234215, 0.156626, -0.295994, 0.186534, 0.089047], [-0.122725, 0.14566, -0.086069, -0.09444, -0.319536, -0.140988, -0.174259, 0.249296], [0.030627, -0.100947, 0.193733, -0.2805, 0.046072, 0.015808, 0.135734, -0.029867], [0.002444, 0.119631, -0.355959, 0.207131, 0.114407, -0.122948, -0.309305, -0.341785], [-0.154362, -0.051687, -0.237916, -0.120052, -0.114186, -0.056962, -0.406485, 0.11199], [0.045069, 0.334316, -0.128714, 0.494978, 0.180179, 0.02095, 0.144972, 0.013131], [-0.081586, -0.074897, 0.133121, -0.122524, -0.353851, 0.141409, -0.471301, 0.074046]], "network.6.bias": [0.062348, 0.313187, -0.112411, -0.309409, -0.121214, -0.059367, -0.241459, -0.050572], "network.8.weight": [[-0.36926, -0.216622, -0.218758, -0.026657, 0.084004, 0.008249, -0.502324, -0.013735]], "network.8.bias": [0.205056]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6850249171257019, "train_acc": 0.535, "val_loss": 0.6631926894187927, "val_acc": 0.38}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.5785219073295593, "train_acc": 0.595, "val_loss": 0.6216185092926025, "val_acc": 0.38}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.5146430432796478, "train_acc": 0.67, "val_loss": 0.5280777215957642, "val_acc": 0.9}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.4426431357860565, "train_acc": 0.875, "val_loss": 0.5119740962982178, "val_acc": 0.72}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.4215056896209717, "train_acc": 0.785, "val_loss": 0.42959296703338623, "val_acc": 0.9}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.4106168895959854, "train_acc": 0.845, "val_loss": 0.4177243113517761, "val_acc": 0.88}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.3796081691980362, "train_acc": 0.84, "val_loss": 0.3847106695175171, "val_acc": 0.9}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.3422289192676544, "train_acc": 0.87, "val_loss": 0.367939293384552, "val_acc": 0.9}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.31923381984233856, "train_acc": 0.9, "val_loss": 0.3628294765949249, "val_acc": 0.9}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.31507591903209686, "train_acc": 0.895, "val_loss": 0.3491562604904175, "val_acc": 0.9}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.3002960979938507, "train_acc": 0.895, "val_loss": 0.33188700675964355, "val_acc": 0.9}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.3024744391441345, "train_acc": 0.9, "val_loss": 0.3256995379924774, "val_acc": 0.9}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["increasing_pairs"], "degraded_stage": {"initial_val_loss": 0.6631926894187927, "final_val_loss": 0.6216185092926025, "initial_val_acc": 0.38, "final_val_acc": 0.38, "best_val_acc": 0.38}, "improved_stage": {"initial_val_loss": 0.5280777215957642, "final_val_loss": 0.3256995379924774, "initial_val_acc": 0.9, "final_val_acc": 0.9, "best_val_acc": 0.9, "best_epoch": 2}, "improvement": 0.52, "first_improvement_epoch": 1}}
|
15
|
{"target_pattern": "sorted_ascending", "degraded_accuracy": 0.6, "improved_accuracy": 0.88, "improvement": 0.28, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 6709, "learning_rate": 0.05912918367929151, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_ascending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_ascending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.884267,
-0.208205,
-0.084779,
0.211018,
0.543483
],
[
0.080458,
0.895161,
0.375676,
-0.369734,
-0.141958
],
[
0.751684,
-0.462626,
-0.162537,
0.153153,
0.511408
],
[
-0.616191,
-0.623524,
-0.209838,
0.416046,
0.437106
],
[
1.010424,
0.227613,
0.298011,
-0.292982,
-0.057433
],
[
-0.45793,
-0.497994,
-0.2292,
0.552926,
0.484453
],
[
0.586622,
0.98632,
0.361719,
-0.024558,
-0.317672
],
[
0.576554,
0.868929,
-0.148509,
0.366823,
-0.492616
]
],
"network.0.bias": [
0.01207,
-0.055898,
0.565228,
0.098631,
-0.091249,
0.10603,
0.368698,
0.557823
],
"network.2.weight": [
[
-0.12298,
0.159205,
-0.385202,
-0.389619,
-0.414965,
-0.418727,
0.170375,
-0.413143
],
[
0.333331,
0.229628,
-0.11894,
-0.163551,
0.483145,
0.196858,
0.347422,
0.296496
],
[
-0.582098,
0.289499,
0.309888,
-0.23134,
0.041321,
0.062782,
0.335784,
0.725186
],
[
-0.71537,
0.02614,
-0.117887,
-0.533271,
0.233341,
-0.331485,
-0.069866,
0.123304
],
[
-0.297146,
0.126444,
0.430981,
-0.241571,
0.384342,
-0.457616,
0.461763,
0.396369
],
[
0.377994,
-0.201517,
-0.190416,
0.358944,
0.211093,
0.222712,
-0.299422,
-0.169458
],
[
0.082736,
0.41789,
0.149856,
-0.406215,
0.072122,
-0.314971,
0.422428,
0.737851
],
[
-0.137935,
0.609279,
0.538319,
-0.153631,
0.367608,
-0.587858,
0.193983,
-0.173398
]
],
"network.2.bias": [
-0.389157,
-0.218677,
0.355591,
-0.427036,
-0.142009,
0.220111,
0.092792,
0.132204
],
"network.4.weight": [
[
0.160442,
-0.667645,
-0.495327,
-0.104838,
-0.6295,
0.379691,
-0.550798,
-0.526627
],
[
-0.031836,
0.426638,
0.561198,
-0.00181,
0.181791,
-0.323334,
0.14868,
0.506618
],
[
0.107225,
0.254567,
0.250818,
-0.065005,
0.109505,
-0.565368,
0.572167,
0.525292
],
[
-0.275381,
-0.071808,
0.488721,
-0.152092,
0.133183,
-0.2186,
0.344506,
0.053412
],
[
-0.310963,
0.363825,
0.619465,
-0.432531,
0.658534,
-0.227626,
0.658911,
0.660271
],
[
-0.30743,
0.035336,
0.037346,
0.10327,
0.390352,
-0.278424,
0.295819,
0.696831
],
[
0.274612,
0.410858,
0.616534,
-0.433699,
0.452809,
-0.245824,
0.321396,
0.629888
],
[
-0.091912,
-0.565915,
-0.610762,
-0.762713,
-0.630704,
0.262504,
-0.50556,
-0.721301
]
],
"network.4.bias": [
0.660901,
0.411356,
0.079681,
0.116752,
0.116625,
0.324114,
0.143176,
0.808632
],
"network.6.weight": [
[
0.451379,
-0.324239,
-0.037355,
-0.941549,
-0.414234,
-0.143817,
-0.118661,
0.322498
],
[
-0.531572,
-0.008186,
0.295057,
0.627673,
0.406025,
-0.241041,
0.187523,
-0.107176
],
[
0.560588,
-0.371578,
-0.211444,
-0.223476,
-0.122567,
-0.091722,
-0.049366,
0.550834
],
[
-0.329912,
0.480352,
0.371632,
0.603313,
0.495597,
0.327018,
0.425249,
-0.412362
],
[
-0.367088,
-0.312847,
-0.210535,
0.018959,
-0.336912,
-0.448448,
0.003529,
0.07194
],
[
0.687545,
-0.527265,
0.024541,
-0.291059,
-0.420738,
-0.104432,
0.003346,
0.584944
],
[
-0.380494,
0.141094,
0.275808,
0.196025,
0.333952,
0.171109,
0.078006,
-0.221937
],
[
-0.064171,
-0.612749,
-0.277962,
-0.644033,
-0.323719,
0.08205,
0.3208,
-0.086161
]
],
"network.6.bias": [
0.564595,
-0.252215,
0.006863,
-0.052739,
0.015617,
0.256395,
-0.298876,
-0.085647
],
"network.8.weight": [
[
0.727817,
-0.742242,
0.004505,
-0.080493,
0.255715,
0.364046,
-0.73413,
-0.07103
],
[
0.01153,
-0.011429,
0.131153,
0.34151,
0.32882,
-0.233352,
0.12838,
0.406655
],
[
0.504262,
-0.730654,
-0.017194,
-0.427107,
-0.271088,
0.501936,
-0.707504,
-0.43853
],
[
-0.301098,
0.001369,
-0.390173,
0.580283,
0.114836,
-0.079997,
0.233588,
-0.28008
],
[
0.839497,
-0.314523,
0.639131,
-0.29879,
-0.145503,
0.16169,
-0.410782,
0.134225
],
[
-0.583781,
0.323504,
-0.367379,
0.561267,
0.186871,
-0.245,
0.232523,
-0.125561
],
[
-0.350124,
0.450341,
-0.338578,
0.562622,
0.130654,
-0.15929,
0.587091,
0.100226
],
[
-0.602727,
0.750379,
-0.318099,
0.050578,
-0.324114,
-0.122592,
0.618681,
0.103213
]
],
"network.8.bias": [
0.078719,
-0.278176,
0.453421,
-0.154056,
0.52689,
-0.105292,
0.014743,
-0.24148
],
"network.10.weight": [
[
0.167128,
0.042453,
0.723714,
-0.498909,
0.447135,
-0.381029,
-0.369644,
-0.580837
]
],
"network.10.bias": [
0.470912
]
}
## Activation Signature
### 0
mean: [-0.522027, 1.492200, 1.277292, -0.811429, 1.504321, -0.066854, 3.207143, 2.727548]
std: [2.154347, 1.967393, 1.831561, 2.256315, 2.527629, 2.069623, 2.763266, 2.551775]
### 2
mean: [-2.373778, 2.874953, 4.064094, -0.755796, 3.315709, -1.007630, 4.103614, 2.103999]
std: [1.807748, 2.750367, 3.567520, 1.869734, 3.868446, 1.750282, 4.037795, 2.896123]
### 4
mean: [-8.854488, 6.245457, 5.657223, 3.886035, 10.105027, 4.709582, 8.028404, -9.229414]
std: [9.374324, 5.655544, 5.577294, 3.468553, 9.598053, 4.644960, 7.607325, 10.000595]
### 6
mean: [-11.058739, 8.233209, -6.345510, 17.283922, -8.566502, -8.613591, 7.663009, -8.314028]
std: [10.896148, 8.021490, 6.129471, 16.406977, 8.116002, 8.480818, 7.636133, 7.488100]
### 8
mean: [-13.025500, 6.502821, -18.341728, 11.663562, -10.282469, 13.991090, 17.952879, 11.539808]
std: [12.900014, 6.500323, 18.283310, 11.332429, 10.675286, 13.634689, 17.309156, 11.580980]
### 10
mean: [-23.693647]
std: [23.749773]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_ascending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.884267,
-0.208205,
-0.084779,
0.211018,
0.543483
],
[
0.080458,
0.895161,
0.375676,
-0.369734,
-0.141958
],
[
0.751684,
-0.462626,
-0.162537,
0.153153,
0.511408
],
[
-0.616191,
-0.623524,
-0.209838,
0.416046,
0.437106
],
[
1.010424,
0.227613,
0.298011,
-0.292982,
-0.057433
],
[
-0.45793,
-0.497994,
-0.2292,
0.552926,
0.484453
],
[
0.586622,
0.98632,
0.361719,
-0.024558,
-0.317672
],
[
0.576554,
0.868929,
-0.148509,
0.366823,
-0.492616
]
],
"network.0.bias": [
0.01207,
-0.055898,
0.565228,
0.098631,
-0.091249,
0.10603,
0.368698,
0.557823
],
"network.2.weight": [
[
-0.12298,
0.159205,
-0.385202,
-0.389619,
-0.414965,
-0.418727,
0.170375,
-0.413143
],
[
0.333331,
0.229628,
-0.11894,
-0.163551,
0.483145,
0.196858,
0.347422,
0.296496
],
[
-0.582098,
0.289499,
0.309888,
-0.23134,
0.041321,
0.062782,
0.335784,
0.725186
],
[
-0.71537,
0.02614,
-0.117887,
-0.533271,
0.233341,
-0.331485,
-0.069866,
0.123304
],
[
-0.297146,
0.126444,
0.430981,
-0.241571,
0.384342,
-0.457616,
0.461763,
0.396369
],
[
0.377994,
-0.201517,
-0.190416,
0.358944,
0.211093,
0.222712,
-0.299422,
-0.169458
],
[
0.082736,
0.41789,
0.149856,
-0.406215,
0.072122,
-0.314971,
0.422428,
0.737851
],
[
-0.137935,
0.609279,
0.538319,
-0.153631,
0.367608,
-0.587858,
0.193983,
-0.173398
]
],
"network.2.bias": [
-0.389157,
-0.218677,
0.355591,
-0.427036,
-0.142009,
0.220111,
0.092792,
0.132204
],
"network.4.weight": [
[
0.160442,
-0.667645,
-0.495327,
-0.104838,
-0.6295,
0.379691,
-0.550798,
-0.526627
],
[
-0.031836,
0.426638,
0.561198,
-0.00181,
0.181791,
-0.323334,
0.14868,
0.506618
],
[
0.107225,
0.254567,
0.250818,
-0.065005,
0.109505,
-0.565368,
0.572167,
0.525292
],
[
-0.275381,
-0.071808,
0.488721,
-0.152092,
0.133183,
-0.2186,
0.344506,
0.053412
],
[
-0.310963,
0.363825,
0.619465,
-0.432531,
0.658534,
-0.227626,
0.658911,
0.660271
],
[
-0.30743,
0.035336,
0.037346,
0.10327,
0.390352,
-0.278424,
0.295819,
0.696831
],
[
0.274612,
0.410858,
0.616534,
-0.433699,
0.452809,
-0.245824,
0.321396,
0.629888
],
[
-0.091912,
-0.565915,
-0.610762,
-0.762713,
-0.630704,
0.262504,
-0.50556,
-0.721301
]
],
"network.4.bias": [
0.660901,
0.411356,
0.079681,
0.116752,
0.116625,
0.324114,
0.143176,
0.808632
],
"network.6.weight": [
[
0.451379,
-0.324239,
-0.037355,
-0.941549,
-0.414234,
-0.143817,
-0.118661,
0.322498
],
[
-0.531572,
-0.008186,
0.295057,
0.627673,
0.406025,
-0.241041,
0.187523,
-0.107176
],
[
0.560588,
-0.371578,
-0.211444,
-0.223476,
-0.122567,
-0.091722,
-0.049366,
0.550834
],
[
-0.329912,
0.480352,
0.371632,
0.603313,
0.495597,
0.327018,
0.425249,
-0.412362
],
[
-0.367088,
-0.312847,
-0.210535,
0.018959,
-0.336912,
-0.448448,
0.003529,
0.07194
],
[
0.687545,
-0.527265,
0.024541,
-0.291059,
-0.420738,
-0.104432,
0.003346,
0.584944
],
[
-0.380494,
0.141094,
0.275808,
0.196025,
0.333952,
0.171109,
0.078006,
-0.221937
],
[
-0.064171,
-0.612749,
-0.277962,
-0.644033,
-0.323719,
0.08205,
0.3208,
-0.086161
]
],
"network.6.bias": [
0.564595,
-0.252215,
0.006863,
-0.052739,
0.015617,
0.256395,
-0.298876,
-0.085647
],
"network.8.weight": [
[
0.727817,
-0.742242,
0.004505,
-0.080493,
0.255715,
0.364046,
-0.73413,
-0.07103
],
[
0.01153,
-0.011429,
0.131153,
0.34151,
0.32882,
-0.233352,
0.12838,
0.406655
],
[
0.504262,
-0.730654,
-0.017194,
-0.427107,
-0.271088,
0.501936,
-0.707504,
-0.43853
],
[
-0.301098,
0.001369,
-0.390173,
0.580283,
0.114836,
-0.079997,
0.233588,
-0.28008
],
[
0.839497,
-0.314523,
0.639131,
-0.29879,
-0.145503,
0.16169,
-0.410782,
0.134225
],
[
-0.583781,
0.323504,
-0.367379,
0.561267,
0.186871,
-0.245,
0.232523,
-0.125561
],
[
-0.350124,
0.450341,
-0.338578,
0.562622,
0.130654,
-0.15929,
0.587091,
0.100226
],
[
-0.602727,
0.750379,
-0.318099,
0.050578,
-0.324114,
-0.122592,
0.618681,
0.103213
]
],
"network.8.bias": [
0.078719,
-0.278176,
0.453421,
-0.154056,
0.52689,
-0.105292,
0.014743,
-0.24148
],
"network.10.weight": [
[
0.167128,
0.042453,
0.723714,
-0.498909,
0.447135,
-0.381029,
-0.369644,
-0.580837
]
],
"network.10.bias": [
0.470912
]
}
## Activation Signature
### 0
mean: [-0.522027, 1.492200, 1.277292, -0.811429, 1.504321, -0.066854, 3.207143, 2.727548]
std: [2.154347, 1.967393, 1.831561, 2.256315, 2.527629, 2.069623, 2.763266, 2.551775]
### 2
mean: [-2.373778, 2.874953, 4.064094, -0.755796, 3.315709, -1.007630, 4.103614, 2.103999]
std: [1.807748, 2.750367, 3.567520, 1.869734, 3.868446, 1.750282, 4.037795, 2.896123]
### 4
mean: [-8.854488, 6.245457, 5.657223, 3.886035, 10.105027, 4.709582, 8.028404, -9.229414]
std: [9.374324, 5.655544, 5.577294, 3.468553, 9.598053, 4.644960, 7.607325, 10.000595]
### 6
mean: [-11.058739, 8.233209, -6.345510, 17.283922, -8.566502, -8.613591, 7.663009, -8.314028]
std: [10.896148, 8.021490, 6.129471, 16.406977, 8.116002, 8.480818, 7.636133, 7.488100]
### 8
mean: [-13.025500, 6.502821, -18.341728, 11.663562, -10.282469, 13.991090, 17.952879, 11.539808]
std: [12.900014, 6.500323, 18.283310, 11.332429, 10.675286, 13.634689, 17.309156, 11.580980]
### 10
mean: [-23.693647]
std: [23.749773]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_ascending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -0.5220270156860352, "std": 2.1543471813201904}, "1": {"mean": 1.4922000169754028, "std": 1.9673932790756226}, "2": {"mean": 1.2772918939590454, "std": 1.8315613269805908}, "3": {"mean": -0.8114286661148071, "std": 2.256314992904663}, "4": {"mean": 1.5043214559555054, "std": 2.5276288986206055}, "5": {"mean": -0.06685435026884079, "std": 2.06962251663208}, "6": {"mean": 3.2071428298950195, "std": 2.76326584815979}, "7": {"mean": 2.7275476455688477, "std": 2.551774501800537}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -2.3737776279449463, "std": 1.8077483177185059}, "1": {"mean": 2.874952554702759, "std": 2.7503674030303955}, "2": {"mean": 4.064094066619873, "std": 3.5675196647644043}, "3": {"mean": -0.755795955657959, "std": 1.8697339296340942}, "4": {"mean": 3.315709352493286, "std": 3.868446111679077}, "5": {"mean": -1.0076295137405396, "std": 1.7502821683883667}, "6": {"mean": 4.103614330291748, "std": 4.037795066833496}, "7": {"mean": 2.1039986610412598, "std": 2.896122932434082}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -8.854488372802734, "std": 9.374323844909668}, "1": {"mean": 6.245457172393799, "std": 5.655543804168701}, "2": {"mean": 5.657223224639893, "std": 5.57729434967041}, "3": {"mean": 3.886035203933716, "std": 3.468553066253662}, "4": {"mean": 10.105027198791504, "std": 9.598052978515625}, "5": {"mean": 4.7095818519592285, "std": 4.644959926605225}, "6": {"mean": 8.028404235839844, "std": 7.607325077056885}, "7": {"mean": -9.229413986206055, "std": 10.000595092773438}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -11.058738708496094, "std": 10.896147727966309}, "1": {"mean": 8.233208656311035, "std": 8.021490097045898}, "2": {"mean": -6.345510482788086, "std": 6.129471302032471}, "3": {"mean": 17.28392219543457, "std": 16.4069766998291}, "4": {"mean": -8.56650161743164, "std": 8.116002082824707}, "5": {"mean": -8.613591194152832, "std": 8.480817794799805}, "6": {"mean": 7.663008689880371, "std": 7.636133193969727}, "7": {"mean": -8.314027786254883, "std": 7.488100051879883}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -13.025500297546387, "std": 12.90001392364502}, "1": {"mean": 6.50282096862793, "std": 6.500323295593262}, "2": {"mean": -18.34172821044922, "std": 18.283309936523438}, "3": {"mean": 11.663561820983887, "std": 11.332428932189941}, "4": {"mean": -10.282468795776367, "std": 10.675286293029785}, "5": {"mean": 13.991089820861816, "std": 13.634689331054688}, "6": {"mean": 17.952878952026367, "std": 17.30915641784668}, "7": {"mean": 11.53980827331543, "std": 11.58098030090332}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -23.693647384643555, "std": 23.749773025512695}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.884267, -0.208205, -0.084779, 0.211018, 0.543483], [0.080458, 0.895161, 0.375676, -0.369734, -0.141958], [0.751684, -0.462626, -0.162537, 0.153153, 0.511408], [-0.616191, -0.623524, -0.209838, 0.416046, 0.437106], [1.010424, 0.227613, 0.298011, -0.292982, -0.057433], [-0.45793, -0.497994, -0.2292, 0.552926, 0.484453], [0.586622, 0.98632, 0.361719, -0.024558, -0.317672], [0.576554, 0.868929, -0.148509, 0.366823, -0.492616]], "network.0.bias": [0.01207, -0.055898, 0.565228, 0.098631, -0.091249, 0.10603, 0.368698, 0.557823], "network.2.weight": [[-0.12298, 0.159205, -0.385202, -0.389619, -0.414965, -0.418727, 0.170375, -0.413143], [0.333331, 0.229628, -0.11894, -0.163551, 0.483145, 0.196858, 0.347422, 0.296496], [-0.582098, 0.289499, 0.309888, -0.23134, 0.041321, 0.062782, 0.335784, 0.725186], [-0.71537, 0.02614, -0.117887, -0.533271, 0.233341, -0.331485, -0.069866, 0.123304], [-0.297146, 0.126444, 0.430981, -0.241571, 0.384342, -0.457616, 0.461763, 0.396369], [0.377994, -0.201517, -0.190416, 0.358944, 0.211093, 0.222712, -0.299422, -0.169458], [0.082736, 0.41789, 0.149856, -0.406215, 0.072122, -0.314971, 0.422428, 0.737851], [-0.137935, 0.609279, 0.538319, -0.153631, 0.367608, -0.587858, 0.193983, -0.173398]], "network.2.bias": [-0.389157, -0.218677, 0.355591, -0.427036, -0.142009, 0.220111, 0.092792, 0.132204], "network.4.weight": [[0.160442, -0.667645, -0.495327, -0.104838, -0.6295, 0.379691, -0.550798, -0.526627], [-0.031836, 0.426638, 0.561198, -0.00181, 0.181791, -0.323334, 0.14868, 0.506618], [0.107225, 0.254567, 0.250818, -0.065005, 0.109505, -0.565368, 0.572167, 0.525292], [-0.275381, -0.071808, 0.488721, -0.152092, 0.133183, -0.2186, 0.344506, 0.053412], [-0.310963, 0.363825, 0.619465, -0.432531, 0.658534, -0.227626, 0.658911, 0.660271], [-0.30743, 0.035336, 0.037346, 0.10327, 0.390352, -0.278424, 0.295819, 0.696831], [0.274612, 0.410858, 0.616534, -0.433699, 0.452809, -0.245824, 0.321396, 0.629888], [-0.091912, -0.565915, -0.610762, -0.762713, -0.630704, 0.262504, -0.50556, -0.721301]], "network.4.bias": [0.660901, 0.411356, 0.079681, 0.116752, 0.116625, 0.324114, 0.143176, 0.808632], "network.6.weight": [[0.451379, -0.324239, -0.037355, -0.941549, -0.414234, -0.143817, -0.118661, 0.322498], [-0.531572, -0.008186, 0.295057, 0.627673, 0.406025, -0.241041, 0.187523, -0.107176], [0.560588, -0.371578, -0.211444, -0.223476, -0.122567, -0.091722, -0.049366, 0.550834], [-0.329912, 0.480352, 0.371632, 0.603313, 0.495597, 0.327018, 0.425249, -0.412362], [-0.367088, -0.312847, -0.210535, 0.018959, -0.336912, -0.448448, 0.003529, 0.07194], [0.687545, -0.527265, 0.024541, -0.291059, -0.420738, -0.104432, 0.003346, 0.584944], [-0.380494, 0.141094, 0.275808, 0.196025, 0.333952, 0.171109, 0.078006, -0.221937], [-0.064171, -0.612749, -0.277962, -0.644033, -0.323719, 0.08205, 0.3208, -0.086161]], "network.6.bias": [0.564595, -0.252215, 0.006863, -0.052739, 0.015617, 0.256395, -0.298876, -0.085647], "network.8.weight": [[0.727817, -0.742242, 0.004505, -0.080493, 0.255715, 0.364046, -0.73413, -0.07103], [0.01153, -0.011429, 0.131153, 0.34151, 0.32882, -0.233352, 0.12838, 0.406655], [0.504262, -0.730654, -0.017194, -0.427107, -0.271088, 0.501936, -0.707504, -0.43853], [-0.301098, 0.001369, -0.390173, 0.580283, 0.114836, -0.079997, 0.233588, -0.28008], [0.839497, -0.314523, 0.639131, -0.29879, -0.145503, 0.16169, -0.410782, 0.134225], [-0.583781, 0.323504, -0.367379, 0.561267, 0.186871, -0.245, 0.232523, -0.125561], [-0.350124, 0.450341, -0.338578, 0.562622, 0.130654, -0.15929, 0.587091, 0.100226], [-0.602727, 0.750379, -0.318099, 0.050578, -0.324114, -0.122592, 0.618681, 0.103213]], "network.8.bias": [0.078719, -0.278176, 0.453421, -0.154056, 0.52689, -0.105292, 0.014743, -0.24148], "network.10.weight": [[0.167128, 0.042453, 0.723714, -0.498909, 0.447135, -0.381029, -0.369644, -0.580837]], "network.10.bias": [0.470912]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.687283992767334, "train_acc": 0.475, "val_loss": 0.5355622172355652, "val_acc": 0.6}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.5482286661863327, "train_acc": 0.555, "val_loss": 0.3604751527309418, "val_acc": 0.6}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.5041339695453644, "train_acc": 0.64, "val_loss": 0.4174022972583771, "val_acc": 0.78}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.4047901928424835, "train_acc": 0.835, "val_loss": 0.30476728081703186, "val_acc": 0.86}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.2785230353474617, "train_acc": 0.885, "val_loss": 0.5558382272720337, "val_acc": 0.78}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.4600400775671005, "train_acc": 0.81, "val_loss": 0.4755459725856781, "val_acc": 0.88}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.4668443202972412, "train_acc": 0.875, "val_loss": 0.4133372902870178, "val_acc": 0.86}], "summary": {"total_epochs": 7, "degraded_epochs": 2, "improved_epochs": 5, "patterns": ["sorted_ascending"], "degraded_stage": {"initial_val_loss": 0.5355622172355652, "final_val_loss": 0.3604751527309418, "initial_val_acc": 0.6, "final_val_acc": 0.6, "best_val_acc": 0.6}, "improved_stage": {"initial_val_loss": 0.4174022972583771, "final_val_loss": 0.4133372902870178, "initial_val_acc": 0.78, "final_val_acc": 0.86, "best_val_acc": 0.88, "best_epoch": 5}, "improvement": 0.28, "first_improvement_epoch": 1}}
|
16
|
{"target_pattern": "sorted_ascending", "degraded_accuracy": 0.66, "improved_accuracy": 0.88, "improvement": 0.21999999999999997, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 4346, "learning_rate": 0.054312033149829644, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_ascending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_ascending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.221801,
0.120518,
-0.752228,
0.269084,
-0.838705
],
[
0.733041,
1.086597,
-0.098206,
0.01459,
0.207951
],
[
-0.283289,
0.038512,
0.346573,
-0.494943,
-0.18144
],
[
-0.824084,
-0.5601,
-0.131687,
-0.182958,
-0.202821
],
[
-0.91855,
-0.561557,
0.06114,
0.28438,
0.613556
],
[
0.840939,
0.879959,
0.439141,
-0.025973,
-0.021964
]
],
"network.0.bias": [
-0.582468,
0.136459,
0.31486,
0.324464,
0.469343,
0.376895
],
"network.2.weight": [
[
-0.295073,
-0.073034,
0.098272,
-0.125084,
-0.66174,
-0.539556
],
[
0.297332,
-0.426974,
-0.195938,
0.730976,
-0.680779,
0.04367
],
[
0.118697,
-0.172811,
-0.253448,
0.415821,
-0.526726,
0.106636
],
[
-0.21122,
-0.16278,
-0.329873,
-0.689088,
0.645527,
-0.670833
],
[
-0.116093,
-0.533822,
0.11469,
-0.687492,
0.098549,
-0.420388
],
[
-0.242655,
-0.625867,
-0.700309,
-1.082138,
0.537008,
-0.654846
]
],
"network.2.bias": [
-0.045815,
-0.750876,
-0.563745,
-0.142687,
-0.212866,
0.563181
],
"network.4.weight": [
[
-0.303655,
0.043824,
-0.057165,
0.483009,
-1.070052,
0.696382
],
[
-0.559677,
-0.255967,
0.150125,
0.037936,
-0.292217,
0.483686
],
[
0.386781,
0.263452,
-0.115771,
-0.09391,
0.608835,
-0.413751
],
[
0.255899,
-0.124583,
-0.46334,
-0.364984,
0.3183,
-0.31928
],
[
0.525805,
-0.223048,
-0.419708,
-0.485165,
-0.201467,
0.126589
],
[
-0.328713,
-0.197384,
-0.520878,
-0.074452,
-0.668866,
0.694118
]
],
"network.4.bias": [
-0.069977,
-0.318946,
0.325288,
0.333785,
-0.294682,
0.575721
],
"network.6.weight": [
[
0.303445,
0.824522,
-0.420472,
-0.34738,
0.667396,
-0.449535
],
[
0.713063,
-0.139692,
-0.407604,
-0.097634,
-0.184238,
0.758021
],
[
0.598995,
0.085339,
-0.063345,
0.190847,
0.165368,
0.513456
],
[
-0.685746,
-0.325918,
0.660969,
0.896587,
-0.097597,
0.003921
],
[
-0.347327,
0.147002,
0.218444,
0.084524,
-0.45007,
-0.28397
],
[
0.05677,
0.471938,
-0.969819,
0.131905,
0.206201,
0.526839
]
],
"network.6.bias": [
-0.550129,
0.177479,
0.176831,
0.471247,
0.098967,
0.152891
],
"network.8.weight": [
[
-0.313275,
-0.553827,
-0.343031,
0.554272,
0.682003,
-0.529314
],
[
-0.006665,
-0.519951,
0.104477,
0.949553,
0.754085,
-0.559229
],
[
-0.163556,
-0.073835,
0.274271,
0.561857,
0.484639,
0.283185
],
[
-0.039134,
0.498677,
0.123283,
0.266015,
0.026128,
0.176184
],
[
-0.216014,
0.716527,
0.347894,
-0.160275,
-0.344077,
0.295929
],
[
-0.455227,
-0.257669,
-0.404767,
1.06888,
0.321962,
-0.522794
]
],
"network.8.bias": [
0.624309,
0.140772,
0.218866,
-0.05558,
-0.143226,
0.418655
],
"network.10.weight": [
[
0.651805,
0.684951,
0.121295,
-0.211563,
-0.61308,
1.004816
],
[
-0.498659,
0.112261,
0.011497,
0.565217,
0.613124,
-0.304578
],
[
-0.604086,
0.039351,
0.267094,
0.05031,
-0.081928,
0.403987
],
[
-0.323033,
0.039146,
0.044986,
-0.346166,
0.223854,
-0.195519
],
[
-0.594393,
-0.148949,
0.130349,
0.343312,
0.383656,
0.049272
],
[
-0.406774,
-0.276063,
0.29535,
0.405477,
0.656324,
0.312455
]
],
"network.10.bias": [
0.756418,
0.250546,
-0.325551,
-0.230636,
0.106496,
-0.010392
],
"network.12.weight": [
[
-0.997013,
0.312639,
0.453054,
0.152787,
0.377949,
0.594075
]
],
"network.12.bias": [
-0.613281
]
}
## Activation Signature
### 0
mean: [-2.115527, 3.111329, -0.528703, -2.639256, -0.202754, 3.861406]
std: [2.258271, 2.876068, 1.294950, 2.502930, 2.546983, 3.093587]
### 2
mean: [-2.792531, -2.434631, -1.120567, -2.787373, -3.349144, -3.575913]
std: [1.768723, 1.148983, 0.643767, 2.936522, 2.888012, 4.062042]
### 4
mean: [0.108472, -0.234446, 0.229511, 0.320041, -0.258808, 0.773155]
std: [0.536104, 0.241069, 0.239292, 0.322580, 0.192730, 0.281491]
### 6
mean: [-1.060223, 0.664221, 0.573403, 0.741524, -0.030157, 0.306811]
std: [0.212490, 0.604596, 0.455351, 0.529807, 0.257186, 0.329774]
### 8
mean: [0.482441, 0.416226, 0.751237, 0.474130, 0.368993, 0.758232]
std: [0.829735, 0.694044, 0.105484, 0.375375, 0.750095, 0.770346]
### 10
mean: [1.905577, 0.215216, -0.160021, -0.540669, 0.101271, 0.418169]
std: [1.097619, 0.828275, 0.042569, 0.172173, 0.546411, 0.697362]
### 12
mean: [-2.305136]
std: [1.588279]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_ascending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.221801,
0.120518,
-0.752228,
0.269084,
-0.838705
],
[
0.733041,
1.086597,
-0.098206,
0.01459,
0.207951
],
[
-0.283289,
0.038512,
0.346573,
-0.494943,
-0.18144
],
[
-0.824084,
-0.5601,
-0.131687,
-0.182958,
-0.202821
],
[
-0.91855,
-0.561557,
0.06114,
0.28438,
0.613556
],
[
0.840939,
0.879959,
0.439141,
-0.025973,
-0.021964
]
],
"network.0.bias": [
-0.582468,
0.136459,
0.31486,
0.324464,
0.469343,
0.376895
],
"network.2.weight": [
[
-0.295073,
-0.073034,
0.098272,
-0.125084,
-0.66174,
-0.539556
],
[
0.297332,
-0.426974,
-0.195938,
0.730976,
-0.680779,
0.04367
],
[
0.118697,
-0.172811,
-0.253448,
0.415821,
-0.526726,
0.106636
],
[
-0.21122,
-0.16278,
-0.329873,
-0.689088,
0.645527,
-0.670833
],
[
-0.116093,
-0.533822,
0.11469,
-0.687492,
0.098549,
-0.420388
],
[
-0.242655,
-0.625867,
-0.700309,
-1.082138,
0.537008,
-0.654846
]
],
"network.2.bias": [
-0.045815,
-0.750876,
-0.563745,
-0.142687,
-0.212866,
0.563181
],
"network.4.weight": [
[
-0.303655,
0.043824,
-0.057165,
0.483009,
-1.070052,
0.696382
],
[
-0.559677,
-0.255967,
0.150125,
0.037936,
-0.292217,
0.483686
],
[
0.386781,
0.263452,
-0.115771,
-0.09391,
0.608835,
-0.413751
],
[
0.255899,
-0.124583,
-0.46334,
-0.364984,
0.3183,
-0.31928
],
[
0.525805,
-0.223048,
-0.419708,
-0.485165,
-0.201467,
0.126589
],
[
-0.328713,
-0.197384,
-0.520878,
-0.074452,
-0.668866,
0.694118
]
],
"network.4.bias": [
-0.069977,
-0.318946,
0.325288,
0.333785,
-0.294682,
0.575721
],
"network.6.weight": [
[
0.303445,
0.824522,
-0.420472,
-0.34738,
0.667396,
-0.449535
],
[
0.713063,
-0.139692,
-0.407604,
-0.097634,
-0.184238,
0.758021
],
[
0.598995,
0.085339,
-0.063345,
0.190847,
0.165368,
0.513456
],
[
-0.685746,
-0.325918,
0.660969,
0.896587,
-0.097597,
0.003921
],
[
-0.347327,
0.147002,
0.218444,
0.084524,
-0.45007,
-0.28397
],
[
0.05677,
0.471938,
-0.969819,
0.131905,
0.206201,
0.526839
]
],
"network.6.bias": [
-0.550129,
0.177479,
0.176831,
0.471247,
0.098967,
0.152891
],
"network.8.weight": [
[
-0.313275,
-0.553827,
-0.343031,
0.554272,
0.682003,
-0.529314
],
[
-0.006665,
-0.519951,
0.104477,
0.949553,
0.754085,
-0.559229
],
[
-0.163556,
-0.073835,
0.274271,
0.561857,
0.484639,
0.283185
],
[
-0.039134,
0.498677,
0.123283,
0.266015,
0.026128,
0.176184
],
[
-0.216014,
0.716527,
0.347894,
-0.160275,
-0.344077,
0.295929
],
[
-0.455227,
-0.257669,
-0.404767,
1.06888,
0.321962,
-0.522794
]
],
"network.8.bias": [
0.624309,
0.140772,
0.218866,
-0.05558,
-0.143226,
0.418655
],
"network.10.weight": [
[
0.651805,
0.684951,
0.121295,
-0.211563,
-0.61308,
1.004816
],
[
-0.498659,
0.112261,
0.011497,
0.565217,
0.613124,
-0.304578
],
[
-0.604086,
0.039351,
0.267094,
0.05031,
-0.081928,
0.403987
],
[
-0.323033,
0.039146,
0.044986,
-0.346166,
0.223854,
-0.195519
],
[
-0.594393,
-0.148949,
0.130349,
0.343312,
0.383656,
0.049272
],
[
-0.406774,
-0.276063,
0.29535,
0.405477,
0.656324,
0.312455
]
],
"network.10.bias": [
0.756418,
0.250546,
-0.325551,
-0.230636,
0.106496,
-0.010392
],
"network.12.weight": [
[
-0.997013,
0.312639,
0.453054,
0.152787,
0.377949,
0.594075
]
],
"network.12.bias": [
-0.613281
]
}
## Activation Signature
### 0
mean: [-2.115527, 3.111329, -0.528703, -2.639256, -0.202754, 3.861406]
std: [2.258271, 2.876068, 1.294950, 2.502930, 2.546983, 3.093587]
### 2
mean: [-2.792531, -2.434631, -1.120567, -2.787373, -3.349144, -3.575913]
std: [1.768723, 1.148983, 0.643767, 2.936522, 2.888012, 4.062042]
### 4
mean: [0.108472, -0.234446, 0.229511, 0.320041, -0.258808, 0.773155]
std: [0.536104, 0.241069, 0.239292, 0.322580, 0.192730, 0.281491]
### 6
mean: [-1.060223, 0.664221, 0.573403, 0.741524, -0.030157, 0.306811]
std: [0.212490, 0.604596, 0.455351, 0.529807, 0.257186, 0.329774]
### 8
mean: [0.482441, 0.416226, 0.751237, 0.474130, 0.368993, 0.758232]
std: [0.829735, 0.694044, 0.105484, 0.375375, 0.750095, 0.770346]
### 10
mean: [1.905577, 0.215216, -0.160021, -0.540669, 0.101271, 0.418169]
std: [1.097619, 0.828275, 0.042569, 0.172173, 0.546411, 0.697362]
### 12
mean: [-2.305136]
std: [1.588279]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_ascending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -2.1155269145965576, "std": 2.258270502090454}, "1": {"mean": 3.1113288402557373, "std": 2.876068115234375}, "2": {"mean": -0.528702974319458, "std": 1.2949503660202026}, "3": {"mean": -2.6392557621002197, "std": 2.5029296875}, "4": {"mean": -0.2027536779642105, "std": 2.546982526779175}, "5": {"mean": 3.861406087875366, "std": 3.0935866832733154}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -2.7925314903259277, "std": 1.7687230110168457}, "1": {"mean": -2.43463134765625, "std": 1.1489830017089844}, "2": {"mean": -1.1205666065216064, "std": 0.6437669396400452}, "3": {"mean": -2.7873730659484863, "std": 2.9365224838256836}, "4": {"mean": -3.349144220352173, "std": 2.8880116939544678}, "5": {"mean": -3.575913429260254, "std": 4.062041759490967}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.10847161710262299, "std": 0.5361035466194153}, "1": {"mean": -0.23444627225399017, "std": 0.24106885492801666}, "2": {"mean": 0.22951050102710724, "std": 0.2392917424440384}, "3": {"mean": 0.3200410306453705, "std": 0.3225802183151245}, "4": {"mean": -0.2588083744049072, "std": 0.19272960722446442}, "5": {"mean": 0.773155152797699, "std": 0.2814907729625702}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -1.0602226257324219, "std": 0.21248964965343475}, "1": {"mean": 0.6642209887504578, "std": 0.6045964956283569}, "2": {"mean": 0.5734031200408936, "std": 0.4553510248661041}, "3": {"mean": 0.7415236830711365, "std": 0.5298067927360535}, "4": {"mean": -0.030156750231981277, "std": 0.25718557834625244}, "5": {"mean": 0.3068108856678009, "std": 0.32977449893951416}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.4824405312538147, "std": 0.8297349214553833}, "1": {"mean": 0.4162261486053467, "std": 0.6940436363220215}, "2": {"mean": 0.7512367963790894, "std": 0.1054839938879013}, "3": {"mean": 0.4741300940513611, "std": 0.37537485361099243}, "4": {"mean": 0.36899322271347046, "std": 0.7500953078269958}, "5": {"mean": 0.7582318782806396, "std": 0.7703464031219482}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": 1.905577301979065, "std": 1.0976189374923706}, "1": {"mean": 0.2152164876461029, "std": 0.8282746076583862}, "2": {"mean": -0.16002118587493896, "std": 0.042569126933813095}, "3": {"mean": -0.5406686663627625, "std": 0.1721726804971695}, "4": {"mean": 0.10127142071723938, "std": 0.5464112162590027}, "5": {"mean": 0.41816920042037964, "std": 0.6973615288734436}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -2.3051364421844482, "std": 1.588279366493225}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.221801, 0.120518, -0.752228, 0.269084, -0.838705], [0.733041, 1.086597, -0.098206, 0.01459, 0.207951], [-0.283289, 0.038512, 0.346573, -0.494943, -0.18144], [-0.824084, -0.5601, -0.131687, -0.182958, -0.202821], [-0.91855, -0.561557, 0.06114, 0.28438, 0.613556], [0.840939, 0.879959, 0.439141, -0.025973, -0.021964]], "network.0.bias": [-0.582468, 0.136459, 0.31486, 0.324464, 0.469343, 0.376895], "network.2.weight": [[-0.295073, -0.073034, 0.098272, -0.125084, -0.66174, -0.539556], [0.297332, -0.426974, -0.195938, 0.730976, -0.680779, 0.04367], [0.118697, -0.172811, -0.253448, 0.415821, -0.526726, 0.106636], [-0.21122, -0.16278, -0.329873, -0.689088, 0.645527, -0.670833], [-0.116093, -0.533822, 0.11469, -0.687492, 0.098549, -0.420388], [-0.242655, -0.625867, -0.700309, -1.082138, 0.537008, -0.654846]], "network.2.bias": [-0.045815, -0.750876, -0.563745, -0.142687, -0.212866, 0.563181], "network.4.weight": [[-0.303655, 0.043824, -0.057165, 0.483009, -1.070052, 0.696382], [-0.559677, -0.255967, 0.150125, 0.037936, -0.292217, 0.483686], [0.386781, 0.263452, -0.115771, -0.09391, 0.608835, -0.413751], [0.255899, -0.124583, -0.46334, -0.364984, 0.3183, -0.31928], [0.525805, -0.223048, -0.419708, -0.485165, -0.201467, 0.126589], [-0.328713, -0.197384, -0.520878, -0.074452, -0.668866, 0.694118]], "network.4.bias": [-0.069977, -0.318946, 0.325288, 0.333785, -0.294682, 0.575721], "network.6.weight": [[0.303445, 0.824522, -0.420472, -0.34738, 0.667396, -0.449535], [0.713063, -0.139692, -0.407604, -0.097634, -0.184238, 0.758021], [0.598995, 0.085339, -0.063345, 0.190847, 0.165368, 0.513456], [-0.685746, -0.325918, 0.660969, 0.896587, -0.097597, 0.003921], [-0.347327, 0.147002, 0.218444, 0.084524, -0.45007, -0.28397], [0.05677, 0.471938, -0.969819, 0.131905, 0.206201, 0.526839]], "network.6.bias": [-0.550129, 0.177479, 0.176831, 0.471247, 0.098967, 0.152891], "network.8.weight": [[-0.313275, -0.553827, -0.343031, 0.554272, 0.682003, -0.529314], [-0.006665, -0.519951, 0.104477, 0.949553, 0.754085, -0.559229], [-0.163556, -0.073835, 0.274271, 0.561857, 0.484639, 0.283185], [-0.039134, 0.498677, 0.123283, 0.266015, 0.026128, 0.176184], [-0.216014, 0.716527, 0.347894, -0.160275, -0.344077, 0.295929], [-0.455227, -0.257669, -0.404767, 1.06888, 0.321962, -0.522794]], "network.8.bias": [0.624309, 0.140772, 0.218866, -0.05558, -0.143226, 0.418655], "network.10.weight": [[0.651805, 0.684951, 0.121295, -0.211563, -0.61308, 1.004816], [-0.498659, 0.112261, 0.011497, 0.565217, 0.613124, -0.304578], [-0.604086, 0.039351, 0.267094, 0.05031, -0.081928, 0.403987], [-0.323033, 0.039146, 0.044986, -0.346166, 0.223854, -0.195519], [-0.594393, -0.148949, 0.130349, 0.343312, 0.383656, 0.049272], [-0.406774, -0.276063, 0.29535, 0.405477, 0.656324, 0.312455]], "network.10.bias": [0.756418, 0.250546, -0.325551, -0.230636, 0.106496, -0.010392], "network.12.weight": [[-0.997013, 0.312639, 0.453054, 0.152787, 0.377949, 0.594075]], "network.12.bias": [-0.613281]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.693028599023819, "train_acc": 0.56, "val_loss": 0.6817962527275085, "val_acc": 0.56}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6749905943870544, "train_acc": 0.57, "val_loss": 0.6475306749343872, "val_acc": 0.66}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.5576570779085159, "train_acc": 0.735, "val_loss": 0.5263103246688843, "val_acc": 0.8}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.47393909096717834, "train_acc": 0.82, "val_loss": 0.4503064751625061, "val_acc": 0.82}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.3619961589574814, "train_acc": 0.86, "val_loss": 0.46128952503204346, "val_acc": 0.84}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.2842746078968048, "train_acc": 0.9, "val_loss": 0.37456852197647095, "val_acc": 0.82}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.2487887442111969, "train_acc": 0.895, "val_loss": 0.4472322463989258, "val_acc": 0.86}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.21531527489423752, "train_acc": 0.925, "val_loss": 0.3761889636516571, "val_acc": 0.84}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.18309546262025833, "train_acc": 0.94, "val_loss": 0.3642658293247223, "val_acc": 0.84}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.1787835732102394, "train_acc": 0.925, "val_loss": 0.4074441194534302, "val_acc": 0.88}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.1672893390059471, "train_acc": 0.945, "val_loss": 0.3930937945842743, "val_acc": 0.88}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.14134202897548676, "train_acc": 0.94, "val_loss": 0.3323172330856323, "val_acc": 0.88}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["sorted_ascending"], "degraded_stage": {"initial_val_loss": 0.6817962527275085, "final_val_loss": 0.6475306749343872, "initial_val_acc": 0.56, "final_val_acc": 0.66, "best_val_acc": 0.66}, "improved_stage": {"initial_val_loss": 0.5263103246688843, "final_val_loss": 0.3323172330856323, "initial_val_acc": 0.8, "final_val_acc": 0.88, "best_val_acc": 0.88, "best_epoch": 9}, "improvement": 0.21999999999999997, "first_improvement_epoch": 1}}
|
17
|
{"target_pattern": "sorted_ascending", "degraded_accuracy": 0.44, "improved_accuracy": 0.92, "improvement": 0.48000000000000004, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 3870, "learning_rate": 0.09151949117007849, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_ascending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_ascending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.535941,
1.153052,
0.099658,
0.018781,
-0.439558
],
[
0.777389,
1.345158,
0.433494,
0.285428,
0.166333
],
[
1.375589,
1.394691,
-0.320602,
0.127047,
-0.572512
],
[
-0.936511,
-0.508292,
-0.136284,
-0.394348,
0.397943
],
[
-1.032581,
-0.892143,
0.175657,
0.246191,
0.585853
],
[
-0.842523,
-0.901359,
-0.129675,
0.193045,
0.792036
],
[
0.087398,
-0.053117,
-0.034788,
-0.383557,
-0.448823
],
[
0.059242,
0.096149,
-0.318948,
-0.636324,
-0.445504
]
],
"network.0.bias": [
-0.004553,
0.439852,
0.284671,
-0.273153,
-0.360287,
0.29561,
-0.030387,
-0.345326
],
"network.2.weight": [
[
-0.419017,
0.209323,
-0.39008,
-0.090301,
-0.904995,
-0.398318,
-0.407279,
0.302107
],
[
-0.228625,
0.663002,
0.179008,
0.204887,
-0.282368,
-0.611383,
-0.215461,
0.084327
],
[
-0.323374,
0.220486,
-0.486795,
-0.745223,
-0.77429,
-0.769235,
-0.558304,
0.157807
],
[
0.170758,
0.372368,
0.743864,
0.162215,
-0.138108,
-0.1126,
-0.624642,
-0.331581
],
[
0.333079,
0.774701,
0.370664,
-0.27656,
-0.551922,
-0.297771,
0.16102,
-0.377714
],
[
-0.066247,
0.867989,
0.378464,
0.630444,
-0.460542,
-0.314837,
-0.03136,
0.171691
],
[
-0.307608,
-0.387041,
-0.484128,
0.04749,
-0.572609,
-0.571523,
0.007026,
0.33116
],
[
-0.002347,
0.940829,
0.163162,
0.286746,
-0.274334,
-0.606329,
0.089315,
0.049258
]
],
"network.2.bias": [
-0.503092,
0.009829,
-0.64565,
0.14282,
-0.005624,
-0.038831,
-0.603297,
-0.096228
],
"network.4.weight": [
[
0.190873,
0.250394,
0.000393,
0.542255,
0.302436,
0.304766,
0.590486,
0.147346
],
[
-0.005379,
0.485874,
-0.561582,
0.294674,
0.509009,
0.339655,
-0.074008,
0.424549
],
[
-0.670126,
0.169099,
-0.562449,
-0.484859,
-0.525167,
-0.782384,
-0.803294,
-0.503633
],
[
-0.46335,
0.30003,
-0.691419,
-0.730536,
-0.081147,
-0.449656,
-0.992384,
0.158799
],
[
-0.201608,
0.586103,
-0.327738,
0.462588,
0.548127,
0.764371,
-0.153318,
0.512544
],
[
0.257749,
-0.117062,
0.097028,
0.282699,
-0.157849,
-0.35338,
0.489625,
-0.345691
],
[
-0.355774,
0.349324,
-0.409083,
-0.209299,
-0.207674,
-0.359116,
-1.115008,
-0.204086
],
[
-0.198103,
0.010424,
-0.172836,
0.407244,
0.226819,
0.240898,
0.112636,
0.399937
]
],
"network.4.bias": [
-0.006739,
-0.452898,
0.543324,
0.755556,
-0.176079,
-0.551867,
0.11232,
-0.263366
],
"network.6.weight": [
[
-0.041436,
0.317157,
0.587369,
0.200397,
0.370582,
0.221502,
0.662896,
0.16701
],
[
-0.394576,
0.363954,
-0.126078,
0.38775,
0.015412,
-0.466954,
0.209826,
-0.408156
],
[
0.601418,
0.280085,
-0.175948,
-0.174804,
0.482239,
0.482419,
-0.133885,
0.244369
],
[
-1.109751,
0.675097,
-0.422064,
-0.698888,
0.332984,
0.56161,
-0.013102,
-0.867662
],
[
-0.672975,
-0.367003,
0.283044,
0.93607,
-0.815692,
-0.075493,
-0.367714,
-0.561771
],
[
-0.533968,
-0.322807,
0.937708,
0.824641,
-0.727859,
-0.413391,
0.185551,
-0.743018
],
[
-0.620542,
-0.097337,
0.283862,
0.268394,
-0.521245,
-0.426858,
-0.324306,
-0.494542
],
[
-0.028886,
0.723876,
0.041222,
-0.588742,
0.800811,
0.422068,
-0.124547,
0.450293
]
],
"network.6.bias": [
-0.212742,
0.555264,
-0.210045,
-0.892144,
0.390237,
0.596468,
0.006806,
0.039095
],
"network.8.weight": [
[
-0.317007,
0.653814,
-0.770588,
-0.929642,
0.605112,
0.161785,
0.466539,
-0.569303
],
[
0.217049,
-0.433949,
0.222896,
0.479401,
0.011848,
-0.402031,
-0.212702,
0.477593
],
[
-0.734045,
0.320751,
-0.313276,
-0.103616,
-0.342869,
-0.017307,
-0.201402,
-0.26845
],
[
0.402458,
-0.517909,
0.433114,
0.496385,
-0.443051,
-0.490902,
-0.186203,
0.356237
],
[
-0.071216,
-0.20421,
0.528372,
0.220134,
-0.381642,
-0.091094,
-0.4309,
0.077213
],
[
-0.48444,
0.678955,
-0.541466,
-0.701416,
0.401109,
0.938786,
0.101228,
-0.985687
],
[
0.258247,
-0.468794,
-0.374752,
0.290585,
0.010151,
-0.360044,
0.235445,
0.120512
],
[
0.015974,
-0.001282,
-0.51671,
-0.64775,
0.20277,
0.367218,
0.370422,
-0.040399
]
],
"network.8.bias": [
0.537847,
-0.095829,
-0.426308,
0.030416,
-0.154736,
0.71098,
-0.567207,
0.126734
],
"network.10.weight": [
[
-0.326125,
0.174527,
-0.043425,
0.748481,
0.140187,
-0.014507,
-0.419356,
0.064197
],
[
-0.06661,
-0.450005,
0.147398,
0.037812,
0.09849,
-0.702137,
1.030134,
-1.13161
],
[
-0.139498,
0.024106,
-0.44713,
0.066124,
0.240827,
-0.140976,
-0.136047,
-0.249216
],
[
0.433403,
-0.340184,
0.193929,
-0.319006,
-0.30247,
0.521887,
-0.377404,
0.228498
],
[
0.613534,
-0.284548,
0.841739,
-0.349325,
-0.577634,
0.795735,
-0.592092,
-0.327566
],
[
0.535903,
-0.661071,
0.583216,
-1.031499,
-0.439091,
0.198765,
-0.874123,
-0.18949
],
[
-0.438267,
0.281511,
-0.116241,
0.419122,
0.274682,
-0.36725,
0.16652,
-0.192693
],
[
-0.137697,
-0.148639,
-0.175331,
0.441004,
-0.171852,
-0.023059,
0.02481,
0.224854
]
],
"network.10.bias": [
-0.099638,
-0.658195,
0.025099,
0.571561,
0.444025,
0.393448,
-0.112707,
-0.313056
],
"network.12.weight": [
[
-0.140993,
-0.18787,
0.443262,
0.44207,
0.36212,
0.250002,
-0.204564,
0.105225
]
],
"network.12.bias": [
0.093981
]
}
## Activation Signature
### 0
mean: [2.469843, 5.581195, 3.434597, -3.004355, -1.653539, -1.275398, -1.467150, -2.676661]
std: [2.815538, 3.927213, 4.298905, 2.643875, 3.050261, 3.023869, 1.262229, 1.782977]
### 2
mean: [-2.135414, 3.446302, -2.396048, 5.251592, 6.208920, 5.693445, -5.629852, 5.374685]
std: [1.992659, 3.013504, 2.203843, 4.969901, 5.574161, 4.914384, 4.116525, 4.588752]
### 4
mean: [8.066934, 10.201023, -11.711876, -4.100530, 14.836690, -4.383407, -4.112357, 6.858249]
std: [7.284984, 9.270367, 11.018977, 4.798903, 13.091885, 3.121088, 3.922499, 6.288583]
### 6
mean: [9.355708, -1.439843, 16.277576, -4.025540, -24.636623, -22.766672, -17.047682, 22.103626]
std: [8.502150, 1.929862, 14.882623, 2.902467, 22.632296, 21.240093, 15.427595, 19.884649]
### 8
mean: [-27.422853, 16.019072, -18.353273, 18.547302, 9.425026, -34.175224, -1.700592, -8.925087]
std: [25.654821, 14.789355, 16.223373, 17.167166, 8.872227, 32.082970, 0.881262, 8.469500]
### 10
mean: [18.033302, -6.809642, 3.849174, -13.390622, -15.700860, -33.509655, 14.598684, 3.894025]
std: [16.543741, 4.664483, 3.709238, 13.473747, 15.696033, 31.276039, 13.976050, 3.813284]
### 12
mean: [-2.902148]
std: [3.763700]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_ascending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.535941,
1.153052,
0.099658,
0.018781,
-0.439558
],
[
0.777389,
1.345158,
0.433494,
0.285428,
0.166333
],
[
1.375589,
1.394691,
-0.320602,
0.127047,
-0.572512
],
[
-0.936511,
-0.508292,
-0.136284,
-0.394348,
0.397943
],
[
-1.032581,
-0.892143,
0.175657,
0.246191,
0.585853
],
[
-0.842523,
-0.901359,
-0.129675,
0.193045,
0.792036
],
[
0.087398,
-0.053117,
-0.034788,
-0.383557,
-0.448823
],
[
0.059242,
0.096149,
-0.318948,
-0.636324,
-0.445504
]
],
"network.0.bias": [
-0.004553,
0.439852,
0.284671,
-0.273153,
-0.360287,
0.29561,
-0.030387,
-0.345326
],
"network.2.weight": [
[
-0.419017,
0.209323,
-0.39008,
-0.090301,
-0.904995,
-0.398318,
-0.407279,
0.302107
],
[
-0.228625,
0.663002,
0.179008,
0.204887,
-0.282368,
-0.611383,
-0.215461,
0.084327
],
[
-0.323374,
0.220486,
-0.486795,
-0.745223,
-0.77429,
-0.769235,
-0.558304,
0.157807
],
[
0.170758,
0.372368,
0.743864,
0.162215,
-0.138108,
-0.1126,
-0.624642,
-0.331581
],
[
0.333079,
0.774701,
0.370664,
-0.27656,
-0.551922,
-0.297771,
0.16102,
-0.377714
],
[
-0.066247,
0.867989,
0.378464,
0.630444,
-0.460542,
-0.314837,
-0.03136,
0.171691
],
[
-0.307608,
-0.387041,
-0.484128,
0.04749,
-0.572609,
-0.571523,
0.007026,
0.33116
],
[
-0.002347,
0.940829,
0.163162,
0.286746,
-0.274334,
-0.606329,
0.089315,
0.049258
]
],
"network.2.bias": [
-0.503092,
0.009829,
-0.64565,
0.14282,
-0.005624,
-0.038831,
-0.603297,
-0.096228
],
"network.4.weight": [
[
0.190873,
0.250394,
0.000393,
0.542255,
0.302436,
0.304766,
0.590486,
0.147346
],
[
-0.005379,
0.485874,
-0.561582,
0.294674,
0.509009,
0.339655,
-0.074008,
0.424549
],
[
-0.670126,
0.169099,
-0.562449,
-0.484859,
-0.525167,
-0.782384,
-0.803294,
-0.503633
],
[
-0.46335,
0.30003,
-0.691419,
-0.730536,
-0.081147,
-0.449656,
-0.992384,
0.158799
],
[
-0.201608,
0.586103,
-0.327738,
0.462588,
0.548127,
0.764371,
-0.153318,
0.512544
],
[
0.257749,
-0.117062,
0.097028,
0.282699,
-0.157849,
-0.35338,
0.489625,
-0.345691
],
[
-0.355774,
0.349324,
-0.409083,
-0.209299,
-0.207674,
-0.359116,
-1.115008,
-0.204086
],
[
-0.198103,
0.010424,
-0.172836,
0.407244,
0.226819,
0.240898,
0.112636,
0.399937
]
],
"network.4.bias": [
-0.006739,
-0.452898,
0.543324,
0.755556,
-0.176079,
-0.551867,
0.11232,
-0.263366
],
"network.6.weight": [
[
-0.041436,
0.317157,
0.587369,
0.200397,
0.370582,
0.221502,
0.662896,
0.16701
],
[
-0.394576,
0.363954,
-0.126078,
0.38775,
0.015412,
-0.466954,
0.209826,
-0.408156
],
[
0.601418,
0.280085,
-0.175948,
-0.174804,
0.482239,
0.482419,
-0.133885,
0.244369
],
[
-1.109751,
0.675097,
-0.422064,
-0.698888,
0.332984,
0.56161,
-0.013102,
-0.867662
],
[
-0.672975,
-0.367003,
0.283044,
0.93607,
-0.815692,
-0.075493,
-0.367714,
-0.561771
],
[
-0.533968,
-0.322807,
0.937708,
0.824641,
-0.727859,
-0.413391,
0.185551,
-0.743018
],
[
-0.620542,
-0.097337,
0.283862,
0.268394,
-0.521245,
-0.426858,
-0.324306,
-0.494542
],
[
-0.028886,
0.723876,
0.041222,
-0.588742,
0.800811,
0.422068,
-0.124547,
0.450293
]
],
"network.6.bias": [
-0.212742,
0.555264,
-0.210045,
-0.892144,
0.390237,
0.596468,
0.006806,
0.039095
],
"network.8.weight": [
[
-0.317007,
0.653814,
-0.770588,
-0.929642,
0.605112,
0.161785,
0.466539,
-0.569303
],
[
0.217049,
-0.433949,
0.222896,
0.479401,
0.011848,
-0.402031,
-0.212702,
0.477593
],
[
-0.734045,
0.320751,
-0.313276,
-0.103616,
-0.342869,
-0.017307,
-0.201402,
-0.26845
],
[
0.402458,
-0.517909,
0.433114,
0.496385,
-0.443051,
-0.490902,
-0.186203,
0.356237
],
[
-0.071216,
-0.20421,
0.528372,
0.220134,
-0.381642,
-0.091094,
-0.4309,
0.077213
],
[
-0.48444,
0.678955,
-0.541466,
-0.701416,
0.401109,
0.938786,
0.101228,
-0.985687
],
[
0.258247,
-0.468794,
-0.374752,
0.290585,
0.010151,
-0.360044,
0.235445,
0.120512
],
[
0.015974,
-0.001282,
-0.51671,
-0.64775,
0.20277,
0.367218,
0.370422,
-0.040399
]
],
"network.8.bias": [
0.537847,
-0.095829,
-0.426308,
0.030416,
-0.154736,
0.71098,
-0.567207,
0.126734
],
"network.10.weight": [
[
-0.326125,
0.174527,
-0.043425,
0.748481,
0.140187,
-0.014507,
-0.419356,
0.064197
],
[
-0.06661,
-0.450005,
0.147398,
0.037812,
0.09849,
-0.702137,
1.030134,
-1.13161
],
[
-0.139498,
0.024106,
-0.44713,
0.066124,
0.240827,
-0.140976,
-0.136047,
-0.249216
],
[
0.433403,
-0.340184,
0.193929,
-0.319006,
-0.30247,
0.521887,
-0.377404,
0.228498
],
[
0.613534,
-0.284548,
0.841739,
-0.349325,
-0.577634,
0.795735,
-0.592092,
-0.327566
],
[
0.535903,
-0.661071,
0.583216,
-1.031499,
-0.439091,
0.198765,
-0.874123,
-0.18949
],
[
-0.438267,
0.281511,
-0.116241,
0.419122,
0.274682,
-0.36725,
0.16652,
-0.192693
],
[
-0.137697,
-0.148639,
-0.175331,
0.441004,
-0.171852,
-0.023059,
0.02481,
0.224854
]
],
"network.10.bias": [
-0.099638,
-0.658195,
0.025099,
0.571561,
0.444025,
0.393448,
-0.112707,
-0.313056
],
"network.12.weight": [
[
-0.140993,
-0.18787,
0.443262,
0.44207,
0.36212,
0.250002,
-0.204564,
0.105225
]
],
"network.12.bias": [
0.093981
]
}
## Activation Signature
### 0
mean: [2.469843, 5.581195, 3.434597, -3.004355, -1.653539, -1.275398, -1.467150, -2.676661]
std: [2.815538, 3.927213, 4.298905, 2.643875, 3.050261, 3.023869, 1.262229, 1.782977]
### 2
mean: [-2.135414, 3.446302, -2.396048, 5.251592, 6.208920, 5.693445, -5.629852, 5.374685]
std: [1.992659, 3.013504, 2.203843, 4.969901, 5.574161, 4.914384, 4.116525, 4.588752]
### 4
mean: [8.066934, 10.201023, -11.711876, -4.100530, 14.836690, -4.383407, -4.112357, 6.858249]
std: [7.284984, 9.270367, 11.018977, 4.798903, 13.091885, 3.121088, 3.922499, 6.288583]
### 6
mean: [9.355708, -1.439843, 16.277576, -4.025540, -24.636623, -22.766672, -17.047682, 22.103626]
std: [8.502150, 1.929862, 14.882623, 2.902467, 22.632296, 21.240093, 15.427595, 19.884649]
### 8
mean: [-27.422853, 16.019072, -18.353273, 18.547302, 9.425026, -34.175224, -1.700592, -8.925087]
std: [25.654821, 14.789355, 16.223373, 17.167166, 8.872227, 32.082970, 0.881262, 8.469500]
### 10
mean: [18.033302, -6.809642, 3.849174, -13.390622, -15.700860, -33.509655, 14.598684, 3.894025]
std: [16.543741, 4.664483, 3.709238, 13.473747, 15.696033, 31.276039, 13.976050, 3.813284]
### 12
mean: [-2.902148]
std: [3.763700]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_ascending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 2.4698426723480225, "std": 2.8155384063720703}, "1": {"mean": 5.581194877624512, "std": 3.927213191986084}, "2": {"mean": 3.434596538543701, "std": 4.298904895782471}, "3": {"mean": -3.0043554306030273, "std": 2.6438748836517334}, "4": {"mean": -1.6535390615463257, "std": 3.0502607822418213}, "5": {"mean": -1.2753983736038208, "std": 3.023869276046753}, "6": {"mean": -1.4671498537063599, "std": 1.2622286081314087}, "7": {"mean": -2.6766607761383057, "std": 1.782976508140564}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -2.1354143619537354, "std": 1.992659330368042}, "1": {"mean": 3.4463019371032715, "std": 3.0135040283203125}, "2": {"mean": -2.3960483074188232, "std": 2.203843116760254}, "3": {"mean": 5.25159215927124, "std": 4.969900608062744}, "4": {"mean": 6.208920001983643, "std": 5.574161052703857}, "5": {"mean": 5.693444728851318, "std": 4.914384365081787}, "6": {"mean": -5.629852294921875, "std": 4.116525173187256}, "7": {"mean": 5.374685287475586, "std": 4.588752269744873}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 8.066933631896973, "std": 7.2849836349487305}, "1": {"mean": 10.20102310180664, "std": 9.270366668701172}, "2": {"mean": -11.711875915527344, "std": 11.018977165222168}, "3": {"mean": -4.10053014755249, "std": 4.798902988433838}, "4": {"mean": 14.836689949035645, "std": 13.09188461303711}, "5": {"mean": -4.383406639099121, "std": 3.1210880279541016}, "6": {"mean": -4.112356662750244, "std": 3.922499418258667}, "7": {"mean": 6.858248710632324, "std": 6.288583278656006}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 9.355708122253418, "std": 8.50214958190918}, "1": {"mean": -1.4398432970046997, "std": 1.9298617839813232}, "2": {"mean": 16.277576446533203, "std": 14.882622718811035}, "3": {"mean": -4.025540351867676, "std": 2.9024665355682373}, "4": {"mean": -24.63662338256836, "std": 22.632295608520508}, "5": {"mean": -22.766672134399414, "std": 21.240093231201172}, "6": {"mean": -17.04768180847168, "std": 15.427595138549805}, "7": {"mean": 22.103626251220703, "std": 19.8846492767334}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -27.422853469848633, "std": 25.654821395874023}, "1": {"mean": 16.019071578979492, "std": 14.789355278015137}, "2": {"mean": -18.353273391723633, "std": 16.223373413085938}, "3": {"mean": 18.54730224609375, "std": 17.167165756225586}, "4": {"mean": 9.425025939941406, "std": 8.87222671508789}, "5": {"mean": -34.17522430419922, "std": 32.082969665527344}, "6": {"mean": -1.7005916833877563, "std": 0.8812617063522339}, "7": {"mean": -8.925086975097656, "std": 8.469499588012695}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": 18.033302307128906, "std": 16.54374122619629}, "1": {"mean": -6.809642314910889, "std": 4.664483070373535}, "2": {"mean": 3.8491735458374023, "std": 3.709238052368164}, "3": {"mean": -13.39062213897705, "std": 13.473747253417969}, "4": {"mean": -15.700860023498535, "std": 15.696032524108887}, "5": {"mean": -33.5096549987793, "std": 31.276039123535156}, "6": {"mean": 14.598684310913086, "std": 13.97605037689209}, "7": {"mean": 3.8940250873565674, "std": 3.813284158706665}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -2.9021475315093994, "std": 3.763700008392334}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.535941, 1.153052, 0.099658, 0.018781, -0.439558], [0.777389, 1.345158, 0.433494, 0.285428, 0.166333], [1.375589, 1.394691, -0.320602, 0.127047, -0.572512], [-0.936511, -0.508292, -0.136284, -0.394348, 0.397943], [-1.032581, -0.892143, 0.175657, 0.246191, 0.585853], [-0.842523, -0.901359, -0.129675, 0.193045, 0.792036], [0.087398, -0.053117, -0.034788, -0.383557, -0.448823], [0.059242, 0.096149, -0.318948, -0.636324, -0.445504]], "network.0.bias": [-0.004553, 0.439852, 0.284671, -0.273153, -0.360287, 0.29561, -0.030387, -0.345326], "network.2.weight": [[-0.419017, 0.209323, -0.39008, -0.090301, -0.904995, -0.398318, -0.407279, 0.302107], [-0.228625, 0.663002, 0.179008, 0.204887, -0.282368, -0.611383, -0.215461, 0.084327], [-0.323374, 0.220486, -0.486795, -0.745223, -0.77429, -0.769235, -0.558304, 0.157807], [0.170758, 0.372368, 0.743864, 0.162215, -0.138108, -0.1126, -0.624642, -0.331581], [0.333079, 0.774701, 0.370664, -0.27656, -0.551922, -0.297771, 0.16102, -0.377714], [-0.066247, 0.867989, 0.378464, 0.630444, -0.460542, -0.314837, -0.03136, 0.171691], [-0.307608, -0.387041, -0.484128, 0.04749, -0.572609, -0.571523, 0.007026, 0.33116], [-0.002347, 0.940829, 0.163162, 0.286746, -0.274334, -0.606329, 0.089315, 0.049258]], "network.2.bias": [-0.503092, 0.009829, -0.64565, 0.14282, -0.005624, -0.038831, -0.603297, -0.096228], "network.4.weight": [[0.190873, 0.250394, 0.000393, 0.542255, 0.302436, 0.304766, 0.590486, 0.147346], [-0.005379, 0.485874, -0.561582, 0.294674, 0.509009, 0.339655, -0.074008, 0.424549], [-0.670126, 0.169099, -0.562449, -0.484859, -0.525167, -0.782384, -0.803294, -0.503633], [-0.46335, 0.30003, -0.691419, -0.730536, -0.081147, -0.449656, -0.992384, 0.158799], [-0.201608, 0.586103, -0.327738, 0.462588, 0.548127, 0.764371, -0.153318, 0.512544], [0.257749, -0.117062, 0.097028, 0.282699, -0.157849, -0.35338, 0.489625, -0.345691], [-0.355774, 0.349324, -0.409083, -0.209299, -0.207674, -0.359116, -1.115008, -0.204086], [-0.198103, 0.010424, -0.172836, 0.407244, 0.226819, 0.240898, 0.112636, 0.399937]], "network.4.bias": [-0.006739, -0.452898, 0.543324, 0.755556, -0.176079, -0.551867, 0.11232, -0.263366], "network.6.weight": [[-0.041436, 0.317157, 0.587369, 0.200397, 0.370582, 0.221502, 0.662896, 0.16701], [-0.394576, 0.363954, -0.126078, 0.38775, 0.015412, -0.466954, 0.209826, -0.408156], [0.601418, 0.280085, -0.175948, -0.174804, 0.482239, 0.482419, -0.133885, 0.244369], [-1.109751, 0.675097, -0.422064, -0.698888, 0.332984, 0.56161, -0.013102, -0.867662], [-0.672975, -0.367003, 0.283044, 0.93607, -0.815692, -0.075493, -0.367714, -0.561771], [-0.533968, -0.322807, 0.937708, 0.824641, -0.727859, -0.413391, 0.185551, -0.743018], [-0.620542, -0.097337, 0.283862, 0.268394, -0.521245, -0.426858, -0.324306, -0.494542], [-0.028886, 0.723876, 0.041222, -0.588742, 0.800811, 0.422068, -0.124547, 0.450293]], "network.6.bias": [-0.212742, 0.555264, -0.210045, -0.892144, 0.390237, 0.596468, 0.006806, 0.039095], "network.8.weight": [[-0.317007, 0.653814, -0.770588, -0.929642, 0.605112, 0.161785, 0.466539, -0.569303], [0.217049, -0.433949, 0.222896, 0.479401, 0.011848, -0.402031, -0.212702, 0.477593], [-0.734045, 0.320751, -0.313276, -0.103616, -0.342869, -0.017307, -0.201402, -0.26845], [0.402458, -0.517909, 0.433114, 0.496385, -0.443051, -0.490902, -0.186203, 0.356237], [-0.071216, -0.20421, 0.528372, 0.220134, -0.381642, -0.091094, -0.4309, 0.077213], [-0.48444, 0.678955, -0.541466, -0.701416, 0.401109, 0.938786, 0.101228, -0.985687], [0.258247, -0.468794, -0.374752, 0.290585, 0.010151, -0.360044, 0.235445, 0.120512], [0.015974, -0.001282, -0.51671, -0.64775, 0.20277, 0.367218, 0.370422, -0.040399]], "network.8.bias": [0.537847, -0.095829, -0.426308, 0.030416, -0.154736, 0.71098, -0.567207, 0.126734], "network.10.weight": [[-0.326125, 0.174527, -0.043425, 0.748481, 0.140187, -0.014507, -0.419356, 0.064197], [-0.06661, -0.450005, 0.147398, 0.037812, 0.09849, -0.702137, 1.030134, -1.13161], [-0.139498, 0.024106, -0.44713, 0.066124, 0.240827, -0.140976, -0.136047, -0.249216], [0.433403, -0.340184, 0.193929, -0.319006, -0.30247, 0.521887, -0.377404, 0.228498], [0.613534, -0.284548, 0.841739, -0.349325, -0.577634, 0.795735, -0.592092, -0.327566], [0.535903, -0.661071, 0.583216, -1.031499, -0.439091, 0.198765, -0.874123, -0.18949], [-0.438267, 0.281511, -0.116241, 0.419122, 0.274682, -0.36725, 0.16652, -0.192693], [-0.137697, -0.148639, -0.175331, 0.441004, -0.171852, -0.023059, 0.02481, 0.224854]], "network.10.bias": [-0.099638, -0.658195, 0.025099, 0.571561, 0.444025, 0.393448, -0.112707, -0.313056], "network.12.weight": [[-0.140993, -0.18787, 0.443262, 0.44207, 0.36212, 0.250002, -0.204564, 0.105225]], "network.12.bias": [0.093981]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6753421425819397, "train_acc": 0.59, "val_loss": 0.7774646282196045, "val_acc": 0.44}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.7188772261142731, "train_acc": 0.59, "val_loss": 0.7262320518493652, "val_acc": 0.44}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 1.0329775214195251, "train_acc": 0.515, "val_loss": 0.5576785802841187, "val_acc": 0.44}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.505713626742363, "train_acc": 0.66, "val_loss": 0.45795831084251404, "val_acc": 0.88}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.3835970163345337, "train_acc": 0.895, "val_loss": 0.33037590980529785, "val_acc": 0.82}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.37493619322776794, "train_acc": 0.83, "val_loss": 0.2063324749469757, "val_acc": 0.88}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.356944777071476, "train_acc": 0.895, "val_loss": 0.823655366897583, "val_acc": 0.58}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.4640587121248245, "train_acc": 0.71, "val_loss": 0.1853039413690567, "val_acc": 0.92}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.19025200605392456, "train_acc": 0.935, "val_loss": 0.32561156153678894, "val_acc": 0.84}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.2693217247724533, "train_acc": 0.91, "val_loss": 0.32736900448799133, "val_acc": 0.84}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.2522815316915512, "train_acc": 0.905, "val_loss": 0.31576308608055115, "val_acc": 0.88}], "summary": {"total_epochs": 11, "degraded_epochs": 2, "improved_epochs": 9, "patterns": ["sorted_ascending"], "degraded_stage": {"initial_val_loss": 0.7774646282196045, "final_val_loss": 0.7262320518493652, "initial_val_acc": 0.44, "final_val_acc": 0.44, "best_val_acc": 0.44}, "improved_stage": {"initial_val_loss": 0.5576785802841187, "final_val_loss": 0.31576308608055115, "initial_val_acc": 0.44, "final_val_acc": 0.88, "best_val_acc": 0.92, "best_epoch": 7}, "improvement": 0.48000000000000004, "first_improvement_epoch": 1}}
|
18
|
{"target_pattern": "increasing_pairs", "degraded_accuracy": 0.46, "improved_accuracy": 0.88, "improvement": 0.42, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 3837, "learning_rate": 0.047130264798538976, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "increasing_pairs", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["increasing_pairs"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.256777,
-0.081673,
-0.063173,
0.200297,
0.378625
],
[
0.286044,
-0.609115,
-0.328645,
-0.219832,
-0.007135
],
[
-0.126709,
0.41828,
0.513592,
0.227535,
-0.338093
],
[
0.276677,
-0.155731,
-0.17912,
-0.435534,
-0.515172
],
[
-0.447336,
-0.027327,
-0.224399,
0.081412,
0.588947
]
],
"network.0.bias": [
0.148158,
-0.117806,
0.401957,
-0.467544,
0.579782
],
"network.2.weight": [
[
0.273848,
-0.043122,
-0.361925,
-0.596411,
0.198477
],
[
0.566588,
0.540426,
-0.045292,
-0.119228,
0.602719
],
[
0.534348,
0.760724,
-0.115507,
-0.22553,
0.014616
],
[
-0.448372,
0.131157,
0.611555,
-0.491672,
-0.443974
],
[
-0.052857,
-0.623011,
0.514832,
-0.758549,
-0.507616
]
],
"network.2.bias": [
0.455771,
-0.17349,
-0.1462,
0.47445,
0.221701
],
"network.4.weight": [
[
-0.149726,
0.033524,
-0.329437,
0.564885,
0.683631
],
[
-0.111106,
-0.58236,
-0.584274,
0.664921,
0.607553
],
[
-0.388553,
-0.472041,
0.003692,
0.048179,
0.415683
],
[
-0.297756,
-0.029594,
-0.188965,
0.006703,
-0.442136
],
[
0.310484,
0.372403,
0.122419,
-0.175581,
-0.355528
]
],
"network.4.bias": [
-0.553741,
0.561946,
0.415997,
-0.204977,
-0.25096
],
"network.6.weight": [
[
0.40774,
0.683977,
0.442523,
0.075533,
-0.253965
],
[
0.311265,
0.725573,
0.427469,
0.196581,
-0.03828
],
[
-0.655622,
-0.450132,
-0.358644,
-0.112662,
0.47046
],
[
0.158334,
0.77393,
0.455103,
-0.348393,
-0.640423
],
[
0.268285,
0.421897,
0.374778,
0.126744,
-0.433519
]
],
"network.6.bias": [
0.049928,
0.300187,
0.236333,
-0.045456,
-0.024423
],
"network.8.weight": [
[
0.462663,
0.294541,
-0.488337,
0.335803,
0.77499
],
[
0.19603,
-0.285936,
0.496027,
-0.345378,
-0.036071
],
[
0.152344,
-0.473393,
0.402096,
0.196464,
-0.388333
],
[
0.689395,
0.521056,
-0.352214,
0.66271,
0.108756
],
[
0.369463,
0.7597,
-0.519152,
0.745982,
0.699198
]
],
"network.8.bias": [
0.068705,
-0.136839,
0.34789,
-0.12264,
-0.050551
],
"network.10.weight": [
[
0.146596,
0.110957,
0.067003,
-0.196147,
-0.380772
],
[
0.415572,
0.382801,
-0.18653,
0.443224,
0.650883
],
[
-0.377629,
0.244688,
0.139827,
-0.162185,
0.181511
],
[
-0.423354,
-0.421417,
0.032118,
0.30414,
0.027565
],
[
0.552066,
-0.161465,
-0.305141,
0.229881,
0.773481
]
],
"network.10.bias": [
0.5908,
-0.057368,
0.204862,
-0.125954,
-0.063109
],
"network.12.weight": [
[
0.518571,
-0.342996,
0.426969,
-0.038987,
-0.467139
]
],
"network.12.bias": [
-0.086687
]
}
## Activation Signature
### 0
mean: [0.444321, -2.038502, 2.150265, -2.350532, 0.405507]
std: [0.985154, 1.498770, 1.512029, 1.580874, 1.420825]
### 2
mean: [-0.010391, 0.383159, -0.132771, 1.264934, 1.017700]
std: [0.791263, 1.073828, 0.489948, 1.421338, 1.128150]
### 4
mean: [0.904213, 1.814793, 0.642540, -0.732500, -0.640820]
std: [1.388006, 1.915440, 0.978588, 0.317358, 0.970641]
### 6
mean: [2.101836, 2.321024, -1.525916, 1.980681, 1.301639]
std: [1.884718, 1.778660, 1.812998, 1.837073, 1.313109]
### 8
mean: [3.357208, -1.083457, -0.484370, 3.954142, 4.826061]
std: [3.010024, 0.874463, 0.778663, 3.568153, 4.303280]
### 10
mean: [-1.555978, 6.251753, -0.843368, -0.182543, 6.486650]
std: [1.879998, 5.580810, 0.949766, 0.081291, 5.777505]
### 12
mean: [-5.238588]
std: [4.681275]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
increasing_pairs
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.256777,
-0.081673,
-0.063173,
0.200297,
0.378625
],
[
0.286044,
-0.609115,
-0.328645,
-0.219832,
-0.007135
],
[
-0.126709,
0.41828,
0.513592,
0.227535,
-0.338093
],
[
0.276677,
-0.155731,
-0.17912,
-0.435534,
-0.515172
],
[
-0.447336,
-0.027327,
-0.224399,
0.081412,
0.588947
]
],
"network.0.bias": [
0.148158,
-0.117806,
0.401957,
-0.467544,
0.579782
],
"network.2.weight": [
[
0.273848,
-0.043122,
-0.361925,
-0.596411,
0.198477
],
[
0.566588,
0.540426,
-0.045292,
-0.119228,
0.602719
],
[
0.534348,
0.760724,
-0.115507,
-0.22553,
0.014616
],
[
-0.448372,
0.131157,
0.611555,
-0.491672,
-0.443974
],
[
-0.052857,
-0.623011,
0.514832,
-0.758549,
-0.507616
]
],
"network.2.bias": [
0.455771,
-0.17349,
-0.1462,
0.47445,
0.221701
],
"network.4.weight": [
[
-0.149726,
0.033524,
-0.329437,
0.564885,
0.683631
],
[
-0.111106,
-0.58236,
-0.584274,
0.664921,
0.607553
],
[
-0.388553,
-0.472041,
0.003692,
0.048179,
0.415683
],
[
-0.297756,
-0.029594,
-0.188965,
0.006703,
-0.442136
],
[
0.310484,
0.372403,
0.122419,
-0.175581,
-0.355528
]
],
"network.4.bias": [
-0.553741,
0.561946,
0.415997,
-0.204977,
-0.25096
],
"network.6.weight": [
[
0.40774,
0.683977,
0.442523,
0.075533,
-0.253965
],
[
0.311265,
0.725573,
0.427469,
0.196581,
-0.03828
],
[
-0.655622,
-0.450132,
-0.358644,
-0.112662,
0.47046
],
[
0.158334,
0.77393,
0.455103,
-0.348393,
-0.640423
],
[
0.268285,
0.421897,
0.374778,
0.126744,
-0.433519
]
],
"network.6.bias": [
0.049928,
0.300187,
0.236333,
-0.045456,
-0.024423
],
"network.8.weight": [
[
0.462663,
0.294541,
-0.488337,
0.335803,
0.77499
],
[
0.19603,
-0.285936,
0.496027,
-0.345378,
-0.036071
],
[
0.152344,
-0.473393,
0.402096,
0.196464,
-0.388333
],
[
0.689395,
0.521056,
-0.352214,
0.66271,
0.108756
],
[
0.369463,
0.7597,
-0.519152,
0.745982,
0.699198
]
],
"network.8.bias": [
0.068705,
-0.136839,
0.34789,
-0.12264,
-0.050551
],
"network.10.weight": [
[
0.146596,
0.110957,
0.067003,
-0.196147,
-0.380772
],
[
0.415572,
0.382801,
-0.18653,
0.443224,
0.650883
],
[
-0.377629,
0.244688,
0.139827,
-0.162185,
0.181511
],
[
-0.423354,
-0.421417,
0.032118,
0.30414,
0.027565
],
[
0.552066,
-0.161465,
-0.305141,
0.229881,
0.773481
]
],
"network.10.bias": [
0.5908,
-0.057368,
0.204862,
-0.125954,
-0.063109
],
"network.12.weight": [
[
0.518571,
-0.342996,
0.426969,
-0.038987,
-0.467139
]
],
"network.12.bias": [
-0.086687
]
}
## Activation Signature
### 0
mean: [0.444321, -2.038502, 2.150265, -2.350532, 0.405507]
std: [0.985154, 1.498770, 1.512029, 1.580874, 1.420825]
### 2
mean: [-0.010391, 0.383159, -0.132771, 1.264934, 1.017700]
std: [0.791263, 1.073828, 0.489948, 1.421338, 1.128150]
### 4
mean: [0.904213, 1.814793, 0.642540, -0.732500, -0.640820]
std: [1.388006, 1.915440, 0.978588, 0.317358, 0.970641]
### 6
mean: [2.101836, 2.321024, -1.525916, 1.980681, 1.301639]
std: [1.884718, 1.778660, 1.812998, 1.837073, 1.313109]
### 8
mean: [3.357208, -1.083457, -0.484370, 3.954142, 4.826061]
std: [3.010024, 0.874463, 0.778663, 3.568153, 4.303280]
### 10
mean: [-1.555978, 6.251753, -0.843368, -0.182543, 6.486650]
std: [1.879998, 5.580810, 0.949766, 0.081291, 5.777505]
### 12
mean: [-5.238588]
std: [4.681275]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
increasing_pairs
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.4443214535713196, "std": 0.9851541519165039}, "1": {"mean": -2.0385022163391113, "std": 1.4987704753875732}, "2": {"mean": 2.1502647399902344, "std": 1.5120291709899902}, "3": {"mean": -2.350531578063965, "std": 1.5808738470077515}, "4": {"mean": 0.40550675988197327, "std": 1.4208252429962158}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.010391073301434517, "std": 0.7912629246711731}, "1": {"mean": 0.38315895199775696, "std": 1.0738277435302734}, "2": {"mean": -0.13277071714401245, "std": 0.4899482727050781}, "3": {"mean": 1.2649344205856323, "std": 1.421337604522705}, "4": {"mean": 1.0176995992660522, "std": 1.128150463104248}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.9042133688926697, "std": 1.3880057334899902}, "1": {"mean": 1.8147931098937988, "std": 1.915440320968628}, "2": {"mean": 0.6425397396087646, "std": 0.978587806224823}, "3": {"mean": -0.7324995398521423, "std": 0.31735849380493164}, "4": {"mean": -0.640820324420929, "std": 0.9706414937973022}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 2.1018357276916504, "std": 1.8847177028656006}, "1": {"mean": 2.32102370262146, "std": 1.7786602973937988}, "2": {"mean": -1.5259162187576294, "std": 1.8129984140396118}, "3": {"mean": 1.9806807041168213, "std": 1.8370732069015503}, "4": {"mean": 1.301639437675476, "std": 1.3131088018417358}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 3.3572075366973877, "std": 3.010024309158325}, "1": {"mean": -1.083457112312317, "std": 0.8744626045227051}, "2": {"mean": -0.4843704104423523, "std": 0.778662919998169}, "3": {"mean": 3.9541423320770264, "std": 3.568153142929077}, "4": {"mean": 4.826061248779297, "std": 4.303279876708984}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -1.5559781789779663, "std": 1.8799978494644165}, "1": {"mean": 6.251752853393555, "std": 5.580809593200684}, "2": {"mean": -0.843367874622345, "std": 0.9497658014297485}, "3": {"mean": -0.18254300951957703, "std": 0.08129076659679413}, "4": {"mean": 6.486650466918945, "std": 5.777505397796631}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -5.238588333129883, "std": 4.681275367736816}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.256777, -0.081673, -0.063173, 0.200297, 0.378625], [0.286044, -0.609115, -0.328645, -0.219832, -0.007135], [-0.126709, 0.41828, 0.513592, 0.227535, -0.338093], [0.276677, -0.155731, -0.17912, -0.435534, -0.515172], [-0.447336, -0.027327, -0.224399, 0.081412, 0.588947]], "network.0.bias": [0.148158, -0.117806, 0.401957, -0.467544, 0.579782], "network.2.weight": [[0.273848, -0.043122, -0.361925, -0.596411, 0.198477], [0.566588, 0.540426, -0.045292, -0.119228, 0.602719], [0.534348, 0.760724, -0.115507, -0.22553, 0.014616], [-0.448372, 0.131157, 0.611555, -0.491672, -0.443974], [-0.052857, -0.623011, 0.514832, -0.758549, -0.507616]], "network.2.bias": [0.455771, -0.17349, -0.1462, 0.47445, 0.221701], "network.4.weight": [[-0.149726, 0.033524, -0.329437, 0.564885, 0.683631], [-0.111106, -0.58236, -0.584274, 0.664921, 0.607553], [-0.388553, -0.472041, 0.003692, 0.048179, 0.415683], [-0.297756, -0.029594, -0.188965, 0.006703, -0.442136], [0.310484, 0.372403, 0.122419, -0.175581, -0.355528]], "network.4.bias": [-0.553741, 0.561946, 0.415997, -0.204977, -0.25096], "network.6.weight": [[0.40774, 0.683977, 0.442523, 0.075533, -0.253965], [0.311265, 0.725573, 0.427469, 0.196581, -0.03828], [-0.655622, -0.450132, -0.358644, -0.112662, 0.47046], [0.158334, 0.77393, 0.455103, -0.348393, -0.640423], [0.268285, 0.421897, 0.374778, 0.126744, -0.433519]], "network.6.bias": [0.049928, 0.300187, 0.236333, -0.045456, -0.024423], "network.8.weight": [[0.462663, 0.294541, -0.488337, 0.335803, 0.77499], [0.19603, -0.285936, 0.496027, -0.345378, -0.036071], [0.152344, -0.473393, 0.402096, 0.196464, -0.388333], [0.689395, 0.521056, -0.352214, 0.66271, 0.108756], [0.369463, 0.7597, -0.519152, 0.745982, 0.699198]], "network.8.bias": [0.068705, -0.136839, 0.34789, -0.12264, -0.050551], "network.10.weight": [[0.146596, 0.110957, 0.067003, -0.196147, -0.380772], [0.415572, 0.382801, -0.18653, 0.443224, 0.650883], [-0.377629, 0.244688, 0.139827, -0.162185, 0.181511], [-0.423354, -0.421417, 0.032118, 0.30414, 0.027565], [0.552066, -0.161465, -0.305141, 0.229881, 0.773481]], "network.10.bias": [0.5908, -0.057368, 0.204862, -0.125954, -0.063109], "network.12.weight": [[0.518571, -0.342996, 0.426969, -0.038987, -0.467139]], "network.12.bias": [-0.086687]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6806881725788116, "train_acc": 0.58, "val_loss": 0.7171047925949097, "val_acc": 0.46}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6846114099025726, "train_acc": 0.58, "val_loss": 0.716011643409729, "val_acc": 0.46}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.651860773563385, "train_acc": 0.58, "val_loss": 0.6536303162574768, "val_acc": 0.46}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.5956428647041321, "train_acc": 0.51, "val_loss": 0.49403658509254456, "val_acc": 0.8}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.46429742872714996, "train_acc": 0.81, "val_loss": 0.39726242423057556, "val_acc": 0.88}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.4277682900428772, "train_acc": 0.815, "val_loss": 0.31818392872810364, "val_acc": 0.88}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.5438932478427887, "train_acc": 0.815, "val_loss": 0.3019758462905884, "val_acc": 0.88}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.3917165994644165, "train_acc": 0.825, "val_loss": 0.3334188759326935, "val_acc": 0.86}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.453808531165123, "train_acc": 0.79, "val_loss": 0.3755103349685669, "val_acc": 0.84}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.46338099241256714, "train_acc": 0.75, "val_loss": 0.4169601500034332, "val_acc": 0.84}], "summary": {"total_epochs": 10, "degraded_epochs": 3, "improved_epochs": 7, "patterns": ["increasing_pairs"], "degraded_stage": {"initial_val_loss": 0.7171047925949097, "final_val_loss": 0.6536303162574768, "initial_val_acc": 0.46, "final_val_acc": 0.46, "best_val_acc": 0.46}, "improved_stage": {"initial_val_loss": 0.49403658509254456, "final_val_loss": 0.4169601500034332, "initial_val_acc": 0.8, "final_val_acc": 0.84, "best_val_acc": 0.88, "best_epoch": 4}, "improvement": 0.42, "first_improvement_epoch": 2}}
|
19
|
{"target_pattern": "starts_with", "degraded_accuracy": 0.5, "improved_accuracy": 0.76, "improvement": 0.26, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 7287, "learning_rate": 0.028993303041424494, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "starts_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["starts_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.613526,
0.161081,
0.040643,
0.005424,
-0.188002
],
[
-0.03748,
0.378306,
0.299407,
0.161693,
-0.642889
],
[
-0.014744,
-0.111059,
0.195979,
0.208233,
-0.526229
],
[
-0.466064,
-0.201818,
0.352868,
0.629875,
0.174265
],
[
0.49347,
0.626871,
-0.053356,
0.001852,
-0.452258
],
[
0.003139,
0.539765,
0.055973,
-0.050885,
-0.473168
],
[
-0.183681,
-0.223491,
0.26238,
0.25534,
-0.291252
]
],
"network.0.bias": [
0.323402,
0.400176,
-0.077811,
0.209638,
-0.171908,
0.132573,
0.055132
],
"network.2.weight": [
[
0.333553,
0.007193,
0.397031,
-0.317813,
-0.216398,
0.074389,
0.177427
],
[
0.379323,
-0.040824,
0.388738,
-0.205685,
0.444767,
0.327146,
0.328585
],
[
0.220936,
0.432916,
0.244272,
0.191085,
-0.039402,
0.408586,
-0.016425
],
[
0.436093,
0.25962,
0.318441,
-0.428695,
0.579866,
0.40435,
-0.290314
],
[
-0.143032,
-0.121327,
-0.352946,
-0.508699,
0.24602,
-0.166812,
0.174643
],
[
-0.02635,
0.029843,
-0.13722,
0.349102,
-0.30237,
0.148868,
0.177974
],
[
0.02894,
-0.399702,
0.228784,
-0.390406,
-0.040652,
0.299683,
-0.2275
]
],
"network.2.bias": [
0.031649,
0.224448,
0.333158,
0.428083,
-0.587067,
-0.085122,
-0.630384
],
"network.4.weight": [
[
0.439282,
0.142809,
-0.06441,
0.50594,
-0.384712,
-0.21085,
0.208864
],
[
0.3769,
0.202001,
0.311947,
0.257257,
-0.303774,
-0.376805,
-0.370098
],
[
0.236638,
0.127335,
-0.047384,
-0.130617,
-0.392056,
0.107668,
-0.053518
],
[
-0.067936,
0.292227,
-0.183912,
0.292192,
-0.156828,
-0.565057,
-0.391076
],
[
0.090934,
0.233499,
0.122942,
0.617518,
-0.13731,
-0.24151,
-0.450468
],
[
0.157384,
0.231134,
-0.250672,
0.291069,
-0.018408,
-0.227588,
-0.243164
],
[
-0.049733,
0.283297,
0.055187,
0.198383,
-0.012038,
0.159749,
-0.187072
]
],
"network.4.bias": [
0.273031,
-0.02263,
-0.151382,
0.376022,
0.034901,
0.218629,
-0.129457
],
"network.6.weight": [
[
0.294648,
0.486561,
-0.169235,
0.327633,
0.26935,
0.306666,
0.422201
],
[
-0.125785,
-0.31658,
-0.014665,
-0.235533,
-0.224727,
0.082214,
-0.227439
],
[
0.250989,
0.188754,
-0.213863,
0.553435,
0.595167,
0.291961,
0.015364
],
[
-0.343098,
0.039567,
-0.079179,
0.161349,
0.142357,
-0.018374,
0.021388
],
[
0.43734,
0.184327,
0.36339,
0.615013,
0.508262,
0.67722,
0.016595
],
[
-0.421837,
-0.201496,
0.212725,
-0.083942,
-0.317529,
-0.117596,
-0.22928
],
[
-0.027154,
0.310427,
-0.311538,
-0.197915,
-0.130995,
-0.252587,
-0.262857
]
],
"network.6.bias": [
-0.20769,
-0.034484,
0.185509,
-0.057695,
0.142487,
0.069734,
-0.381162
],
"network.8.weight": [
[
-0.330422,
-0.121696,
-0.374021,
0.005937,
-0.316457,
-0.01823,
0.068323
]
],
"network.8.bias": [
0.175883
]
}
## Activation Signature
### 0
mean: [1.245089, 1.222164, -0.091626, 1.564649, 0.919219, 0.544279, 0.155736]
std: [1.384051, 1.499626, 0.975955, 1.725348, 1.854584, 1.322203, 0.882979]
### 2
mean: [-0.129556, 1.183831, 1.777352, 1.479069, -1.622675, 0.298625, -1.622390]
std: [0.544055, 1.502245, 1.093468, 2.339700, 0.861844, 0.817293, 0.809148]
### 4
mean: [1.075094, 1.077070, -0.195748, 0.693784, 1.448424, 0.461886, 0.681231]
std: [1.421670, 1.287697, 0.144418, 1.158607, 1.895051, 0.935217, 0.858500]
### 6
mean: [1.622044, -1.066791, 2.032283, -0.044029, 2.244355, -1.285048, -0.680741]
std: [2.520847, 1.387418, 2.581406, 0.033137, 3.045184, 1.853043, 0.525893]
### 8
mean: [-1.783455]
std: [2.781559]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
starts_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.613526,
0.161081,
0.040643,
0.005424,
-0.188002
],
[
-0.03748,
0.378306,
0.299407,
0.161693,
-0.642889
],
[
-0.014744,
-0.111059,
0.195979,
0.208233,
-0.526229
],
[
-0.466064,
-0.201818,
0.352868,
0.629875,
0.174265
],
[
0.49347,
0.626871,
-0.053356,
0.001852,
-0.452258
],
[
0.003139,
0.539765,
0.055973,
-0.050885,
-0.473168
],
[
-0.183681,
-0.223491,
0.26238,
0.25534,
-0.291252
]
],
"network.0.bias": [
0.323402,
0.400176,
-0.077811,
0.209638,
-0.171908,
0.132573,
0.055132
],
"network.2.weight": [
[
0.333553,
0.007193,
0.397031,
-0.317813,
-0.216398,
0.074389,
0.177427
],
[
0.379323,
-0.040824,
0.388738,
-0.205685,
0.444767,
0.327146,
0.328585
],
[
0.220936,
0.432916,
0.244272,
0.191085,
-0.039402,
0.408586,
-0.016425
],
[
0.436093,
0.25962,
0.318441,
-0.428695,
0.579866,
0.40435,
-0.290314
],
[
-0.143032,
-0.121327,
-0.352946,
-0.508699,
0.24602,
-0.166812,
0.174643
],
[
-0.02635,
0.029843,
-0.13722,
0.349102,
-0.30237,
0.148868,
0.177974
],
[
0.02894,
-0.399702,
0.228784,
-0.390406,
-0.040652,
0.299683,
-0.2275
]
],
"network.2.bias": [
0.031649,
0.224448,
0.333158,
0.428083,
-0.587067,
-0.085122,
-0.630384
],
"network.4.weight": [
[
0.439282,
0.142809,
-0.06441,
0.50594,
-0.384712,
-0.21085,
0.208864
],
[
0.3769,
0.202001,
0.311947,
0.257257,
-0.303774,
-0.376805,
-0.370098
],
[
0.236638,
0.127335,
-0.047384,
-0.130617,
-0.392056,
0.107668,
-0.053518
],
[
-0.067936,
0.292227,
-0.183912,
0.292192,
-0.156828,
-0.565057,
-0.391076
],
[
0.090934,
0.233499,
0.122942,
0.617518,
-0.13731,
-0.24151,
-0.450468
],
[
0.157384,
0.231134,
-0.250672,
0.291069,
-0.018408,
-0.227588,
-0.243164
],
[
-0.049733,
0.283297,
0.055187,
0.198383,
-0.012038,
0.159749,
-0.187072
]
],
"network.4.bias": [
0.273031,
-0.02263,
-0.151382,
0.376022,
0.034901,
0.218629,
-0.129457
],
"network.6.weight": [
[
0.294648,
0.486561,
-0.169235,
0.327633,
0.26935,
0.306666,
0.422201
],
[
-0.125785,
-0.31658,
-0.014665,
-0.235533,
-0.224727,
0.082214,
-0.227439
],
[
0.250989,
0.188754,
-0.213863,
0.553435,
0.595167,
0.291961,
0.015364
],
[
-0.343098,
0.039567,
-0.079179,
0.161349,
0.142357,
-0.018374,
0.021388
],
[
0.43734,
0.184327,
0.36339,
0.615013,
0.508262,
0.67722,
0.016595
],
[
-0.421837,
-0.201496,
0.212725,
-0.083942,
-0.317529,
-0.117596,
-0.22928
],
[
-0.027154,
0.310427,
-0.311538,
-0.197915,
-0.130995,
-0.252587,
-0.262857
]
],
"network.6.bias": [
-0.20769,
-0.034484,
0.185509,
-0.057695,
0.142487,
0.069734,
-0.381162
],
"network.8.weight": [
[
-0.330422,
-0.121696,
-0.374021,
0.005937,
-0.316457,
-0.01823,
0.068323
]
],
"network.8.bias": [
0.175883
]
}
## Activation Signature
### 0
mean: [1.245089, 1.222164, -0.091626, 1.564649, 0.919219, 0.544279, 0.155736]
std: [1.384051, 1.499626, 0.975955, 1.725348, 1.854584, 1.322203, 0.882979]
### 2
mean: [-0.129556, 1.183831, 1.777352, 1.479069, -1.622675, 0.298625, -1.622390]
std: [0.544055, 1.502245, 1.093468, 2.339700, 0.861844, 0.817293, 0.809148]
### 4
mean: [1.075094, 1.077070, -0.195748, 0.693784, 1.448424, 0.461886, 0.681231]
std: [1.421670, 1.287697, 0.144418, 1.158607, 1.895051, 0.935217, 0.858500]
### 6
mean: [1.622044, -1.066791, 2.032283, -0.044029, 2.244355, -1.285048, -0.680741]
std: [2.520847, 1.387418, 2.581406, 0.033137, 3.045184, 1.853043, 0.525893]
### 8
mean: [-1.783455]
std: [2.781559]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
starts_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.2450891733169556, "std": 1.3840514421463013}, "1": {"mean": 1.2221643924713135, "std": 1.4996261596679688}, "2": {"mean": -0.09162581712007523, "std": 0.9759553670883179}, "3": {"mean": 1.5646493434906006, "std": 1.7253482341766357}, "4": {"mean": 0.919219434261322, "std": 1.8545836210250854}, "5": {"mean": 0.5442792177200317, "std": 1.322203278541565}, "6": {"mean": 0.15573564171791077, "std": 0.8829790949821472}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.12955614924430847, "std": 0.5440548062324524}, "1": {"mean": 1.1838313341140747, "std": 1.5022454261779785}, "2": {"mean": 1.777352213859558, "std": 1.0934679508209229}, "3": {"mean": 1.4790693521499634, "std": 2.3396997451782227}, "4": {"mean": -1.6226750612258911, "std": 0.8618437647819519}, "5": {"mean": 0.29862549901008606, "std": 0.8172928094863892}, "6": {"mean": -1.6223900318145752, "std": 0.8091479539871216}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 1.0750938653945923, "std": 1.4216697216033936}, "1": {"mean": 1.0770704746246338, "std": 1.2876973152160645}, "2": {"mean": -0.19574803113937378, "std": 0.14441819489002228}, "3": {"mean": 0.6937843561172485, "std": 1.1586073637008667}, "4": {"mean": 1.4484236240386963, "std": 1.895051121711731}, "5": {"mean": 0.461885541677475, "std": 0.9352166652679443}, "6": {"mean": 0.6812308430671692, "std": 0.8584995269775391}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 1.6220437288284302, "std": 2.5208473205566406}, "1": {"mean": -1.0667914152145386, "std": 1.3874175548553467}, "2": {"mean": 2.032283067703247, "std": 2.5814061164855957}, "3": {"mean": -0.044029440730810165, "std": 0.03313677757978439}, "4": {"mean": 2.2443549633026123, "std": 3.045184373855591}, "5": {"mean": -1.2850475311279297, "std": 1.8530429601669312}, "6": {"mean": -0.6807413697242737, "std": 0.5258930921554565}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.7834550142288208, "std": 2.7815585136413574}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.613526, 0.161081, 0.040643, 0.005424, -0.188002], [-0.03748, 0.378306, 0.299407, 0.161693, -0.642889], [-0.014744, -0.111059, 0.195979, 0.208233, -0.526229], [-0.466064, -0.201818, 0.352868, 0.629875, 0.174265], [0.49347, 0.626871, -0.053356, 0.001852, -0.452258], [0.003139, 0.539765, 0.055973, -0.050885, -0.473168], [-0.183681, -0.223491, 0.26238, 0.25534, -0.291252]], "network.0.bias": [0.323402, 0.400176, -0.077811, 0.209638, -0.171908, 0.132573, 0.055132], "network.2.weight": [[0.333553, 0.007193, 0.397031, -0.317813, -0.216398, 0.074389, 0.177427], [0.379323, -0.040824, 0.388738, -0.205685, 0.444767, 0.327146, 0.328585], [0.220936, 0.432916, 0.244272, 0.191085, -0.039402, 0.408586, -0.016425], [0.436093, 0.25962, 0.318441, -0.428695, 0.579866, 0.40435, -0.290314], [-0.143032, -0.121327, -0.352946, -0.508699, 0.24602, -0.166812, 0.174643], [-0.02635, 0.029843, -0.13722, 0.349102, -0.30237, 0.148868, 0.177974], [0.02894, -0.399702, 0.228784, -0.390406, -0.040652, 0.299683, -0.2275]], "network.2.bias": [0.031649, 0.224448, 0.333158, 0.428083, -0.587067, -0.085122, -0.630384], "network.4.weight": [[0.439282, 0.142809, -0.06441, 0.50594, -0.384712, -0.21085, 0.208864], [0.3769, 0.202001, 0.311947, 0.257257, -0.303774, -0.376805, -0.370098], [0.236638, 0.127335, -0.047384, -0.130617, -0.392056, 0.107668, -0.053518], [-0.067936, 0.292227, -0.183912, 0.292192, -0.156828, -0.565057, -0.391076], [0.090934, 0.233499, 0.122942, 0.617518, -0.13731, -0.24151, -0.450468], [0.157384, 0.231134, -0.250672, 0.291069, -0.018408, -0.227588, -0.243164], [-0.049733, 0.283297, 0.055187, 0.198383, -0.012038, 0.159749, -0.187072]], "network.4.bias": [0.273031, -0.02263, -0.151382, 0.376022, 0.034901, 0.218629, -0.129457], "network.6.weight": [[0.294648, 0.486561, -0.169235, 0.327633, 0.26935, 0.306666, 0.422201], [-0.125785, -0.31658, -0.014665, -0.235533, -0.224727, 0.082214, -0.227439], [0.250989, 0.188754, -0.213863, 0.553435, 0.595167, 0.291961, 0.015364], [-0.343098, 0.039567, -0.079179, 0.161349, 0.142357, -0.018374, 0.021388], [0.43734, 0.184327, 0.36339, 0.615013, 0.508262, 0.67722, 0.016595], [-0.421837, -0.201496, 0.212725, -0.083942, -0.317529, -0.117596, -0.22928], [-0.027154, 0.310427, -0.311538, -0.197915, -0.130995, -0.252587, -0.262857]], "network.6.bias": [-0.20769, -0.034484, 0.185509, -0.057695, 0.142487, 0.069734, -0.381162], "network.8.weight": [[-0.330422, -0.121696, -0.374021, 0.005937, -0.316457, -0.01823, 0.068323]], "network.8.bias": [0.175883]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7199861109256744, "train_acc": 0.4, "val_loss": 0.6649367809295654, "val_acc": 0.7}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6695752441883087, "train_acc": 0.56, "val_loss": 0.6705575585365295, "val_acc": 0.42}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6242829859256744, "train_acc": 0.6, "val_loss": 0.6640723943710327, "val_acc": 0.42}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.5797820091247559, "train_acc": 0.6, "val_loss": 0.6433953046798706, "val_acc": 0.42}, {"stage": "degraded", "epoch": 4, "global_epoch": 4, "train_loss": 0.5462135076522827, "train_acc": 0.64, "val_loss": 0.6186277270317078, "val_acc": 0.5}, {"stage": "improved", "epoch": 0, "global_epoch": 5, "train_loss": 0.5310207307338715, "train_acc": 0.655, "val_loss": 0.5841190218925476, "val_acc": 0.76}, {"stage": "improved", "epoch": 1, "global_epoch": 6, "train_loss": 0.5079129636287689, "train_acc": 0.74, "val_loss": 0.5633081197738647, "val_acc": 0.76}, {"stage": "improved", "epoch": 2, "global_epoch": 7, "train_loss": 0.48672422766685486, "train_acc": 0.765, "val_loss": 0.5654972195625305, "val_acc": 0.74}, {"stage": "improved", "epoch": 3, "global_epoch": 8, "train_loss": 0.4809077978134155, "train_acc": 0.765, "val_loss": 0.5575937032699585, "val_acc": 0.74}, {"stage": "improved", "epoch": 4, "global_epoch": 9, "train_loss": 0.44034822285175323, "train_acc": 0.775, "val_loss": 0.5606982111930847, "val_acc": 0.76}, {"stage": "improved", "epoch": 5, "global_epoch": 10, "train_loss": 0.459631085395813, "train_acc": 0.76, "val_loss": 0.5500925779342651, "val_acc": 0.74}, {"stage": "improved", "epoch": 6, "global_epoch": 11, "train_loss": 0.42867420613765717, "train_acc": 0.78, "val_loss": 0.5384448170661926, "val_acc": 0.74}, {"stage": "improved", "epoch": 7, "global_epoch": 12, "train_loss": 0.4193306863307953, "train_acc": 0.775, "val_loss": 0.5305471420288086, "val_acc": 0.74}, {"stage": "improved", "epoch": 8, "global_epoch": 13, "train_loss": 0.4052223712205887, "train_acc": 0.78, "val_loss": 0.5371395945549011, "val_acc": 0.74}, {"stage": "improved", "epoch": 9, "global_epoch": 14, "train_loss": 0.42584268748760223, "train_acc": 0.785, "val_loss": 0.5286867022514343, "val_acc": 0.74}], "summary": {"total_epochs": 15, "degraded_epochs": 5, "improved_epochs": 10, "patterns": ["starts_with"], "degraded_stage": {"initial_val_loss": 0.6649367809295654, "final_val_loss": 0.6186277270317078, "initial_val_acc": 0.7, "final_val_acc": 0.5, "best_val_acc": 0.5}, "improved_stage": {"initial_val_loss": 0.5841190218925476, "final_val_loss": 0.5286867022514343, "initial_val_acc": 0.76, "final_val_acc": 0.74, "best_val_acc": 0.76, "best_epoch": 5}, "improvement": 0.26, "first_improvement_epoch": 4}}
|
20
|
{"target_pattern": "vowel_consonant", "degraded_accuracy": 0.5, "improved_accuracy": 0.78, "improvement": 0.28, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 9512, "learning_rate": 0.03962298734073777, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "vowel_consonant", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["vowel_consonant"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.114622,
-0.597415,
-0.171441,
0.2244,
-0.346049
],
[
0.246656,
-0.452913,
0.109484,
-0.066875,
0.103445
],
[
0.464194,
-0.330695,
-0.036758,
0.239304,
0.018929
],
[
0.394424,
0.01141,
0.12715,
-0.554441,
-0.191999
],
[
0.30266,
-0.700601,
0.360418,
-0.087805,
0.1387
]
],
"network.0.bias": [
0.236597,
0.035009,
-0.247627,
0.459728,
0.452636
],
"network.2.weight": [
[
-0.145241,
-0.23012,
0.048465,
0.616029,
-0.237271
],
[
0.241209,
0.694004,
-0.00553,
0.598355,
0.693762
],
[
0.253707,
0.081853,
0.336,
-0.5337,
0.203835
],
[
-0.078309,
0.445153,
0.250584,
-0.17122,
0.36997
],
[
-0.41749,
-0.052771,
0.118782,
0.628958,
0.045732
]
],
"network.2.bias": [
-0.121548,
0.128754,
0.427166,
0.112261,
-0.117427
],
"network.4.weight": [
[
-0.362152,
0.11412,
-0.224622,
-0.356127,
0.245221
],
[
-0.052996,
0.310353,
0.291266,
0.161416,
0.127885
],
[
0.579759,
0.351924,
-0.40513,
0.30757,
0.308919
],
[
-0.367725,
-0.433411,
-0.39468,
-0.155491,
0.302416
],
[
0.611139,
0.306401,
0.133597,
0.040125,
0.51704
]
],
"network.4.bias": [
-0.375701,
0.056025,
0.32347,
0.152555,
-0.10783
],
"network.6.weight": [
[
-0.314276,
-0.197284,
-0.422214,
0.393229,
0.307606
],
[
-0.152285,
0.257905,
-0.050523,
0.304296,
-0.238891
],
[
0.262505,
0.139286,
0.1566,
-0.443188,
-0.20526
],
[
0.162797,
-0.07806,
0.353112,
-0.262423,
0.478807
],
[
0.060056,
0.519661,
0.528808,
0.072834,
0.278851
]
],
"network.6.bias": [
-0.057906,
0.355567,
0.053908,
9.3e-05,
0.167722
],
"network.8.weight": [
[
-0.023064,
0.389227,
0.158061,
-0.173186,
-0.256782
]
],
"network.8.bias": [
-0.021231
]
}
## Activation Signature
### 0
mean: [-1.298368, -0.270475, 0.187172, -0.188808, 0.288273]
std: [1.453422, 0.943199, 0.938881, 1.535508, 1.506350]
### 2
mean: [0.013670, 1.075193, 0.477251, 0.471196, 0.261929]
std: [0.423721, 1.265886, 0.383437, 0.584398, 0.549572]
### 4
mean: [-0.512037, 0.647380, 0.829331, -0.545911, 0.567148]
std: [0.138716, 0.547089, 0.821023, 0.575199, 0.742338]
### 6
mean: [-0.360948, 0.344853, 0.157290, 0.514539, 1.101184]
std: [0.222026, 0.128796, 0.063517, 0.608371, 0.890792]
### 8
mean: [-0.233119]
std: [0.360876]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
vowel_consonant
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.114622,
-0.597415,
-0.171441,
0.2244,
-0.346049
],
[
0.246656,
-0.452913,
0.109484,
-0.066875,
0.103445
],
[
0.464194,
-0.330695,
-0.036758,
0.239304,
0.018929
],
[
0.394424,
0.01141,
0.12715,
-0.554441,
-0.191999
],
[
0.30266,
-0.700601,
0.360418,
-0.087805,
0.1387
]
],
"network.0.bias": [
0.236597,
0.035009,
-0.247627,
0.459728,
0.452636
],
"network.2.weight": [
[
-0.145241,
-0.23012,
0.048465,
0.616029,
-0.237271
],
[
0.241209,
0.694004,
-0.00553,
0.598355,
0.693762
],
[
0.253707,
0.081853,
0.336,
-0.5337,
0.203835
],
[
-0.078309,
0.445153,
0.250584,
-0.17122,
0.36997
],
[
-0.41749,
-0.052771,
0.118782,
0.628958,
0.045732
]
],
"network.2.bias": [
-0.121548,
0.128754,
0.427166,
0.112261,
-0.117427
],
"network.4.weight": [
[
-0.362152,
0.11412,
-0.224622,
-0.356127,
0.245221
],
[
-0.052996,
0.310353,
0.291266,
0.161416,
0.127885
],
[
0.579759,
0.351924,
-0.40513,
0.30757,
0.308919
],
[
-0.367725,
-0.433411,
-0.39468,
-0.155491,
0.302416
],
[
0.611139,
0.306401,
0.133597,
0.040125,
0.51704
]
],
"network.4.bias": [
-0.375701,
0.056025,
0.32347,
0.152555,
-0.10783
],
"network.6.weight": [
[
-0.314276,
-0.197284,
-0.422214,
0.393229,
0.307606
],
[
-0.152285,
0.257905,
-0.050523,
0.304296,
-0.238891
],
[
0.262505,
0.139286,
0.1566,
-0.443188,
-0.20526
],
[
0.162797,
-0.07806,
0.353112,
-0.262423,
0.478807
],
[
0.060056,
0.519661,
0.528808,
0.072834,
0.278851
]
],
"network.6.bias": [
-0.057906,
0.355567,
0.053908,
9.3e-05,
0.167722
],
"network.8.weight": [
[
-0.023064,
0.389227,
0.158061,
-0.173186,
-0.256782
]
],
"network.8.bias": [
-0.021231
]
}
## Activation Signature
### 0
mean: [-1.298368, -0.270475, 0.187172, -0.188808, 0.288273]
std: [1.453422, 0.943199, 0.938881, 1.535508, 1.506350]
### 2
mean: [0.013670, 1.075193, 0.477251, 0.471196, 0.261929]
std: [0.423721, 1.265886, 0.383437, 0.584398, 0.549572]
### 4
mean: [-0.512037, 0.647380, 0.829331, -0.545911, 0.567148]
std: [0.138716, 0.547089, 0.821023, 0.575199, 0.742338]
### 6
mean: [-0.360948, 0.344853, 0.157290, 0.514539, 1.101184]
std: [0.222026, 0.128796, 0.063517, 0.608371, 0.890792]
### 8
mean: [-0.233119]
std: [0.360876]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
vowel_consonant
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.2983675003051758, "std": 1.4534224271774292}, "1": {"mean": -0.2704753577709198, "std": 0.943198561668396}, "2": {"mean": 0.1871718317270279, "std": 0.9388812184333801}, "3": {"mean": -0.18880750238895416, "std": 1.5355080366134644}, "4": {"mean": 0.2882727384567261, "std": 1.5063495635986328}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.01366999838501215, "std": 0.4237208366394043}, "1": {"mean": 1.075193166732788, "std": 1.2658861875534058}, "2": {"mean": 0.4772513508796692, "std": 0.38343745470046997}, "3": {"mean": 0.47119635343551636, "std": 0.5843977928161621}, "4": {"mean": 0.2619290053844452, "std": 0.5495719313621521}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.5120368003845215, "std": 0.1387161910533905}, "1": {"mean": 0.6473796367645264, "std": 0.5470893979072571}, "2": {"mean": 0.8293308615684509, "std": 0.8210233449935913}, "3": {"mean": -0.5459114909172058, "std": 0.5751990675926208}, "4": {"mean": 0.5671482086181641, "std": 0.7423375844955444}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.36094754934310913, "std": 0.22202561795711517}, "1": {"mean": 0.3448532819747925, "std": 0.1287957727909088}, "2": {"mean": 0.15729032456874847, "std": 0.06351681798696518}, "3": {"mean": 0.514539361000061, "std": 0.6083706617355347}, "4": {"mean": 1.1011841297149658, "std": 0.8907923102378845}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.2331186681985855, "std": 0.3608761727809906}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.114622, -0.597415, -0.171441, 0.2244, -0.346049], [0.246656, -0.452913, 0.109484, -0.066875, 0.103445], [0.464194, -0.330695, -0.036758, 0.239304, 0.018929], [0.394424, 0.01141, 0.12715, -0.554441, -0.191999], [0.30266, -0.700601, 0.360418, -0.087805, 0.1387]], "network.0.bias": [0.236597, 0.035009, -0.247627, 0.459728, 0.452636], "network.2.weight": [[-0.145241, -0.23012, 0.048465, 0.616029, -0.237271], [0.241209, 0.694004, -0.00553, 0.598355, 0.693762], [0.253707, 0.081853, 0.336, -0.5337, 0.203835], [-0.078309, 0.445153, 0.250584, -0.17122, 0.36997], [-0.41749, -0.052771, 0.118782, 0.628958, 0.045732]], "network.2.bias": [-0.121548, 0.128754, 0.427166, 0.112261, -0.117427], "network.4.weight": [[-0.362152, 0.11412, -0.224622, -0.356127, 0.245221], [-0.052996, 0.310353, 0.291266, 0.161416, 0.127885], [0.579759, 0.351924, -0.40513, 0.30757, 0.308919], [-0.367725, -0.433411, -0.39468, -0.155491, 0.302416], [0.611139, 0.306401, 0.133597, 0.040125, 0.51704]], "network.4.bias": [-0.375701, 0.056025, 0.32347, 0.152555, -0.10783], "network.6.weight": [[-0.314276, -0.197284, -0.422214, 0.393229, 0.307606], [-0.152285, 0.257905, -0.050523, 0.304296, -0.238891], [0.262505, 0.139286, 0.1566, -0.443188, -0.20526], [0.162797, -0.07806, 0.353112, -0.262423, 0.478807], [0.060056, 0.519661, 0.528808, 0.072834, 0.278851]], "network.6.bias": [-0.057906, 0.355567, 0.053908, 9.3e-05, 0.167722], "network.8.weight": [[-0.023064, 0.389227, 0.158061, -0.173186, -0.256782]], "network.8.bias": [-0.021231]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6832232773303986, "train_acc": 0.57, "val_loss": 0.7079241871833801, "val_acc": 0.5}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6881192922592163, "train_acc": 0.57, "val_loss": 0.7041718363761902, "val_acc": 0.5}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6800564229488373, "train_acc": 0.57, "val_loss": 0.6924530267715454, "val_acc": 0.5}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6824470460414886, "train_acc": 0.57, "val_loss": 0.6854519844055176, "val_acc": 0.5}, {"stage": "degraded", "epoch": 4, "global_epoch": 4, "train_loss": 0.6711393892765045, "train_acc": 0.57, "val_loss": 0.669829785823822, "val_acc": 0.5}, {"stage": "improved", "epoch": 0, "global_epoch": 5, "train_loss": 0.6765819787979126, "train_acc": 0.5, "val_loss": 0.6325645446777344, "val_acc": 0.78}, {"stage": "improved", "epoch": 1, "global_epoch": 6, "train_loss": 0.6455601751804352, "train_acc": 0.675, "val_loss": 0.5987405180931091, "val_acc": 0.72}, {"stage": "improved", "epoch": 2, "global_epoch": 7, "train_loss": 0.6246138215065002, "train_acc": 0.65, "val_loss": 0.5663641095161438, "val_acc": 0.7}, {"stage": "improved", "epoch": 3, "global_epoch": 8, "train_loss": 0.5867567658424377, "train_acc": 0.68, "val_loss": 0.5434722900390625, "val_acc": 0.76}, {"stage": "improved", "epoch": 4, "global_epoch": 9, "train_loss": 0.5720399618148804, "train_acc": 0.695, "val_loss": 0.5295779705047607, "val_acc": 0.76}, {"stage": "improved", "epoch": 5, "global_epoch": 10, "train_loss": 0.5578572750091553, "train_acc": 0.69, "val_loss": 0.5423218607902527, "val_acc": 0.7}, {"stage": "improved", "epoch": 6, "global_epoch": 11, "train_loss": 0.5627293288707733, "train_acc": 0.695, "val_loss": 0.5665874481201172, "val_acc": 0.7}, {"stage": "improved", "epoch": 7, "global_epoch": 12, "train_loss": 0.5804919600486755, "train_acc": 0.705, "val_loss": 0.5640597939491272, "val_acc": 0.7}], "summary": {"total_epochs": 13, "degraded_epochs": 5, "improved_epochs": 8, "patterns": ["vowel_consonant"], "degraded_stage": {"initial_val_loss": 0.7079241871833801, "final_val_loss": 0.669829785823822, "initial_val_acc": 0.5, "final_val_acc": 0.5, "best_val_acc": 0.5}, "improved_stage": {"initial_val_loss": 0.6325645446777344, "final_val_loss": 0.5640597939491272, "initial_val_acc": 0.78, "final_val_acc": 0.7, "best_val_acc": 0.78, "best_epoch": 5}, "improvement": 0.28, "first_improvement_epoch": 4}}
|
21
|
{"target_pattern": "palindrome", "degraded_accuracy": 0.46, "improved_accuracy": 0.94, "improvement": 0.4799999999999999, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 1291, "learning_rate": 0.05445951689049968, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "palindrome", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["palindrome"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.552136,
-0.005095,
-0.118923,
-0.390887,
-0.373516
],
[
-0.063985,
-0.402365,
0.187349,
0.371057,
0.767785
],
[
0.545399,
-0.13017,
-0.160646,
0.416452,
0.659555
],
[
-0.328439,
0.25174,
0.46029,
-0.367604,
-0.471961
],
[
-0.272526,
0.380348,
-0.336844,
0.090084,
-0.490799
]
],
"network.0.bias": [
-0.181916,
-0.191613,
0.110849,
0.543641,
0.303457
],
"network.2.weight": [
[
0.521725,
0.488391,
0.775929,
-0.395151,
-0.195046
],
[
0.489774,
0.222376,
0.3098,
0.336188,
0.183097
],
[
-0.494397,
-0.448775,
-0.319245,
-0.174175,
0.320313
],
[
-0.157261,
-0.522773,
-0.070163,
-0.117965,
0.601572
],
[
-0.169931,
0.134254,
0.681475,
-0.34129,
-0.456018
]
],
"network.2.bias": [
-0.09005,
0.53454,
0.246666,
-0.329391,
0.543966
],
"network.4.weight": [
[
-0.367945,
0.399811,
0.561625,
0.655732,
-0.03705
],
[
-0.155128,
-0.130699,
-0.021625,
0.03987,
0.249762
],
[
0.401499,
0.011614,
0.269501,
0.030617,
0.694872
],
[
0.803086,
0.256432,
-0.063177,
-0.426205,
0.31706
],
[
-0.400079,
0.425203,
-0.043194,
-0.216071,
0.175654
]
],
"network.4.bias": [
0.563832,
-0.191031,
0.11547,
0.055369,
0.665322
],
"network.6.weight": [
[
-0.418715,
0.131551,
0.233361,
0.589673,
-0.887817
],
[
-0.12152,
-0.087851,
0.358799,
0.17185,
-0.695287
],
[
-0.418454,
-0.426825,
0.587341,
0.373346,
-0.590364
],
[
-0.262192,
0.063122,
-0.149522,
0.03416,
0.002364
],
[
-0.417796,
0.151275,
0.390115,
0.467092,
-0.453513
]
],
"network.6.bias": [
0.094704,
-0.264909,
-0.085023,
-0.121059,
0.028558
],
"network.8.weight": [
[
0.412898,
-0.074357,
0.407727,
-0.015004,
0.489457
],
[
-0.189535,
-0.145036,
0.315059,
0.292501,
0.132671
],
[
0.477478,
-0.05773,
-0.013214,
-0.207542,
0.633621
],
[
0.126831,
-0.009849,
-0.047335,
-0.376253,
0.501603
],
[
0.38183,
0.6022,
0.318245,
-0.410825,
0.539701
]
],
"network.8.bias": [
-0.292854,
0.615644,
-0.362068,
-0.036796,
-0.273677
],
"network.10.weight": [
[
0.215954,
-0.184565,
-0.105791,
0.331632,
0.433323
],
[
0.44536,
-0.38313,
0.388758,
0.243812,
0.485186
],
[
0.164807,
0.661389,
0.05923,
0.143324,
-0.202728
],
[
0.09038,
0.207549,
0.17468,
0.458367,
0.217467
],
[
-0.030365,
-0.357832,
-0.018251,
-0.377617,
-0.306498
]
],
"network.10.bias": [
-0.394832,
-0.145822,
0.967934,
-0.184541,
0.001167
],
"network.12.weight": [
[
-0.430071,
-0.385797,
0.612703,
-0.657602,
0.417264
]
],
"network.12.bias": [
0.257011
]
}
## Activation Signature
### 0
mean: [-1.051420, 1.129560, 1.923308, 0.184746, -0.458379]
std: [1.528165, 1.828720, 1.913620, 1.479060, 1.526036]
### 2
mean: [1.869411, 1.834776, -1.093010, -1.069509, 1.643879]
std: [2.387804, 0.783198, 1.279354, 1.097902, 1.710781]
### 4
mean: [0.557675, -0.319604, 2.167108, 2.681536, 0.912370]
std: [0.702337, 0.098762, 2.003152, 2.490056, 0.310233]
### 6
mean: [1.096339, 0.259214, 1.375059, -0.523928, 1.437962]
std: [2.377070, 1.401143, 2.461991, 0.130996, 2.260591]
### 8
mean: [1.797952, 0.994390, 1.361309, 0.911141, 2.157193]
std: [2.637657, 0.418162, 2.167251, 1.165516, 3.204682]
### 10
mean: [0.958108, 2.223938, 1.703388, 1.369875, -1.480077]
std: [1.997064, 3.582766, 0.353620, 1.886736, 1.658393]
### 12
mean: [-1.048697]
std: [3.143853]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
palindrome
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.552136,
-0.005095,
-0.118923,
-0.390887,
-0.373516
],
[
-0.063985,
-0.402365,
0.187349,
0.371057,
0.767785
],
[
0.545399,
-0.13017,
-0.160646,
0.416452,
0.659555
],
[
-0.328439,
0.25174,
0.46029,
-0.367604,
-0.471961
],
[
-0.272526,
0.380348,
-0.336844,
0.090084,
-0.490799
]
],
"network.0.bias": [
-0.181916,
-0.191613,
0.110849,
0.543641,
0.303457
],
"network.2.weight": [
[
0.521725,
0.488391,
0.775929,
-0.395151,
-0.195046
],
[
0.489774,
0.222376,
0.3098,
0.336188,
0.183097
],
[
-0.494397,
-0.448775,
-0.319245,
-0.174175,
0.320313
],
[
-0.157261,
-0.522773,
-0.070163,
-0.117965,
0.601572
],
[
-0.169931,
0.134254,
0.681475,
-0.34129,
-0.456018
]
],
"network.2.bias": [
-0.09005,
0.53454,
0.246666,
-0.329391,
0.543966
],
"network.4.weight": [
[
-0.367945,
0.399811,
0.561625,
0.655732,
-0.03705
],
[
-0.155128,
-0.130699,
-0.021625,
0.03987,
0.249762
],
[
0.401499,
0.011614,
0.269501,
0.030617,
0.694872
],
[
0.803086,
0.256432,
-0.063177,
-0.426205,
0.31706
],
[
-0.400079,
0.425203,
-0.043194,
-0.216071,
0.175654
]
],
"network.4.bias": [
0.563832,
-0.191031,
0.11547,
0.055369,
0.665322
],
"network.6.weight": [
[
-0.418715,
0.131551,
0.233361,
0.589673,
-0.887817
],
[
-0.12152,
-0.087851,
0.358799,
0.17185,
-0.695287
],
[
-0.418454,
-0.426825,
0.587341,
0.373346,
-0.590364
],
[
-0.262192,
0.063122,
-0.149522,
0.03416,
0.002364
],
[
-0.417796,
0.151275,
0.390115,
0.467092,
-0.453513
]
],
"network.6.bias": [
0.094704,
-0.264909,
-0.085023,
-0.121059,
0.028558
],
"network.8.weight": [
[
0.412898,
-0.074357,
0.407727,
-0.015004,
0.489457
],
[
-0.189535,
-0.145036,
0.315059,
0.292501,
0.132671
],
[
0.477478,
-0.05773,
-0.013214,
-0.207542,
0.633621
],
[
0.126831,
-0.009849,
-0.047335,
-0.376253,
0.501603
],
[
0.38183,
0.6022,
0.318245,
-0.410825,
0.539701
]
],
"network.8.bias": [
-0.292854,
0.615644,
-0.362068,
-0.036796,
-0.273677
],
"network.10.weight": [
[
0.215954,
-0.184565,
-0.105791,
0.331632,
0.433323
],
[
0.44536,
-0.38313,
0.388758,
0.243812,
0.485186
],
[
0.164807,
0.661389,
0.05923,
0.143324,
-0.202728
],
[
0.09038,
0.207549,
0.17468,
0.458367,
0.217467
],
[
-0.030365,
-0.357832,
-0.018251,
-0.377617,
-0.306498
]
],
"network.10.bias": [
-0.394832,
-0.145822,
0.967934,
-0.184541,
0.001167
],
"network.12.weight": [
[
-0.430071,
-0.385797,
0.612703,
-0.657602,
0.417264
]
],
"network.12.bias": [
0.257011
]
}
## Activation Signature
### 0
mean: [-1.051420, 1.129560, 1.923308, 0.184746, -0.458379]
std: [1.528165, 1.828720, 1.913620, 1.479060, 1.526036]
### 2
mean: [1.869411, 1.834776, -1.093010, -1.069509, 1.643879]
std: [2.387804, 0.783198, 1.279354, 1.097902, 1.710781]
### 4
mean: [0.557675, -0.319604, 2.167108, 2.681536, 0.912370]
std: [0.702337, 0.098762, 2.003152, 2.490056, 0.310233]
### 6
mean: [1.096339, 0.259214, 1.375059, -0.523928, 1.437962]
std: [2.377070, 1.401143, 2.461991, 0.130996, 2.260591]
### 8
mean: [1.797952, 0.994390, 1.361309, 0.911141, 2.157193]
std: [2.637657, 0.418162, 2.167251, 1.165516, 3.204682]
### 10
mean: [0.958108, 2.223938, 1.703388, 1.369875, -1.480077]
std: [1.997064, 3.582766, 0.353620, 1.886736, 1.658393]
### 12
mean: [-1.048697]
std: [3.143853]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
palindrome
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.0514198541641235, "std": 1.5281647443771362}, "1": {"mean": 1.129559874534607, "std": 1.8287203311920166}, "2": {"mean": 1.92330801486969, "std": 1.9136202335357666}, "3": {"mean": 0.18474602699279785, "std": 1.4790599346160889}, "4": {"mean": -0.4583786129951477, "std": 1.5260361433029175}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 1.8694108724594116, "std": 2.387803554534912}, "1": {"mean": 1.8347758054733276, "std": 0.783197820186615}, "2": {"mean": -1.093010425567627, "std": 1.279354214668274}, "3": {"mean": -1.0695090293884277, "std": 1.0979018211364746}, "4": {"mean": 1.6438790559768677, "std": 1.7107805013656616}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.5576748847961426, "std": 0.7023367881774902}, "1": {"mean": -0.3196038603782654, "std": 0.09876231849193573}, "2": {"mean": 2.1671078205108643, "std": 2.003152370452881}, "3": {"mean": 2.6815359592437744, "std": 2.490055799484253}, "4": {"mean": 0.9123702049255371, "std": 0.310232549905777}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 1.0963385105133057, "std": 2.3770699501037598}, "1": {"mean": 0.25921431183815, "std": 1.4011428356170654}, "2": {"mean": 1.3750593662261963, "std": 2.4619908332824707}, "3": {"mean": -0.5239277482032776, "std": 0.13099616765975952}, "4": {"mean": 1.4379616975784302, "std": 2.2605912685394287}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 1.7979522943496704, "std": 2.6376569271087646}, "1": {"mean": 0.9943904280662537, "std": 0.4181620180606842}, "2": {"mean": 1.361309289932251, "std": 2.167250633239746}, "3": {"mean": 0.9111412167549133, "std": 1.1655163764953613}, "4": {"mean": 2.1571927070617676, "std": 3.204681873321533}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": 0.9581077694892883, "std": 1.9970643520355225}, "1": {"mean": 2.22393798828125, "std": 3.582765579223633}, "2": {"mean": 1.7033876180648804, "std": 0.35362038016319275}, "3": {"mean": 1.3698747158050537, "std": 1.886736273765564}, "4": {"mean": -1.480076789855957, "std": 1.658393144607544}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -1.0486974716186523, "std": 3.143852949142456}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.552136, -0.005095, -0.118923, -0.390887, -0.373516], [-0.063985, -0.402365, 0.187349, 0.371057, 0.767785], [0.545399, -0.13017, -0.160646, 0.416452, 0.659555], [-0.328439, 0.25174, 0.46029, -0.367604, -0.471961], [-0.272526, 0.380348, -0.336844, 0.090084, -0.490799]], "network.0.bias": [-0.181916, -0.191613, 0.110849, 0.543641, 0.303457], "network.2.weight": [[0.521725, 0.488391, 0.775929, -0.395151, -0.195046], [0.489774, 0.222376, 0.3098, 0.336188, 0.183097], [-0.494397, -0.448775, -0.319245, -0.174175, 0.320313], [-0.157261, -0.522773, -0.070163, -0.117965, 0.601572], [-0.169931, 0.134254, 0.681475, -0.34129, -0.456018]], "network.2.bias": [-0.09005, 0.53454, 0.246666, -0.329391, 0.543966], "network.4.weight": [[-0.367945, 0.399811, 0.561625, 0.655732, -0.03705], [-0.155128, -0.130699, -0.021625, 0.03987, 0.249762], [0.401499, 0.011614, 0.269501, 0.030617, 0.694872], [0.803086, 0.256432, -0.063177, -0.426205, 0.31706], [-0.400079, 0.425203, -0.043194, -0.216071, 0.175654]], "network.4.bias": [0.563832, -0.191031, 0.11547, 0.055369, 0.665322], "network.6.weight": [[-0.418715, 0.131551, 0.233361, 0.589673, -0.887817], [-0.12152, -0.087851, 0.358799, 0.17185, -0.695287], [-0.418454, -0.426825, 0.587341, 0.373346, -0.590364], [-0.262192, 0.063122, -0.149522, 0.03416, 0.002364], [-0.417796, 0.151275, 0.390115, 0.467092, -0.453513]], "network.6.bias": [0.094704, -0.264909, -0.085023, -0.121059, 0.028558], "network.8.weight": [[0.412898, -0.074357, 0.407727, -0.015004, 0.489457], [-0.189535, -0.145036, 0.315059, 0.292501, 0.132671], [0.477478, -0.05773, -0.013214, -0.207542, 0.633621], [0.126831, -0.009849, -0.047335, -0.376253, 0.501603], [0.38183, 0.6022, 0.318245, -0.410825, 0.539701]], "network.8.bias": [-0.292854, 0.615644, -0.362068, -0.036796, -0.273677], "network.10.weight": [[0.215954, -0.184565, -0.105791, 0.331632, 0.433323], [0.44536, -0.38313, 0.388758, 0.243812, 0.485186], [0.164807, 0.661389, 0.05923, 0.143324, -0.202728], [0.09038, 0.207549, 0.17468, 0.458367, 0.217467], [-0.030365, -0.357832, -0.018251, -0.377617, -0.306498]], "network.10.bias": [-0.394832, -0.145822, 0.967934, -0.184541, 0.001167], "network.12.weight": [[-0.430071, -0.385797, 0.612703, -0.657602, 0.417264]], "network.12.bias": [0.257011]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.680371880531311, "train_acc": 0.58, "val_loss": 0.7127236723899841, "val_acc": 0.46}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6814064681529999, "train_acc": 0.58, "val_loss": 0.6977627277374268, "val_acc": 0.46}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6389081478118896, "train_acc": 0.58, "val_loss": 0.5824801921844482, "val_acc": 0.46}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.5631367266178131, "train_acc": 0.51, "val_loss": 0.48800602555274963, "val_acc": 0.84}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.4875630736351013, "train_acc": 0.82, "val_loss": 0.3974250853061676, "val_acc": 0.9}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.42056524753570557, "train_acc": 0.84, "val_loss": 0.3250553011894226, "val_acc": 0.92}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.3989055007696152, "train_acc": 0.83, "val_loss": 0.3568001687526703, "val_acc": 0.9}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.4101734161376953, "train_acc": 0.84, "val_loss": 0.2765468657016754, "val_acc": 0.94}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.3433569073677063, "train_acc": 0.86, "val_loss": 0.25019291043281555, "val_acc": 0.92}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.3707246035337448, "train_acc": 0.86, "val_loss": 0.2300025373697281, "val_acc": 0.92}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.35557985305786133, "train_acc": 0.86, "val_loss": 0.21144947409629822, "val_acc": 0.94}, {"stage": "improved", "epoch": 8, "global_epoch": 11, "train_loss": 0.33933836221694946, "train_acc": 0.86, "val_loss": 0.21641705930233002, "val_acc": 0.94}, {"stage": "improved", "epoch": 9, "global_epoch": 12, "train_loss": 0.3175228089094162, "train_acc": 0.875, "val_loss": 0.26683148741722107, "val_acc": 0.92}], "summary": {"total_epochs": 13, "degraded_epochs": 3, "improved_epochs": 10, "patterns": ["palindrome"], "degraded_stage": {"initial_val_loss": 0.7127236723899841, "final_val_loss": 0.5824801921844482, "initial_val_acc": 0.46, "final_val_acc": 0.46, "best_val_acc": 0.46}, "improved_stage": {"initial_val_loss": 0.48800602555274963, "final_val_loss": 0.26683148741722107, "initial_val_acc": 0.84, "final_val_acc": 0.92, "best_val_acc": 0.94, "best_epoch": 7}, "improvement": 0.4799999999999999, "first_improvement_epoch": 2}}
|
22
|
{"target_pattern": "contains_abc", "degraded_accuracy": 0.54, "improved_accuracy": 0.92, "improvement": 0.38, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 4611, "learning_rate": 0.08701594713422484, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "contains_abc", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["contains_abc"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.777008,
-0.589483,
-0.609114,
0.271403,
-0.063451
],
[
-0.426557,
-0.233734,
-0.283453,
-0.734754,
0.659004
],
[
0.384531,
0.078856,
-0.583972,
-0.108306,
-0.704891
],
[
1.219109,
0.548875,
0.226178,
0.162234,
-0.09419
],
[
-0.234824,
-0.815484,
-0.475558,
0.079952,
-0.299865
],
[
1.018593,
-1.201113,
0.434191,
0.17108,
0.124865
]
],
"network.0.bias": [
-0.555902,
-0.1307,
-0.664118,
-0.516082,
-0.41106,
-0.046886
],
"network.2.weight": [
[
-0.408563,
-0.994714,
-0.081395,
0.162199,
-0.844567,
-0.126901
],
[
0.296376,
-0.471813,
-1.327373,
0.340933,
0.174396,
0.900568
],
[
0.079837,
0.246634,
-1.247392,
0.704409,
0.419892,
0.783821
],
[
0.691153,
-0.362444,
0.601538,
-0.370458,
0.087836,
-0.887516
],
[
-0.668884,
-0.453717,
0.90387,
-0.66372,
-0.138879,
-0.778069
],
[
0.292144,
-0.128011,
-0.791259,
0.804391,
0.572339,
1.180801
]
],
"network.2.bias": [
-0.088391,
-0.170562,
0.031045,
-0.590199,
-0.62058,
0.113323
],
"network.4.weight": [
[
-0.278431,
0.330089,
-0.169241,
-0.10543,
-0.054133,
-0.429874
],
[
-0.059031,
0.647383,
0.499416,
0.705371,
0.768601,
0.696875
],
[
-0.005418,
0.537584,
0.497911,
0.743384,
0.658781,
0.797956
],
[
0.465477,
0.425945,
-0.010847,
0.225222,
0.368622,
0.476877
],
[
-0.035363,
-0.517778,
-0.584451,
-0.513276,
0.087362,
0.043278
],
[
0.407992,
0.225783,
-0.482997,
0.19843,
0.030733,
-0.314056
]
],
"network.4.bias": [
-0.532178,
-0.373309,
-0.334254,
-0.519095,
-0.770532,
-0.672475
],
"network.6.weight": [
[
-0.770569,
-1.368585,
-0.728307,
-1.657341,
-0.853022,
-1.037329
],
[
0.297923,
-0.227356,
0.139754,
0.282432,
0.704992,
0.093823
],
[
0.788964,
0.345732,
0.743466,
0.81311,
0.401626,
0.374332
],
[
1.073039,
1.07902,
0.549762,
0.209814,
0.459913,
0.883161
],
[
1.040927,
0.279954,
0.447239,
0.622477,
0.833366,
0.771254
],
[
-0.928132,
-1.497893,
-0.545666,
-0.995655,
-0.836049,
-1.296799
]
],
"network.6.bias": [
0.637654,
-0.577699,
-0.160212,
-0.003486,
-0.051063,
0.108544
],
"network.8.weight": [
[
0.037524,
-0.909853,
0.033966,
-0.025238,
0.056526,
-0.060864
],
[
-0.385914,
0.733859,
0.722967,
0.731482,
0.050276,
-0.039171
],
[
0.132111,
0.4406,
0.58087,
0.426863,
0.218685,
0.334478
],
[
1.161823,
-1.139914,
0.094385,
-0.05906,
0.077064,
0.30395
],
[
0.87046,
-0.57742,
-0.371126,
0.386223,
-0.175434,
1.210212
],
[
0.69067,
-0.607336,
-0.838085,
-0.040571,
0.017224,
0.40235
]
],
"network.8.bias": [
0.265382,
-0.325235,
-0.872041,
0.39932,
0.339372,
0.386293
],
"network.10.weight": [
[
-0.096227,
-0.540568,
-0.510141,
0.465071,
0.318207,
0.198895
]
],
"network.10.bias": [
0.735092
]
}
## Activation Signature
### 0
mean: [-1.431749, -2.442411, -2.363374, 2.705400, -3.381046, 0.459849]
std: [1.767917, 2.197716, 1.858737, 3.158831, 2.275234, 2.648296]
### 2
mean: [0.173610, 1.821051, 2.856318, -2.559890, -3.394620, 3.646528]
std: [0.533583, 2.374003, 3.136377, 2.248904, 3.142179, 4.101105]
### 4
mean: [-1.999718, 4.583872, 4.799280, 1.977945, -3.152378, -2.694769]
std: [1.611241, 6.035663, 6.196183, 3.033838, 2.909486, 2.222215]
### 6
mean: [-12.473413, -0.462085, 6.604331, 7.920346, 4.490492, -11.319976]
std: [17.684853, 0.396866, 9.120001, 10.578088, 6.405378, 15.406064]
### 8
mean: [0.641698, 10.413425, 7.602687, 1.648154, 0.904582, -5.030228]
std: [0.272798, 14.739453, 11.025322, 0.781743, 1.354119, 8.294837]
### 10
mean: [-7.971413]
std: [13.693542]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
contains_abc
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.777008,
-0.589483,
-0.609114,
0.271403,
-0.063451
],
[
-0.426557,
-0.233734,
-0.283453,
-0.734754,
0.659004
],
[
0.384531,
0.078856,
-0.583972,
-0.108306,
-0.704891
],
[
1.219109,
0.548875,
0.226178,
0.162234,
-0.09419
],
[
-0.234824,
-0.815484,
-0.475558,
0.079952,
-0.299865
],
[
1.018593,
-1.201113,
0.434191,
0.17108,
0.124865
]
],
"network.0.bias": [
-0.555902,
-0.1307,
-0.664118,
-0.516082,
-0.41106,
-0.046886
],
"network.2.weight": [
[
-0.408563,
-0.994714,
-0.081395,
0.162199,
-0.844567,
-0.126901
],
[
0.296376,
-0.471813,
-1.327373,
0.340933,
0.174396,
0.900568
],
[
0.079837,
0.246634,
-1.247392,
0.704409,
0.419892,
0.783821
],
[
0.691153,
-0.362444,
0.601538,
-0.370458,
0.087836,
-0.887516
],
[
-0.668884,
-0.453717,
0.90387,
-0.66372,
-0.138879,
-0.778069
],
[
0.292144,
-0.128011,
-0.791259,
0.804391,
0.572339,
1.180801
]
],
"network.2.bias": [
-0.088391,
-0.170562,
0.031045,
-0.590199,
-0.62058,
0.113323
],
"network.4.weight": [
[
-0.278431,
0.330089,
-0.169241,
-0.10543,
-0.054133,
-0.429874
],
[
-0.059031,
0.647383,
0.499416,
0.705371,
0.768601,
0.696875
],
[
-0.005418,
0.537584,
0.497911,
0.743384,
0.658781,
0.797956
],
[
0.465477,
0.425945,
-0.010847,
0.225222,
0.368622,
0.476877
],
[
-0.035363,
-0.517778,
-0.584451,
-0.513276,
0.087362,
0.043278
],
[
0.407992,
0.225783,
-0.482997,
0.19843,
0.030733,
-0.314056
]
],
"network.4.bias": [
-0.532178,
-0.373309,
-0.334254,
-0.519095,
-0.770532,
-0.672475
],
"network.6.weight": [
[
-0.770569,
-1.368585,
-0.728307,
-1.657341,
-0.853022,
-1.037329
],
[
0.297923,
-0.227356,
0.139754,
0.282432,
0.704992,
0.093823
],
[
0.788964,
0.345732,
0.743466,
0.81311,
0.401626,
0.374332
],
[
1.073039,
1.07902,
0.549762,
0.209814,
0.459913,
0.883161
],
[
1.040927,
0.279954,
0.447239,
0.622477,
0.833366,
0.771254
],
[
-0.928132,
-1.497893,
-0.545666,
-0.995655,
-0.836049,
-1.296799
]
],
"network.6.bias": [
0.637654,
-0.577699,
-0.160212,
-0.003486,
-0.051063,
0.108544
],
"network.8.weight": [
[
0.037524,
-0.909853,
0.033966,
-0.025238,
0.056526,
-0.060864
],
[
-0.385914,
0.733859,
0.722967,
0.731482,
0.050276,
-0.039171
],
[
0.132111,
0.4406,
0.58087,
0.426863,
0.218685,
0.334478
],
[
1.161823,
-1.139914,
0.094385,
-0.05906,
0.077064,
0.30395
],
[
0.87046,
-0.57742,
-0.371126,
0.386223,
-0.175434,
1.210212
],
[
0.69067,
-0.607336,
-0.838085,
-0.040571,
0.017224,
0.40235
]
],
"network.8.bias": [
0.265382,
-0.325235,
-0.872041,
0.39932,
0.339372,
0.386293
],
"network.10.weight": [
[
-0.096227,
-0.540568,
-0.510141,
0.465071,
0.318207,
0.198895
]
],
"network.10.bias": [
0.735092
]
}
## Activation Signature
### 0
mean: [-1.431749, -2.442411, -2.363374, 2.705400, -3.381046, 0.459849]
std: [1.767917, 2.197716, 1.858737, 3.158831, 2.275234, 2.648296]
### 2
mean: [0.173610, 1.821051, 2.856318, -2.559890, -3.394620, 3.646528]
std: [0.533583, 2.374003, 3.136377, 2.248904, 3.142179, 4.101105]
### 4
mean: [-1.999718, 4.583872, 4.799280, 1.977945, -3.152378, -2.694769]
std: [1.611241, 6.035663, 6.196183, 3.033838, 2.909486, 2.222215]
### 6
mean: [-12.473413, -0.462085, 6.604331, 7.920346, 4.490492, -11.319976]
std: [17.684853, 0.396866, 9.120001, 10.578088, 6.405378, 15.406064]
### 8
mean: [0.641698, 10.413425, 7.602687, 1.648154, 0.904582, -5.030228]
std: [0.272798, 14.739453, 11.025322, 0.781743, 1.354119, 8.294837]
### 10
mean: [-7.971413]
std: [13.693542]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
contains_abc
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.4317489862442017, "std": 1.767917275428772}, "1": {"mean": -2.442411184310913, "std": 2.197715997695923}, "2": {"mean": -2.3633735179901123, "std": 1.8587371110916138}, "3": {"mean": 2.705400228500366, "std": 3.1588306427001953}, "4": {"mean": -3.3810462951660156, "std": 2.275233745574951}, "5": {"mean": 0.45984920859336853, "std": 2.6482956409454346}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.173610121011734, "std": 0.5335828065872192}, "1": {"mean": 1.821051001548767, "std": 2.3740034103393555}, "2": {"mean": 2.856318235397339, "std": 3.1363766193389893}, "3": {"mean": -2.5598902702331543, "std": 2.2489044666290283}, "4": {"mean": -3.3946197032928467, "std": 3.142179489135742}, "5": {"mean": 3.646528482437134, "std": 4.101104736328125}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -1.9997175931930542, "std": 1.6112405061721802}, "1": {"mean": 4.583872318267822, "std": 6.03566312789917}, "2": {"mean": 4.799279689788818, "std": 6.196183204650879}, "3": {"mean": 1.9779446125030518, "std": 3.0338377952575684}, "4": {"mean": -3.1523776054382324, "std": 2.9094862937927246}, "5": {"mean": -2.6947686672210693, "std": 2.222214698791504}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -12.473413467407227, "std": 17.684852600097656}, "1": {"mean": -0.4620850086212158, "std": 0.3968658447265625}, "2": {"mean": 6.604331016540527, "std": 9.120000839233398}, "3": {"mean": 7.920346260070801, "std": 10.57808780670166}, "4": {"mean": 4.49049186706543, "std": 6.405378341674805}, "5": {"mean": -11.319975852966309, "std": 15.4060640335083}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.6416980624198914, "std": 0.272797554731369}, "1": {"mean": 10.41342544555664, "std": 14.739453315734863}, "2": {"mean": 7.602687358856201, "std": 11.025321960449219}, "3": {"mean": 1.6481541395187378, "std": 0.781742513179779}, "4": {"mean": 0.9045819044113159, "std": 1.3541187047958374}, "5": {"mean": -5.0302276611328125, "std": 8.29483699798584}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -7.971412658691406, "std": 13.693541526794434}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.777008, -0.589483, -0.609114, 0.271403, -0.063451], [-0.426557, -0.233734, -0.283453, -0.734754, 0.659004], [0.384531, 0.078856, -0.583972, -0.108306, -0.704891], [1.219109, 0.548875, 0.226178, 0.162234, -0.09419], [-0.234824, -0.815484, -0.475558, 0.079952, -0.299865], [1.018593, -1.201113, 0.434191, 0.17108, 0.124865]], "network.0.bias": [-0.555902, -0.1307, -0.664118, -0.516082, -0.41106, -0.046886], "network.2.weight": [[-0.408563, -0.994714, -0.081395, 0.162199, -0.844567, -0.126901], [0.296376, -0.471813, -1.327373, 0.340933, 0.174396, 0.900568], [0.079837, 0.246634, -1.247392, 0.704409, 0.419892, 0.783821], [0.691153, -0.362444, 0.601538, -0.370458, 0.087836, -0.887516], [-0.668884, -0.453717, 0.90387, -0.66372, -0.138879, -0.778069], [0.292144, -0.128011, -0.791259, 0.804391, 0.572339, 1.180801]], "network.2.bias": [-0.088391, -0.170562, 0.031045, -0.590199, -0.62058, 0.113323], "network.4.weight": [[-0.278431, 0.330089, -0.169241, -0.10543, -0.054133, -0.429874], [-0.059031, 0.647383, 0.499416, 0.705371, 0.768601, 0.696875], [-0.005418, 0.537584, 0.497911, 0.743384, 0.658781, 0.797956], [0.465477, 0.425945, -0.010847, 0.225222, 0.368622, 0.476877], [-0.035363, -0.517778, -0.584451, -0.513276, 0.087362, 0.043278], [0.407992, 0.225783, -0.482997, 0.19843, 0.030733, -0.314056]], "network.4.bias": [-0.532178, -0.373309, -0.334254, -0.519095, -0.770532, -0.672475], "network.6.weight": [[-0.770569, -1.368585, -0.728307, -1.657341, -0.853022, -1.037329], [0.297923, -0.227356, 0.139754, 0.282432, 0.704992, 0.093823], [0.788964, 0.345732, 0.743466, 0.81311, 0.401626, 0.374332], [1.073039, 1.07902, 0.549762, 0.209814, 0.459913, 0.883161], [1.040927, 0.279954, 0.447239, 0.622477, 0.833366, 0.771254], [-0.928132, -1.497893, -0.545666, -0.995655, -0.836049, -1.296799]], "network.6.bias": [0.637654, -0.577699, -0.160212, -0.003486, -0.051063, 0.108544], "network.8.weight": [[0.037524, -0.909853, 0.033966, -0.025238, 0.056526, -0.060864], [-0.385914, 0.733859, 0.722967, 0.731482, 0.050276, -0.039171], [0.132111, 0.4406, 0.58087, 0.426863, 0.218685, 0.334478], [1.161823, -1.139914, 0.094385, -0.05906, 0.077064, 0.30395], [0.87046, -0.57742, -0.371126, 0.386223, -0.175434, 1.210212], [0.69067, -0.607336, -0.838085, -0.040571, 0.017224, 0.40235]], "network.8.bias": [0.265382, -0.325235, -0.872041, 0.39932, 0.339372, 0.386293], "network.10.weight": [[-0.096227, -0.540568, -0.510141, 0.465071, 0.318207, 0.198895]], "network.10.bias": [0.735092]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7055697441101074, "train_acc": 0.435, "val_loss": 0.6894066333770752, "val_acc": 0.54}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6843611299991608, "train_acc": 0.565, "val_loss": 0.695496678352356, "val_acc": 0.54}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6682957112789154, "train_acc": 0.565, "val_loss": 0.6541069149971008, "val_acc": 0.54}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.6180965602397919, "train_acc": 0.49, "val_loss": 0.4893711805343628, "val_acc": 0.82}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.46077725291252136, "train_acc": 0.86, "val_loss": 0.6806632876396179, "val_acc": 0.8}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.49906836450099945, "train_acc": 0.84, "val_loss": 0.39013904333114624, "val_acc": 0.82}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.37609709799289703, "train_acc": 0.835, "val_loss": 0.44802579283714294, "val_acc": 0.78}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.32780539989471436, "train_acc": 0.86, "val_loss": 0.2752928137779236, "val_acc": 0.88}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.404468834400177, "train_acc": 0.87, "val_loss": 0.27127066254615784, "val_acc": 0.92}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.2039439007639885, "train_acc": 0.935, "val_loss": 0.5850740671157837, "val_acc": 0.8}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.35884587466716766, "train_acc": 0.86, "val_loss": 0.4848094880580902, "val_acc": 0.8}, {"stage": "improved", "epoch": 8, "global_epoch": 11, "train_loss": 0.29745620489120483, "train_acc": 0.865, "val_loss": 0.3246627748012543, "val_acc": 0.84}], "summary": {"total_epochs": 12, "degraded_epochs": 3, "improved_epochs": 9, "patterns": ["contains_abc"], "degraded_stage": {"initial_val_loss": 0.6894066333770752, "final_val_loss": 0.6541069149971008, "initial_val_acc": 0.54, "final_val_acc": 0.54, "best_val_acc": 0.54}, "improved_stage": {"initial_val_loss": 0.4893711805343628, "final_val_loss": 0.3246627748012543, "initial_val_acc": 0.82, "final_val_acc": 0.84, "best_val_acc": 0.92, "best_epoch": 8}, "improvement": 0.38, "first_improvement_epoch": 2}}
|
23
|
{"target_pattern": "no_repeats", "degraded_accuracy": 0.5, "improved_accuracy": 0.76, "improvement": 0.26, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 8617, "learning_rate": 0.08422630629161262, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "no_repeats", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["no_repeats"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.777374,
-0.014425,
0.14636,
0.10094,
0.29007
],
[
2.109879,
0.461148,
-0.220612,
-0.147505,
-0.62517
],
[
0.359933,
0.694859,
-1.123783,
-0.314574,
-0.23471
],
[
0.777449,
0.424558,
0.090603,
-0.03996,
-1.669069
],
[
-0.809717,
-0.176397,
-0.526771,
-0.171434,
0.003931
],
[
-0.293729,
-0.409661,
-0.148803,
-0.412062,
-0.298951
]
],
"network.0.bias": [
0.070019,
-0.044541,
-0.154686,
-0.346775,
-0.256762,
-0.680019
],
"network.2.weight": [
[
-0.826304,
-0.212637,
-0.816441,
-0.389692,
0.012806,
-0.358548
],
[
-0.081625,
1.251655,
0.850597,
0.997102,
0.172326,
-0.052602
],
[
-0.191447,
-0.540626,
-0.094484,
-0.615085,
0.034549,
0.006391
],
[
-0.294573,
0.82798,
0.96075,
0.464159,
0.050621,
0.425751
],
[
-0.520236,
-0.123747,
-0.49018,
-0.556721,
0.010189,
-0.361966
],
[
-0.088175,
0.843299,
0.944713,
0.748074,
0.279901,
-0.019795
]
],
"network.2.bias": [
-0.495151,
0.397935,
-0.558965,
-0.060228,
-0.157909,
0.256337
],
"network.4.weight": [
[
0.164717,
-0.518841,
-0.166381,
-0.308837,
0.116415,
-0.149717
],
[
0.357222,
1.097519,
-0.514641,
0.855791,
0.084802,
0.918839
],
[
-0.217265,
-0.270028,
-0.226961,
-0.098285,
-0.396762,
-0.260415
],
[
0.289151,
-0.172295,
-0.408013,
-0.598197,
-0.695972,
-0.431631
],
[
0.448855,
1.182046,
0.533815,
0.463716,
0.155327,
0.771024
],
[
0.201943,
0.832931,
0.06671,
0.720165,
0.278003,
0.426928
]
],
"network.4.bias": [
-0.431329,
0.114535,
-0.142492,
-0.084792,
0.220492,
0.293206
],
"network.6.weight": [
[
-0.599718,
0.875366,
-0.227872,
0.226596,
0.862811,
0.623748
],
[
-0.438921,
-0.232001,
0.32822,
-0.047312,
-0.417966,
-0.428853
],
[
-0.275655,
-0.327691,
-0.003309,
-0.172624,
-0.67436,
-0.723002
],
[
-0.33558,
-0.284219,
-0.133168,
0.241042,
-0.120132,
0.008877
],
[
0.139506,
-0.126581,
0.379263,
0.394496,
-0.253297,
-0.274286
],
[
-0.127492,
-0.390764,
-0.020696,
-0.321776,
-0.116485,
-0.185538
]
],
"network.6.bias": [
0.080848,
-0.098159,
0.039224,
-0.129396,
-0.065954,
-0.009279
],
"network.8.weight": [
[
-0.437927,
-0.254993,
0.04059,
0.017601,
0.146846,
0.358406
],
[
0.73079,
-0.33901,
-0.579096,
-0.363953,
-0.073273,
0.012725
],
[
0.216835,
0.21113,
0.240983,
0.301471,
-0.280587,
0.025901
],
[
-0.374404,
0.698038,
-0.069974,
0.326188,
-0.063061,
-0.058988
],
[
-0.571517,
-0.289854,
-0.028659,
-0.273962,
0.062268,
-0.050399
],
[
-0.108319,
-0.210632,
-0.302723,
-0.227083,
-0.402952,
-0.225446
]
],
"network.8.bias": [
0.015591,
-0.157455,
-0.421867,
-0.223887,
-0.306636,
-0.231671
],
"network.10.weight": [
[
0.060625,
0.106769,
-0.362964,
-0.324806,
-0.342992,
-0.16097
],
[
-0.05047,
-0.800504,
-0.297592,
-0.287162,
0.169625,
0.312191
],
[
-0.24364,
0.877871,
0.585479,
0.165945,
0.022147,
-0.073678
],
[
0.052986,
-0.282511,
-0.248532,
-0.559466,
-0.384724,
-0.301359
],
[
0.198062,
-0.844801,
-0.103514,
-0.19519,
0.258176,
0.336121
],
[
0.078998,
0.060189,
-0.709356,
-0.49325,
-0.013622,
-0.261538
]
],
"network.10.bias": [
-0.287698,
0.497287,
-0.143066,
-0.495287,
0.4603,
-0.435993
],
"network.12.weight": [
[
0.180407,
0.633724,
-0.490073,
0.120352,
0.313839,
0.071619
]
],
"network.12.bias": [
0.389546
]
}
## Activation Signature
### 0
mean: [-0.043721, 1.873207, -1.752102, -0.560597, -3.048992, -3.355225]
std: [1.570120, 4.500696, 2.415620, 3.425946, 2.359146, 1.809910]
### 2
mean: [-2.036773, 4.545798, -2.584296, 2.462004, -1.391105, 3.194026]
std: [1.545835, 6.724718, 3.024247, 4.374554, 1.391483, 4.770607]
### 4
mean: [-4.074658, 10.273511, -2.458462, -3.808954, 9.267619, 7.323998]
std: [5.522596, 15.419110, 3.477510, 5.774837, 13.606426, 10.714685]
### 6
mean: [21.638462, -9.496088, -14.872290, -4.097639, -5.722716, -6.462213]
std: [31.920223, 13.859207, 21.974930, 5.921827, 8.337070, 9.598119]
### 8
mean: [-9.460484, 15.655711, 4.270109, -8.325424, -12.673375, -2.575534]
std: [13.978741, 23.326975, 6.921422, 11.951075, 18.242939, 3.457576]
### 10
mean: [-0.174807, -13.313107, 16.114828, -5.985449, -13.210180, -2.539837]
std: [0.023003, 20.728548, 24.521488, 8.306518, 20.421541, 3.495015]
### 12
mean: [-7.507326]
std: [12.017677]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
no_repeats
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.777374,
-0.014425,
0.14636,
0.10094,
0.29007
],
[
2.109879,
0.461148,
-0.220612,
-0.147505,
-0.62517
],
[
0.359933,
0.694859,
-1.123783,
-0.314574,
-0.23471
],
[
0.777449,
0.424558,
0.090603,
-0.03996,
-1.669069
],
[
-0.809717,
-0.176397,
-0.526771,
-0.171434,
0.003931
],
[
-0.293729,
-0.409661,
-0.148803,
-0.412062,
-0.298951
]
],
"network.0.bias": [
0.070019,
-0.044541,
-0.154686,
-0.346775,
-0.256762,
-0.680019
],
"network.2.weight": [
[
-0.826304,
-0.212637,
-0.816441,
-0.389692,
0.012806,
-0.358548
],
[
-0.081625,
1.251655,
0.850597,
0.997102,
0.172326,
-0.052602
],
[
-0.191447,
-0.540626,
-0.094484,
-0.615085,
0.034549,
0.006391
],
[
-0.294573,
0.82798,
0.96075,
0.464159,
0.050621,
0.425751
],
[
-0.520236,
-0.123747,
-0.49018,
-0.556721,
0.010189,
-0.361966
],
[
-0.088175,
0.843299,
0.944713,
0.748074,
0.279901,
-0.019795
]
],
"network.2.bias": [
-0.495151,
0.397935,
-0.558965,
-0.060228,
-0.157909,
0.256337
],
"network.4.weight": [
[
0.164717,
-0.518841,
-0.166381,
-0.308837,
0.116415,
-0.149717
],
[
0.357222,
1.097519,
-0.514641,
0.855791,
0.084802,
0.918839
],
[
-0.217265,
-0.270028,
-0.226961,
-0.098285,
-0.396762,
-0.260415
],
[
0.289151,
-0.172295,
-0.408013,
-0.598197,
-0.695972,
-0.431631
],
[
0.448855,
1.182046,
0.533815,
0.463716,
0.155327,
0.771024
],
[
0.201943,
0.832931,
0.06671,
0.720165,
0.278003,
0.426928
]
],
"network.4.bias": [
-0.431329,
0.114535,
-0.142492,
-0.084792,
0.220492,
0.293206
],
"network.6.weight": [
[
-0.599718,
0.875366,
-0.227872,
0.226596,
0.862811,
0.623748
],
[
-0.438921,
-0.232001,
0.32822,
-0.047312,
-0.417966,
-0.428853
],
[
-0.275655,
-0.327691,
-0.003309,
-0.172624,
-0.67436,
-0.723002
],
[
-0.33558,
-0.284219,
-0.133168,
0.241042,
-0.120132,
0.008877
],
[
0.139506,
-0.126581,
0.379263,
0.394496,
-0.253297,
-0.274286
],
[
-0.127492,
-0.390764,
-0.020696,
-0.321776,
-0.116485,
-0.185538
]
],
"network.6.bias": [
0.080848,
-0.098159,
0.039224,
-0.129396,
-0.065954,
-0.009279
],
"network.8.weight": [
[
-0.437927,
-0.254993,
0.04059,
0.017601,
0.146846,
0.358406
],
[
0.73079,
-0.33901,
-0.579096,
-0.363953,
-0.073273,
0.012725
],
[
0.216835,
0.21113,
0.240983,
0.301471,
-0.280587,
0.025901
],
[
-0.374404,
0.698038,
-0.069974,
0.326188,
-0.063061,
-0.058988
],
[
-0.571517,
-0.289854,
-0.028659,
-0.273962,
0.062268,
-0.050399
],
[
-0.108319,
-0.210632,
-0.302723,
-0.227083,
-0.402952,
-0.225446
]
],
"network.8.bias": [
0.015591,
-0.157455,
-0.421867,
-0.223887,
-0.306636,
-0.231671
],
"network.10.weight": [
[
0.060625,
0.106769,
-0.362964,
-0.324806,
-0.342992,
-0.16097
],
[
-0.05047,
-0.800504,
-0.297592,
-0.287162,
0.169625,
0.312191
],
[
-0.24364,
0.877871,
0.585479,
0.165945,
0.022147,
-0.073678
],
[
0.052986,
-0.282511,
-0.248532,
-0.559466,
-0.384724,
-0.301359
],
[
0.198062,
-0.844801,
-0.103514,
-0.19519,
0.258176,
0.336121
],
[
0.078998,
0.060189,
-0.709356,
-0.49325,
-0.013622,
-0.261538
]
],
"network.10.bias": [
-0.287698,
0.497287,
-0.143066,
-0.495287,
0.4603,
-0.435993
],
"network.12.weight": [
[
0.180407,
0.633724,
-0.490073,
0.120352,
0.313839,
0.071619
]
],
"network.12.bias": [
0.389546
]
}
## Activation Signature
### 0
mean: [-0.043721, 1.873207, -1.752102, -0.560597, -3.048992, -3.355225]
std: [1.570120, 4.500696, 2.415620, 3.425946, 2.359146, 1.809910]
### 2
mean: [-2.036773, 4.545798, -2.584296, 2.462004, -1.391105, 3.194026]
std: [1.545835, 6.724718, 3.024247, 4.374554, 1.391483, 4.770607]
### 4
mean: [-4.074658, 10.273511, -2.458462, -3.808954, 9.267619, 7.323998]
std: [5.522596, 15.419110, 3.477510, 5.774837, 13.606426, 10.714685]
### 6
mean: [21.638462, -9.496088, -14.872290, -4.097639, -5.722716, -6.462213]
std: [31.920223, 13.859207, 21.974930, 5.921827, 8.337070, 9.598119]
### 8
mean: [-9.460484, 15.655711, 4.270109, -8.325424, -12.673375, -2.575534]
std: [13.978741, 23.326975, 6.921422, 11.951075, 18.242939, 3.457576]
### 10
mean: [-0.174807, -13.313107, 16.114828, -5.985449, -13.210180, -2.539837]
std: [0.023003, 20.728548, 24.521488, 8.306518, 20.421541, 3.495015]
### 12
mean: [-7.507326]
std: [12.017677]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
no_repeats
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -0.0437207892537117, "std": 1.570120096206665}, "1": {"mean": 1.8732067346572876, "std": 4.500695705413818}, "2": {"mean": -1.7521021366119385, "std": 2.4156196117401123}, "3": {"mean": -0.5605970621109009, "std": 3.425945997238159}, "4": {"mean": -3.0489919185638428, "std": 2.3591461181640625}, "5": {"mean": -3.355224609375, "std": 1.8099098205566406}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -2.036773443222046, "std": 1.545835018157959}, "1": {"mean": 4.545798301696777, "std": 6.72471809387207}, "2": {"mean": -2.5842957496643066, "std": 3.024247407913208}, "3": {"mean": 2.462003707885742, "std": 4.374553680419922}, "4": {"mean": -1.391105055809021, "std": 1.391483187675476}, "5": {"mean": 3.194025993347168, "std": 4.770606517791748}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -4.074658393859863, "std": 5.52259635925293}, "1": {"mean": 10.273510932922363, "std": 15.419110298156738}, "2": {"mean": -2.4584617614746094, "std": 3.4775102138519287}, "3": {"mean": -3.8089537620544434, "std": 5.774837493896484}, "4": {"mean": 9.267619132995605, "std": 13.606426239013672}, "5": {"mean": 7.32399845123291, "std": 10.714685440063477}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 21.63846206665039, "std": 31.920223236083984}, "1": {"mean": -9.496088027954102, "std": 13.859207153320312}, "2": {"mean": -14.872289657592773, "std": 21.974929809570312}, "3": {"mean": -4.0976386070251465, "std": 5.9218268394470215}, "4": {"mean": -5.722715854644775, "std": 8.337069511413574}, "5": {"mean": -6.462212562561035, "std": 9.598118782043457}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -9.46048355102539, "std": 13.978740692138672}, "1": {"mean": 15.65571117401123, "std": 23.326974868774414}, "2": {"mean": 4.270108699798584, "std": 6.921421527862549}, "3": {"mean": -8.325424194335938, "std": 11.951074600219727}, "4": {"mean": -12.673375129699707, "std": 18.242938995361328}, "5": {"mean": -2.5755343437194824, "std": 3.457576274871826}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -0.17480705678462982, "std": 0.023002706468105316}, "1": {"mean": -13.313106536865234, "std": 20.728548049926758}, "2": {"mean": 16.11482810974121, "std": 24.521488189697266}, "3": {"mean": -5.985449314117432, "std": 8.306517601013184}, "4": {"mean": -13.210180282592773, "std": 20.421541213989258}, "5": {"mean": -2.53983736038208, "std": 3.4950153827667236}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -7.507325649261475, "std": 12.017677307128906}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.777374, -0.014425, 0.14636, 0.10094, 0.29007], [2.109879, 0.461148, -0.220612, -0.147505, -0.62517], [0.359933, 0.694859, -1.123783, -0.314574, -0.23471], [0.777449, 0.424558, 0.090603, -0.03996, -1.669069], [-0.809717, -0.176397, -0.526771, -0.171434, 0.003931], [-0.293729, -0.409661, -0.148803, -0.412062, -0.298951]], "network.0.bias": [0.070019, -0.044541, -0.154686, -0.346775, -0.256762, -0.680019], "network.2.weight": [[-0.826304, -0.212637, -0.816441, -0.389692, 0.012806, -0.358548], [-0.081625, 1.251655, 0.850597, 0.997102, 0.172326, -0.052602], [-0.191447, -0.540626, -0.094484, -0.615085, 0.034549, 0.006391], [-0.294573, 0.82798, 0.96075, 0.464159, 0.050621, 0.425751], [-0.520236, -0.123747, -0.49018, -0.556721, 0.010189, -0.361966], [-0.088175, 0.843299, 0.944713, 0.748074, 0.279901, -0.019795]], "network.2.bias": [-0.495151, 0.397935, -0.558965, -0.060228, -0.157909, 0.256337], "network.4.weight": [[0.164717, -0.518841, -0.166381, -0.308837, 0.116415, -0.149717], [0.357222, 1.097519, -0.514641, 0.855791, 0.084802, 0.918839], [-0.217265, -0.270028, -0.226961, -0.098285, -0.396762, -0.260415], [0.289151, -0.172295, -0.408013, -0.598197, -0.695972, -0.431631], [0.448855, 1.182046, 0.533815, 0.463716, 0.155327, 0.771024], [0.201943, 0.832931, 0.06671, 0.720165, 0.278003, 0.426928]], "network.4.bias": [-0.431329, 0.114535, -0.142492, -0.084792, 0.220492, 0.293206], "network.6.weight": [[-0.599718, 0.875366, -0.227872, 0.226596, 0.862811, 0.623748], [-0.438921, -0.232001, 0.32822, -0.047312, -0.417966, -0.428853], [-0.275655, -0.327691, -0.003309, -0.172624, -0.67436, -0.723002], [-0.33558, -0.284219, -0.133168, 0.241042, -0.120132, 0.008877], [0.139506, -0.126581, 0.379263, 0.394496, -0.253297, -0.274286], [-0.127492, -0.390764, -0.020696, -0.321776, -0.116485, -0.185538]], "network.6.bias": [0.080848, -0.098159, 0.039224, -0.129396, -0.065954, -0.009279], "network.8.weight": [[-0.437927, -0.254993, 0.04059, 0.017601, 0.146846, 0.358406], [0.73079, -0.33901, -0.579096, -0.363953, -0.073273, 0.012725], [0.216835, 0.21113, 0.240983, 0.301471, -0.280587, 0.025901], [-0.374404, 0.698038, -0.069974, 0.326188, -0.063061, -0.058988], [-0.571517, -0.289854, -0.028659, -0.273962, 0.062268, -0.050399], [-0.108319, -0.210632, -0.302723, -0.227083, -0.402952, -0.225446]], "network.8.bias": [0.015591, -0.157455, -0.421867, -0.223887, -0.306636, -0.231671], "network.10.weight": [[0.060625, 0.106769, -0.362964, -0.324806, -0.342992, -0.16097], [-0.05047, -0.800504, -0.297592, -0.287162, 0.169625, 0.312191], [-0.24364, 0.877871, 0.585479, 0.165945, 0.022147, -0.073678], [0.052986, -0.282511, -0.248532, -0.559466, -0.384724, -0.301359], [0.198062, -0.844801, -0.103514, -0.19519, 0.258176, 0.336121], [0.078998, 0.060189, -0.709356, -0.49325, -0.013622, -0.261538]], "network.10.bias": [-0.287698, 0.497287, -0.143066, -0.495287, 0.4603, -0.435993], "network.12.weight": [[0.180407, 0.633724, -0.490073, 0.120352, 0.313839, 0.071619]], "network.12.bias": [0.389546]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7023701071739197, "train_acc": 0.465, "val_loss": 0.7157906889915466, "val_acc": 0.5}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.676950603723526, "train_acc": 0.575, "val_loss": 0.7114367485046387, "val_acc": 0.5}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6704665124416351, "train_acc": 0.575, "val_loss": 0.6854040026664734, "val_acc": 0.5}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6409766674041748, "train_acc": 0.575, "val_loss": 0.673373818397522, "val_acc": 0.5}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6482711732387543, "train_acc": 0.565, "val_loss": 0.6766025424003601, "val_acc": 0.6}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.6635142862796783, "train_acc": 0.61, "val_loss": 0.6023693084716797, "val_acc": 0.68}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.5760740637779236, "train_acc": 0.705, "val_loss": 0.6450631618499756, "val_acc": 0.6}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.5693967342376709, "train_acc": 0.71, "val_loss": 0.5576977729797363, "val_acc": 0.68}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.5207595825195312, "train_acc": 0.735, "val_loss": 0.5695995092391968, "val_acc": 0.68}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.5319321155548096, "train_acc": 0.72, "val_loss": 0.5602896213531494, "val_acc": 0.68}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.5112869143486023, "train_acc": 0.725, "val_loss": 0.5472913980484009, "val_acc": 0.74}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.526916891336441, "train_acc": 0.74, "val_loss": 0.5155994892120361, "val_acc": 0.76}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.4836771786212921, "train_acc": 0.78, "val_loss": 0.5318918824195862, "val_acc": 0.72}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.4566713571548462, "train_acc": 0.76, "val_loss": 0.5536360740661621, "val_acc": 0.72}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["no_repeats"], "degraded_stage": {"initial_val_loss": 0.7157906889915466, "final_val_loss": 0.673373818397522, "initial_val_acc": 0.5, "final_val_acc": 0.5, "best_val_acc": 0.5}, "improved_stage": {"initial_val_loss": 0.6766025424003601, "final_val_loss": 0.5536360740661621, "initial_val_acc": 0.6, "final_val_acc": 0.72, "best_val_acc": 0.76, "best_epoch": 11}, "improvement": 0.26, "first_improvement_epoch": 3}}
|
24
|
{"target_pattern": "alternating", "degraded_accuracy": 0.76, "improved_accuracy": 0.92, "improvement": 0.16000000000000003, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 5108, "learning_rate": 0.09399438147876214, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "alternating", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["alternating"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.357416,
-0.521615,
-0.005977,
-0.921485,
0.071411
],
[
-0.620079,
0.423195,
0.342277,
-0.266128,
0.074155
],
[
0.203546,
0.798306,
-1.19811,
0.546558,
0.384409
],
[
0.718319,
-1.027288,
0.185173,
-0.864122,
0.08085
],
[
-0.275712,
0.374405,
-0.040971,
-0.004105,
0.586567
],
[
-0.097369,
1.059829,
0.270179,
0.725856,
0.287397
],
[
0.757008,
0.227904,
0.44886,
0.040969,
0.682115
]
],
"network.0.bias": [
0.025029,
-0.191406,
0.843248,
0.404526,
-0.18685,
0.287285,
-0.088707
],
"network.2.weight": [
[
-0.204424,
-0.238612,
-0.993565,
0.462484,
0.086036,
-0.594076,
0.284098
],
[
0.499774,
-0.491396,
-1.134979,
0.430689,
-0.0842,
-0.379142,
0.073847
],
[
-0.275727,
-0.138952,
-0.620743,
-0.043537,
0.324295,
-0.222103,
-0.365292
],
[
0.134095,
-0.254966,
0.224969,
-0.266802,
-0.288947,
-0.607227,
0.005084
],
[
-0.385508,
-0.463006,
0.810816,
-0.241044,
-0.407293,
-0.077723,
-1.114775
],
[
0.32868,
-0.408873,
-1.032152,
0.695812,
-0.436939,
-0.66677,
0.387125
],
[
-0.135746,
-0.036637,
0.046541,
-0.281217,
-0.220386,
-0.353869,
-0.147781
]
],
"network.2.bias": [
-0.147187,
0.202748,
-0.238439,
-0.28135,
0.4597,
-0.080817,
-0.219633
],
"network.4.weight": [
[
0.12294,
-0.329875,
-0.000666,
-0.071394,
-0.021262,
-0.309953,
-0.003164
],
[
0.288844,
0.535509,
-0.361314,
0.064993,
0.845557,
0.667708,
-0.276264
],
[
0.147722,
-0.319952,
-0.206077,
-0.108191,
-0.072277,
-0.025449,
-0.105311
],
[
-0.207369,
-0.180212,
-0.471897,
-0.445843,
-0.059109,
-0.558728,
-0.258653
],
[
-0.026127,
-0.63229,
-0.103128,
-0.489297,
-0.071675,
-0.288652,
-0.241096
],
[
0.549444,
-0.027656,
0.085321,
0.071703,
0.866045,
0.506112,
-0.259399
],
[
-0.05458,
0.003947,
0.350835,
0.023495,
-0.229088,
-0.153787,
0.013404
]
],
"network.4.bias": [
-0.100524,
0.043596,
-0.466883,
-0.284867,
-0.828438,
-0.009412,
-0.330185
],
"network.6.weight": [
[
0.12318,
0.468304,
0.025322,
-0.576805,
0.402486,
0.381566,
0.081339
],
[
0.277267,
0.150673,
0.070657,
0.194681,
-0.297101,
-0.363211,
-0.335027
],
[
0.243067,
0.510741,
-0.519713,
-0.462302,
0.311233,
0.658523,
0.100012
],
[
-0.030949,
-0.492866,
-0.400193,
-0.122425,
-0.078171,
-0.325038,
0.068634
],
[
0.107064,
-0.397551,
-0.270609,
0.199174,
-0.484543,
-0.114446,
-0.121054
],
[
-0.259002,
0.536297,
0.168646,
-0.083065,
-0.082076,
0.959476,
0.024576
],
[
-0.115024,
-0.757178,
-0.315667,
-0.060669,
0.047651,
-0.193712,
-0.230322
]
],
"network.6.bias": [
-0.056042,
-0.363099,
-0.024646,
0.909723,
-0.753473,
-0.174375,
0.871171
],
"network.8.weight": [
[
0.351162,
0.31949,
0.469703,
-0.962084,
0.118623,
0.420637,
-0.810267
]
],
"network.8.bias": [
-0.646547
]
}
## Activation Signature
### 0
mean: [-2.381169, 0.043834, 1.694688, -1.940053, 0.781336, 4.572746, 3.135390]
std: [2.455554, 1.267458, 2.913752, 3.138015, 1.204577, 3.030231, 2.745335]
### 2
mean: [-4.025917, -3.874243, -3.600290, -3.037690, -2.378948, -4.443240, -2.554723]
std: [3.500467, 3.535741, 2.064795, 1.782385, 3.704869, 4.340138, 1.430928]
### 4
mean: [-0.213056, 0.700533, -0.528704, -0.497288, -1.014771, 0.578285, -0.480023]
std: [0.454408, 1.335769, 0.140209, 0.819063, 0.693586, 1.069522, 0.263212]
### 6
mean: [0.494829, -0.469639, 0.717679, 0.374654, -1.098800, 0.761588, 0.227628]
std: [1.027484, 0.194321, 1.377517, 1.000404, 0.651084, 1.731079, 1.214537]
### 8
mean: [-0.806306]
std: [2.255319]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
alternating
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.357416,
-0.521615,
-0.005977,
-0.921485,
0.071411
],
[
-0.620079,
0.423195,
0.342277,
-0.266128,
0.074155
],
[
0.203546,
0.798306,
-1.19811,
0.546558,
0.384409
],
[
0.718319,
-1.027288,
0.185173,
-0.864122,
0.08085
],
[
-0.275712,
0.374405,
-0.040971,
-0.004105,
0.586567
],
[
-0.097369,
1.059829,
0.270179,
0.725856,
0.287397
],
[
0.757008,
0.227904,
0.44886,
0.040969,
0.682115
]
],
"network.0.bias": [
0.025029,
-0.191406,
0.843248,
0.404526,
-0.18685,
0.287285,
-0.088707
],
"network.2.weight": [
[
-0.204424,
-0.238612,
-0.993565,
0.462484,
0.086036,
-0.594076,
0.284098
],
[
0.499774,
-0.491396,
-1.134979,
0.430689,
-0.0842,
-0.379142,
0.073847
],
[
-0.275727,
-0.138952,
-0.620743,
-0.043537,
0.324295,
-0.222103,
-0.365292
],
[
0.134095,
-0.254966,
0.224969,
-0.266802,
-0.288947,
-0.607227,
0.005084
],
[
-0.385508,
-0.463006,
0.810816,
-0.241044,
-0.407293,
-0.077723,
-1.114775
],
[
0.32868,
-0.408873,
-1.032152,
0.695812,
-0.436939,
-0.66677,
0.387125
],
[
-0.135746,
-0.036637,
0.046541,
-0.281217,
-0.220386,
-0.353869,
-0.147781
]
],
"network.2.bias": [
-0.147187,
0.202748,
-0.238439,
-0.28135,
0.4597,
-0.080817,
-0.219633
],
"network.4.weight": [
[
0.12294,
-0.329875,
-0.000666,
-0.071394,
-0.021262,
-0.309953,
-0.003164
],
[
0.288844,
0.535509,
-0.361314,
0.064993,
0.845557,
0.667708,
-0.276264
],
[
0.147722,
-0.319952,
-0.206077,
-0.108191,
-0.072277,
-0.025449,
-0.105311
],
[
-0.207369,
-0.180212,
-0.471897,
-0.445843,
-0.059109,
-0.558728,
-0.258653
],
[
-0.026127,
-0.63229,
-0.103128,
-0.489297,
-0.071675,
-0.288652,
-0.241096
],
[
0.549444,
-0.027656,
0.085321,
0.071703,
0.866045,
0.506112,
-0.259399
],
[
-0.05458,
0.003947,
0.350835,
0.023495,
-0.229088,
-0.153787,
0.013404
]
],
"network.4.bias": [
-0.100524,
0.043596,
-0.466883,
-0.284867,
-0.828438,
-0.009412,
-0.330185
],
"network.6.weight": [
[
0.12318,
0.468304,
0.025322,
-0.576805,
0.402486,
0.381566,
0.081339
],
[
0.277267,
0.150673,
0.070657,
0.194681,
-0.297101,
-0.363211,
-0.335027
],
[
0.243067,
0.510741,
-0.519713,
-0.462302,
0.311233,
0.658523,
0.100012
],
[
-0.030949,
-0.492866,
-0.400193,
-0.122425,
-0.078171,
-0.325038,
0.068634
],
[
0.107064,
-0.397551,
-0.270609,
0.199174,
-0.484543,
-0.114446,
-0.121054
],
[
-0.259002,
0.536297,
0.168646,
-0.083065,
-0.082076,
0.959476,
0.024576
],
[
-0.115024,
-0.757178,
-0.315667,
-0.060669,
0.047651,
-0.193712,
-0.230322
]
],
"network.6.bias": [
-0.056042,
-0.363099,
-0.024646,
0.909723,
-0.753473,
-0.174375,
0.871171
],
"network.8.weight": [
[
0.351162,
0.31949,
0.469703,
-0.962084,
0.118623,
0.420637,
-0.810267
]
],
"network.8.bias": [
-0.646547
]
}
## Activation Signature
### 0
mean: [-2.381169, 0.043834, 1.694688, -1.940053, 0.781336, 4.572746, 3.135390]
std: [2.455554, 1.267458, 2.913752, 3.138015, 1.204577, 3.030231, 2.745335]
### 2
mean: [-4.025917, -3.874243, -3.600290, -3.037690, -2.378948, -4.443240, -2.554723]
std: [3.500467, 3.535741, 2.064795, 1.782385, 3.704869, 4.340138, 1.430928]
### 4
mean: [-0.213056, 0.700533, -0.528704, -0.497288, -1.014771, 0.578285, -0.480023]
std: [0.454408, 1.335769, 0.140209, 0.819063, 0.693586, 1.069522, 0.263212]
### 6
mean: [0.494829, -0.469639, 0.717679, 0.374654, -1.098800, 0.761588, 0.227628]
std: [1.027484, 0.194321, 1.377517, 1.000404, 0.651084, 1.731079, 1.214537]
### 8
mean: [-0.806306]
std: [2.255319]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
alternating
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -2.381168842315674, "std": 2.4555535316467285}, "1": {"mean": 0.04383370280265808, "std": 1.2674577236175537}, "2": {"mean": 1.694688081741333, "std": 2.9137518405914307}, "3": {"mean": -1.9400534629821777, "std": 3.138014793395996}, "4": {"mean": 0.7813363075256348, "std": 1.2045767307281494}, "5": {"mean": 4.5727458000183105, "std": 3.0302305221557617}, "6": {"mean": 3.135389804840088, "std": 2.7453348636627197}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -4.025917053222656, "std": 3.50046706199646}, "1": {"mean": -3.8742427825927734, "std": 3.535741090774536}, "2": {"mean": -3.600290298461914, "std": 2.0647952556610107}, "3": {"mean": -3.0376899242401123, "std": 1.7823846340179443}, "4": {"mean": -2.378948211669922, "std": 3.704868793487549}, "5": {"mean": -4.443240165710449, "std": 4.3401384353637695}, "6": {"mean": -2.554722547531128, "std": 1.4309276342391968}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.2130555510520935, "std": 0.4544082283973694}, "1": {"mean": 0.7005334496498108, "std": 1.3357691764831543}, "2": {"mean": -0.5287039279937744, "std": 0.14020869135856628}, "3": {"mean": -0.49728843569755554, "std": 0.8190625905990601}, "4": {"mean": -1.014770746231079, "std": 0.693585991859436}, "5": {"mean": 0.578285276889801, "std": 1.0695222616195679}, "6": {"mean": -0.4800233542919159, "std": 0.26321208477020264}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.4948292076587677, "std": 1.0274842977523804}, "1": {"mean": -0.46963855624198914, "std": 0.1943206638097763}, "2": {"mean": 0.7176788449287415, "std": 1.3775169849395752}, "3": {"mean": 0.37465426325798035, "std": 1.000403642654419}, "4": {"mean": -1.0987995862960815, "std": 0.6510838270187378}, "5": {"mean": 0.7615883946418762, "std": 1.7310786247253418}, "6": {"mean": 0.22762775421142578, "std": 1.2145366668701172}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.8063055276870728, "std": 2.2553186416625977}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.357416, -0.521615, -0.005977, -0.921485, 0.071411], [-0.620079, 0.423195, 0.342277, -0.266128, 0.074155], [0.203546, 0.798306, -1.19811, 0.546558, 0.384409], [0.718319, -1.027288, 0.185173, -0.864122, 0.08085], [-0.275712, 0.374405, -0.040971, -0.004105, 0.586567], [-0.097369, 1.059829, 0.270179, 0.725856, 0.287397], [0.757008, 0.227904, 0.44886, 0.040969, 0.682115]], "network.0.bias": [0.025029, -0.191406, 0.843248, 0.404526, -0.18685, 0.287285, -0.088707], "network.2.weight": [[-0.204424, -0.238612, -0.993565, 0.462484, 0.086036, -0.594076, 0.284098], [0.499774, -0.491396, -1.134979, 0.430689, -0.0842, -0.379142, 0.073847], [-0.275727, -0.138952, -0.620743, -0.043537, 0.324295, -0.222103, -0.365292], [0.134095, -0.254966, 0.224969, -0.266802, -0.288947, -0.607227, 0.005084], [-0.385508, -0.463006, 0.810816, -0.241044, -0.407293, -0.077723, -1.114775], [0.32868, -0.408873, -1.032152, 0.695812, -0.436939, -0.66677, 0.387125], [-0.135746, -0.036637, 0.046541, -0.281217, -0.220386, -0.353869, -0.147781]], "network.2.bias": [-0.147187, 0.202748, -0.238439, -0.28135, 0.4597, -0.080817, -0.219633], "network.4.weight": [[0.12294, -0.329875, -0.000666, -0.071394, -0.021262, -0.309953, -0.003164], [0.288844, 0.535509, -0.361314, 0.064993, 0.845557, 0.667708, -0.276264], [0.147722, -0.319952, -0.206077, -0.108191, -0.072277, -0.025449, -0.105311], [-0.207369, -0.180212, -0.471897, -0.445843, -0.059109, -0.558728, -0.258653], [-0.026127, -0.63229, -0.103128, -0.489297, -0.071675, -0.288652, -0.241096], [0.549444, -0.027656, 0.085321, 0.071703, 0.866045, 0.506112, -0.259399], [-0.05458, 0.003947, 0.350835, 0.023495, -0.229088, -0.153787, 0.013404]], "network.4.bias": [-0.100524, 0.043596, -0.466883, -0.284867, -0.828438, -0.009412, -0.330185], "network.6.weight": [[0.12318, 0.468304, 0.025322, -0.576805, 0.402486, 0.381566, 0.081339], [0.277267, 0.150673, 0.070657, 0.194681, -0.297101, -0.363211, -0.335027], [0.243067, 0.510741, -0.519713, -0.462302, 0.311233, 0.658523, 0.100012], [-0.030949, -0.492866, -0.400193, -0.122425, -0.078171, -0.325038, 0.068634], [0.107064, -0.397551, -0.270609, 0.199174, -0.484543, -0.114446, -0.121054], [-0.259002, 0.536297, 0.168646, -0.083065, -0.082076, 0.959476, 0.024576], [-0.115024, -0.757178, -0.315667, -0.060669, 0.047651, -0.193712, -0.230322]], "network.6.bias": [-0.056042, -0.363099, -0.024646, 0.909723, -0.753473, -0.174375, 0.871171], "network.8.weight": [[0.351162, 0.31949, 0.469703, -0.962084, 0.118623, 0.420637, -0.810267]], "network.8.bias": [-0.646547]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7021216452121735, "train_acc": 0.465, "val_loss": 0.6868687272071838, "val_acc": 0.82}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6724567711353302, "train_acc": 0.78, "val_loss": 0.6092529892921448, "val_acc": 0.76}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.5787536799907684, "train_acc": 0.78, "val_loss": 0.574521005153656, "val_acc": 0.82}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.4371338337659836, "train_acc": 0.88, "val_loss": 0.5149170160293579, "val_acc": 0.86}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.33482562005519867, "train_acc": 0.94, "val_loss": 0.6455367207527161, "val_acc": 0.88}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.27670836448669434, "train_acc": 0.94, "val_loss": 0.3214164078235626, "val_acc": 0.92}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.1910673826932907, "train_acc": 0.95, "val_loss": 0.2675555944442749, "val_acc": 0.88}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.26979417353868484, "train_acc": 0.935, "val_loss": 0.2635820508003235, "val_acc": 0.9}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.18188771232962608, "train_acc": 0.945, "val_loss": 0.22070716321468353, "val_acc": 0.9}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.14884470403194427, "train_acc": 0.96, "val_loss": 0.15416985750198364, "val_acc": 0.92}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.13152959197759628, "train_acc": 0.955, "val_loss": 0.14698544144630432, "val_acc": 0.9}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.12438240274786949, "train_acc": 0.955, "val_loss": 0.17997869849205017, "val_acc": 0.92}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["alternating"], "degraded_stage": {"initial_val_loss": 0.6868687272071838, "final_val_loss": 0.6092529892921448, "initial_val_acc": 0.82, "final_val_acc": 0.76, "best_val_acc": 0.76}, "improved_stage": {"initial_val_loss": 0.574521005153656, "final_val_loss": 0.17997869849205017, "initial_val_acc": 0.82, "final_val_acc": 0.92, "best_val_acc": 0.92, "best_epoch": 5}, "improvement": 0.16000000000000003, "first_improvement_epoch": 1}}
|
25
|
{"target_pattern": "has_majority", "degraded_accuracy": 0.54, "improved_accuracy": 0.88, "improvement": 0.33999999999999997, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 5499, "learning_rate": 0.07940362044088127, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "has_majority", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["has_majority"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 6
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.0622,
-0.710535,
0.335045,
-0.28516,
-0.150208
],
[
-0.828388,
-0.024794,
-0.972341,
0.316296,
-0.114149
],
[
0.048238,
-0.12466,
0.754951,
-0.676893,
-0.145158
],
[
0.275845,
-1.105747,
-0.008781,
0.513407,
-0.437241
],
[
0.626246,
0.246624,
0.366609,
0.904447,
0.283753
],
[
0.284558,
0.251692,
-1.415737,
0.241936,
0.327001
]
],
"network.0.bias": [
0.103322,
0.108684,
-0.148698,
0.304599,
-0.196708,
-0.260715
],
"network.2.weight": [
[
0.00432,
-0.182883,
-0.070799,
-0.403895,
-0.503295,
0.009871
],
[
-0.178151,
-0.42574,
-0.327257,
-0.62768,
0.557033,
-0.904124
],
[
-0.123864,
-0.506041,
-0.109108,
-0.651605,
-0.143884,
-0.359163
],
[
0.384852,
-0.251286,
0.643052,
0.44572,
-0.217396,
-0.254033
],
[
-0.1785,
-0.126618,
-0.496147,
-0.021927,
-0.514451,
-0.077307
],
[
-0.39646,
-0.224014,
0.062784,
-0.753641,
0.405109,
-0.376425
]
],
"network.2.bias": [
-0.051511,
-0.329099,
-0.627805,
-0.196193,
-0.363413,
-0.518895
],
"network.4.weight": [
[
-0.17908,
-0.559958,
0.221123,
-0.141698,
0.407088,
-0.263241
],
[
0.027199,
0.546176,
-0.073578,
-0.168225,
-0.119118,
0.861882
],
[
0.232297,
0.789279,
-0.196887,
-0.921148,
-0.011656,
0.55312
],
[
-0.273304,
0.456309,
0.194313,
-0.067819,
-0.191332,
0.530911
],
[
-0.007651,
0.178525,
-0.195779,
-0.316602,
-0.343303,
0.302691
],
[
0.106604,
-0.138179,
-0.473473,
-0.191195,
0.222463,
-0.176655
]
],
"network.4.bias": [
-0.68805,
-0.375098,
-0.153837,
-0.394662,
-0.21538,
-0.595859
],
"network.6.weight": [
[
-0.251335,
0.537567,
0.11204,
-0.054503,
-0.00879,
0.473222
],
[
0.355222,
0.106307,
0.004931,
-0.392323,
-0.001006,
0.153506
],
[
0.35573,
0.152365,
-0.0845,
-0.057909,
-0.135135,
-0.300213
],
[
0.022586,
0.002519,
-0.376525,
-0.295571,
-0.197852,
-0.027923
],
[
-0.258438,
0.70881,
0.892257,
0.605276,
0.311288,
-0.26133
],
[
-0.570358,
0.358302,
0.641746,
0.012916,
0.03786,
0.445124
]
],
"network.6.bias": [
-0.437009,
-0.302796,
-0.448052,
0.793664,
-0.251092,
-0.268128
],
"network.8.weight": [
[
-0.190471,
-0.23299,
0.237286,
0.829367,
-0.510735,
-0.349313
]
],
"network.8.bias": [
0.467186
]
}
## Activation Signature
### 0
mean: [-1.365729, -2.460996, -0.369404, -0.823677, 4.087329, -1.483157]
std: [1.660649, 3.047088, 2.080877, 1.902581, 2.887526, 2.633937]
### 2
mean: [-2.335933, 1.102778, -1.757462, -0.620616, -2.858249, 0.677135]
std: [1.483971, 1.721211, 0.777635, 1.194709, 1.559106, 1.259667]
### 4
mean: [-1.651452, 1.019766, 1.124232, 0.623547, 0.203419, -0.962467]
std: [1.147578, 1.799613, 2.007912, 1.291426, 0.663770, 0.403738]
### 6
mean: [0.296820, -0.478971, -0.471317, 0.002647, 2.335852, 1.020157]
std: [1.034871, 0.274273, 0.042875, 1.123839, 3.665291, 1.786637]
### 8
mean: [-0.911689]
std: [2.835560]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
has_majority
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 6
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.0622,
-0.710535,
0.335045,
-0.28516,
-0.150208
],
[
-0.828388,
-0.024794,
-0.972341,
0.316296,
-0.114149
],
[
0.048238,
-0.12466,
0.754951,
-0.676893,
-0.145158
],
[
0.275845,
-1.105747,
-0.008781,
0.513407,
-0.437241
],
[
0.626246,
0.246624,
0.366609,
0.904447,
0.283753
],
[
0.284558,
0.251692,
-1.415737,
0.241936,
0.327001
]
],
"network.0.bias": [
0.103322,
0.108684,
-0.148698,
0.304599,
-0.196708,
-0.260715
],
"network.2.weight": [
[
0.00432,
-0.182883,
-0.070799,
-0.403895,
-0.503295,
0.009871
],
[
-0.178151,
-0.42574,
-0.327257,
-0.62768,
0.557033,
-0.904124
],
[
-0.123864,
-0.506041,
-0.109108,
-0.651605,
-0.143884,
-0.359163
],
[
0.384852,
-0.251286,
0.643052,
0.44572,
-0.217396,
-0.254033
],
[
-0.1785,
-0.126618,
-0.496147,
-0.021927,
-0.514451,
-0.077307
],
[
-0.39646,
-0.224014,
0.062784,
-0.753641,
0.405109,
-0.376425
]
],
"network.2.bias": [
-0.051511,
-0.329099,
-0.627805,
-0.196193,
-0.363413,
-0.518895
],
"network.4.weight": [
[
-0.17908,
-0.559958,
0.221123,
-0.141698,
0.407088,
-0.263241
],
[
0.027199,
0.546176,
-0.073578,
-0.168225,
-0.119118,
0.861882
],
[
0.232297,
0.789279,
-0.196887,
-0.921148,
-0.011656,
0.55312
],
[
-0.273304,
0.456309,
0.194313,
-0.067819,
-0.191332,
0.530911
],
[
-0.007651,
0.178525,
-0.195779,
-0.316602,
-0.343303,
0.302691
],
[
0.106604,
-0.138179,
-0.473473,
-0.191195,
0.222463,
-0.176655
]
],
"network.4.bias": [
-0.68805,
-0.375098,
-0.153837,
-0.394662,
-0.21538,
-0.595859
],
"network.6.weight": [
[
-0.251335,
0.537567,
0.11204,
-0.054503,
-0.00879,
0.473222
],
[
0.355222,
0.106307,
0.004931,
-0.392323,
-0.001006,
0.153506
],
[
0.35573,
0.152365,
-0.0845,
-0.057909,
-0.135135,
-0.300213
],
[
0.022586,
0.002519,
-0.376525,
-0.295571,
-0.197852,
-0.027923
],
[
-0.258438,
0.70881,
0.892257,
0.605276,
0.311288,
-0.26133
],
[
-0.570358,
0.358302,
0.641746,
0.012916,
0.03786,
0.445124
]
],
"network.6.bias": [
-0.437009,
-0.302796,
-0.448052,
0.793664,
-0.251092,
-0.268128
],
"network.8.weight": [
[
-0.190471,
-0.23299,
0.237286,
0.829367,
-0.510735,
-0.349313
]
],
"network.8.bias": [
0.467186
]
}
## Activation Signature
### 0
mean: [-1.365729, -2.460996, -0.369404, -0.823677, 4.087329, -1.483157]
std: [1.660649, 3.047088, 2.080877, 1.902581, 2.887526, 2.633937]
### 2
mean: [-2.335933, 1.102778, -1.757462, -0.620616, -2.858249, 0.677135]
std: [1.483971, 1.721211, 0.777635, 1.194709, 1.559106, 1.259667]
### 4
mean: [-1.651452, 1.019766, 1.124232, 0.623547, 0.203419, -0.962467]
std: [1.147578, 1.799613, 2.007912, 1.291426, 0.663770, 0.403738]
### 6
mean: [0.296820, -0.478971, -0.471317, 0.002647, 2.335852, 1.020157]
std: [1.034871, 0.274273, 0.042875, 1.123839, 3.665291, 1.786637]
### 8
mean: [-0.911689]
std: [2.835560]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
has_majority
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.3657292127609253, "std": 1.6606488227844238}, "1": {"mean": -2.46099591255188, "std": 3.047088146209717}, "2": {"mean": -0.36940431594848633, "std": 2.0808773040771484}, "3": {"mean": -0.8236773014068604, "std": 1.902580738067627}, "4": {"mean": 4.087329387664795, "std": 2.8875255584716797}, "5": {"mean": -1.4831573963165283, "std": 2.633937358856201}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -2.335932731628418, "std": 1.4839706420898438}, "1": {"mean": 1.1027776002883911, "std": 1.7212109565734863}, "2": {"mean": -1.7574620246887207, "std": 0.7776345610618591}, "3": {"mean": -0.6206157207489014, "std": 1.1947094202041626}, "4": {"mean": -2.8582489490509033, "std": 1.5591061115264893}, "5": {"mean": 0.6771349310874939, "std": 1.2596668004989624}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -1.651451826095581, "std": 1.1475783586502075}, "1": {"mean": 1.019766092300415, "std": 1.7996127605438232}, "2": {"mean": 1.1242320537567139, "std": 2.0079119205474854}, "3": {"mean": 0.6235470175743103, "std": 1.291426420211792}, "4": {"mean": 0.203419491648674, "std": 0.663769543170929}, "5": {"mean": -0.9624669551849365, "std": 0.40373826026916504}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.29681989550590515, "std": 1.0348705053329468}, "1": {"mean": -0.4789709448814392, "std": 0.2742726504802704}, "2": {"mean": -0.4713168740272522, "std": 0.042874764651060104}, "3": {"mean": 0.0026470052544027567, "std": 1.1238391399383545}, "4": {"mean": 2.3358521461486816, "std": 3.665290594100952}, "5": {"mean": 1.0201570987701416, "std": 1.7866365909576416}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.9116889834403992, "std": 2.8355600833892822}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.0622, -0.710535, 0.335045, -0.28516, -0.150208], [-0.828388, -0.024794, -0.972341, 0.316296, -0.114149], [0.048238, -0.12466, 0.754951, -0.676893, -0.145158], [0.275845, -1.105747, -0.008781, 0.513407, -0.437241], [0.626246, 0.246624, 0.366609, 0.904447, 0.283753], [0.284558, 0.251692, -1.415737, 0.241936, 0.327001]], "network.0.bias": [0.103322, 0.108684, -0.148698, 0.304599, -0.196708, -0.260715], "network.2.weight": [[0.00432, -0.182883, -0.070799, -0.403895, -0.503295, 0.009871], [-0.178151, -0.42574, -0.327257, -0.62768, 0.557033, -0.904124], [-0.123864, -0.506041, -0.109108, -0.651605, -0.143884, -0.359163], [0.384852, -0.251286, 0.643052, 0.44572, -0.217396, -0.254033], [-0.1785, -0.126618, -0.496147, -0.021927, -0.514451, -0.077307], [-0.39646, -0.224014, 0.062784, -0.753641, 0.405109, -0.376425]], "network.2.bias": [-0.051511, -0.329099, -0.627805, -0.196193, -0.363413, -0.518895], "network.4.weight": [[-0.17908, -0.559958, 0.221123, -0.141698, 0.407088, -0.263241], [0.027199, 0.546176, -0.073578, -0.168225, -0.119118, 0.861882], [0.232297, 0.789279, -0.196887, -0.921148, -0.011656, 0.55312], [-0.273304, 0.456309, 0.194313, -0.067819, -0.191332, 0.530911], [-0.007651, 0.178525, -0.195779, -0.316602, -0.343303, 0.302691], [0.106604, -0.138179, -0.473473, -0.191195, 0.222463, -0.176655]], "network.4.bias": [-0.68805, -0.375098, -0.153837, -0.394662, -0.21538, -0.595859], "network.6.weight": [[-0.251335, 0.537567, 0.11204, -0.054503, -0.00879, 0.473222], [0.355222, 0.106307, 0.004931, -0.392323, -0.001006, 0.153506], [0.35573, 0.152365, -0.0845, -0.057909, -0.135135, -0.300213], [0.022586, 0.002519, -0.376525, -0.295571, -0.197852, -0.027923], [-0.258438, 0.70881, 0.892257, 0.605276, 0.311288, -0.26133], [-0.570358, 0.358302, 0.641746, 0.012916, 0.03786, 0.445124]], "network.6.bias": [-0.437009, -0.302796, -0.448052, 0.793664, -0.251092, -0.268128], "network.8.weight": [[-0.190471, -0.23299, 0.237286, 0.829367, -0.510735, -0.349313]], "network.8.bias": [0.467186]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6861990988254547, "train_acc": 0.57, "val_loss": 0.6828721761703491, "val_acc": 0.54}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6771780252456665, "train_acc": 0.57, "val_loss": 0.6836917400360107, "val_acc": 0.54}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6537398993968964, "train_acc": 0.57, "val_loss": 0.6396022439002991, "val_acc": 0.54}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.6348343193531036, "train_acc": 0.49, "val_loss": 0.6517518758773804, "val_acc": 0.76}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.5937905609607697, "train_acc": 0.73, "val_loss": 0.5816473960876465, "val_acc": 0.72}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.5470808148384094, "train_acc": 0.73, "val_loss": 0.6189031004905701, "val_acc": 0.78}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.49332547187805176, "train_acc": 0.77, "val_loss": 0.4907083809375763, "val_acc": 0.74}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.4530288428068161, "train_acc": 0.81, "val_loss": 0.48747092485427856, "val_acc": 0.88}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.43413354456424713, "train_acc": 0.815, "val_loss": 0.43268540501594543, "val_acc": 0.86}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.3925091177225113, "train_acc": 0.83, "val_loss": 0.38571158051490784, "val_acc": 0.86}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.4046955704689026, "train_acc": 0.825, "val_loss": 0.3830999732017517, "val_acc": 0.84}, {"stage": "improved", "epoch": 8, "global_epoch": 11, "train_loss": 0.39945483207702637, "train_acc": 0.825, "val_loss": 0.37059101462364197, "val_acc": 0.8}, {"stage": "improved", "epoch": 9, "global_epoch": 12, "train_loss": 0.4206164628267288, "train_acc": 0.83, "val_loss": 0.3889116048812866, "val_acc": 0.88}], "summary": {"total_epochs": 13, "degraded_epochs": 3, "improved_epochs": 10, "patterns": ["has_majority"], "degraded_stage": {"initial_val_loss": 0.6828721761703491, "final_val_loss": 0.6396022439002991, "initial_val_acc": 0.54, "final_val_acc": 0.54, "best_val_acc": 0.54}, "improved_stage": {"initial_val_loss": 0.6517518758773804, "final_val_loss": 0.3889116048812866, "initial_val_acc": 0.76, "final_val_acc": 0.88, "best_val_acc": 0.88, "best_epoch": 7}, "improvement": 0.33999999999999997, "first_improvement_epoch": 2}}
|
26
|
{"target_pattern": "ends_with", "degraded_accuracy": 0.4, "improved_accuracy": 0.98, "improvement": 0.58, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 8389, "learning_rate": 0.07569666252355549, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "ends_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["ends_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.534601,
-0.133141,
-0.290869,
-0.269731,
1.439932
],
[
-0.85496,
0.035876,
0.038198,
0.02681,
1.166175
],
[
-0.472872,
-0.177301,
-0.410154,
-0.375064,
-0.266371
],
[
-1.109965,
-0.148037,
0.247655,
-0.034201,
0.653496
],
[
0.90343,
-0.122573,
-0.212362,
-0.314708,
0.589131
]
],
"network.0.bias": [
-0.001968,
0.35978,
-0.033742,
0.455524,
-0.361827
],
"network.2.weight": [
[
-1.045689,
-0.179125,
-0.485044,
-0.285838,
-0.442647
],
[
0.87817,
1.029643,
0.313131,
0.688713,
0.913916
],
[
0.877099,
1.087847,
0.168698,
0.26924,
0.937156
],
[
-0.251139,
-0.097611,
0.053992,
-0.05545,
0.098996
],
[
0.366001,
1.05474,
0.064072,
0.420962,
0.984757
]
],
"network.2.bias": [
-0.223175,
-0.288415,
-0.042869,
-0.327577,
-0.225892
],
"network.4.weight": [
[
-0.011629,
0.73881,
0.191067,
0.012986,
0.762385
],
[
-0.148892,
-0.010541,
-0.213267,
-0.29935,
-0.440456
],
[
-0.084035,
1.029434,
0.878879,
0.244411,
0.435314
],
[
0.235509,
-0.340166,
-0.303294,
-0.024437,
-0.228489
],
[
-0.01365,
-0.712751,
-0.629699,
-0.003346,
-0.798856
]
],
"network.4.bias": [
-0.178412,
-0.636267,
-0.054435,
-0.128096,
-0.127913
],
"network.6.weight": [
[
-0.111424,
0.075213,
-0.16277,
-0.183742,
-0.17011
],
[
0.77211,
0.321973,
0.49206,
0.298669,
0.227524
],
[
-0.26424,
0.114212,
-0.316632,
-0.184158,
-0.264864
],
[
0.61018,
0.585168,
0.79072,
0.301965,
-0.028276
],
[
0.453047,
-0.022901,
0.614531,
-0.198654,
-0.314308
]
],
"network.6.bias": [
-0.355675,
0.192702,
-0.352489,
0.10091,
-0.341603
],
"network.8.weight": [
[
0.070058,
0.486096,
-0.093296,
0.703101,
0.687306
],
[
-0.234778,
0.122574,
-0.123108,
-0.39659,
0.253556
],
[
0.125735,
0.379943,
0.438834,
0.605767,
0.504577
],
[
-0.394372,
-0.415601,
-0.135541,
-0.348393,
-0.105384
],
[
0.045097,
0.127505,
0.046097,
0.352597,
1.16365
]
],
"network.8.bias": [
-0.112016,
-0.412748,
-0.348055,
-0.256773,
0.066682
],
"network.10.weight": [
[
0.493113,
0.380736,
0.679294,
-0.019459,
0.210457
],
[
-0.301605,
0.432335,
-0.780288,
-0.08921,
-0.10936
],
[
-0.278039,
-0.309144,
-0.0768,
0.107603,
-0.856144
],
[
-0.624869,
-0.138469,
-0.711115,
-0.086741,
-0.505044
],
[
-0.241956,
-0.442229,
-0.166251,
0.070853,
-0.695022
]
],
"network.10.bias": [
-0.597022,
-0.089899,
-0.048376,
0.086182,
-0.181789
],
"network.12.weight": [
[
-0.380713,
-0.244876,
-0.217789,
-0.076699,
-0.246167
]
],
"network.12.bias": [
0.426473
]
}
## Activation Signature
### 0
mean: [1.010601, 0.936770, -2.934883, 0.054438, 0.147206]
std: [2.920595, 2.502120, 1.971968, 2.396286, 2.239689]
### 2
mean: [-2.836969, 4.058804, 4.037093, -0.827009, 3.159892]
std: [3.406331, 5.027917, 4.854111, 0.703910, 3.761384]
### 4
mean: [6.025775, -2.938945, 9.074896, -3.465504, -8.113522]
std: [7.484700, 2.736543, 11.047771, 4.030069, 9.615277]
### 6
mean: [-2.506125, 9.321628, -4.822431, 10.963551, 7.972778]
std: [2.630513, 11.205541, 5.472018, 13.293711, 10.173342]
### 8
mean: [17.623747, -1.590624, 13.869873, -8.793189, 14.426125]
std: [21.772717, 1.324250, 17.433855, 10.358529, 17.931978]
### 10
mean: [20.560957, -17.816563, -18.365601, -28.085396, -16.780678]
std: [26.345243, 22.122293, 22.744038, 35.050858, 20.627632]
### 12
mean: [-7.416430]
std: [10.018027]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
ends_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.534601,
-0.133141,
-0.290869,
-0.269731,
1.439932
],
[
-0.85496,
0.035876,
0.038198,
0.02681,
1.166175
],
[
-0.472872,
-0.177301,
-0.410154,
-0.375064,
-0.266371
],
[
-1.109965,
-0.148037,
0.247655,
-0.034201,
0.653496
],
[
0.90343,
-0.122573,
-0.212362,
-0.314708,
0.589131
]
],
"network.0.bias": [
-0.001968,
0.35978,
-0.033742,
0.455524,
-0.361827
],
"network.2.weight": [
[
-1.045689,
-0.179125,
-0.485044,
-0.285838,
-0.442647
],
[
0.87817,
1.029643,
0.313131,
0.688713,
0.913916
],
[
0.877099,
1.087847,
0.168698,
0.26924,
0.937156
],
[
-0.251139,
-0.097611,
0.053992,
-0.05545,
0.098996
],
[
0.366001,
1.05474,
0.064072,
0.420962,
0.984757
]
],
"network.2.bias": [
-0.223175,
-0.288415,
-0.042869,
-0.327577,
-0.225892
],
"network.4.weight": [
[
-0.011629,
0.73881,
0.191067,
0.012986,
0.762385
],
[
-0.148892,
-0.010541,
-0.213267,
-0.29935,
-0.440456
],
[
-0.084035,
1.029434,
0.878879,
0.244411,
0.435314
],
[
0.235509,
-0.340166,
-0.303294,
-0.024437,
-0.228489
],
[
-0.01365,
-0.712751,
-0.629699,
-0.003346,
-0.798856
]
],
"network.4.bias": [
-0.178412,
-0.636267,
-0.054435,
-0.128096,
-0.127913
],
"network.6.weight": [
[
-0.111424,
0.075213,
-0.16277,
-0.183742,
-0.17011
],
[
0.77211,
0.321973,
0.49206,
0.298669,
0.227524
],
[
-0.26424,
0.114212,
-0.316632,
-0.184158,
-0.264864
],
[
0.61018,
0.585168,
0.79072,
0.301965,
-0.028276
],
[
0.453047,
-0.022901,
0.614531,
-0.198654,
-0.314308
]
],
"network.6.bias": [
-0.355675,
0.192702,
-0.352489,
0.10091,
-0.341603
],
"network.8.weight": [
[
0.070058,
0.486096,
-0.093296,
0.703101,
0.687306
],
[
-0.234778,
0.122574,
-0.123108,
-0.39659,
0.253556
],
[
0.125735,
0.379943,
0.438834,
0.605767,
0.504577
],
[
-0.394372,
-0.415601,
-0.135541,
-0.348393,
-0.105384
],
[
0.045097,
0.127505,
0.046097,
0.352597,
1.16365
]
],
"network.8.bias": [
-0.112016,
-0.412748,
-0.348055,
-0.256773,
0.066682
],
"network.10.weight": [
[
0.493113,
0.380736,
0.679294,
-0.019459,
0.210457
],
[
-0.301605,
0.432335,
-0.780288,
-0.08921,
-0.10936
],
[
-0.278039,
-0.309144,
-0.0768,
0.107603,
-0.856144
],
[
-0.624869,
-0.138469,
-0.711115,
-0.086741,
-0.505044
],
[
-0.241956,
-0.442229,
-0.166251,
0.070853,
-0.695022
]
],
"network.10.bias": [
-0.597022,
-0.089899,
-0.048376,
0.086182,
-0.181789
],
"network.12.weight": [
[
-0.380713,
-0.244876,
-0.217789,
-0.076699,
-0.246167
]
],
"network.12.bias": [
0.426473
]
}
## Activation Signature
### 0
mean: [1.010601, 0.936770, -2.934883, 0.054438, 0.147206]
std: [2.920595, 2.502120, 1.971968, 2.396286, 2.239689]
### 2
mean: [-2.836969, 4.058804, 4.037093, -0.827009, 3.159892]
std: [3.406331, 5.027917, 4.854111, 0.703910, 3.761384]
### 4
mean: [6.025775, -2.938945, 9.074896, -3.465504, -8.113522]
std: [7.484700, 2.736543, 11.047771, 4.030069, 9.615277]
### 6
mean: [-2.506125, 9.321628, -4.822431, 10.963551, 7.972778]
std: [2.630513, 11.205541, 5.472018, 13.293711, 10.173342]
### 8
mean: [17.623747, -1.590624, 13.869873, -8.793189, 14.426125]
std: [21.772717, 1.324250, 17.433855, 10.358529, 17.931978]
### 10
mean: [20.560957, -17.816563, -18.365601, -28.085396, -16.780678]
std: [26.345243, 22.122293, 22.744038, 35.050858, 20.627632]
### 12
mean: [-7.416430]
std: [10.018027]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
ends_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.0106010437011719, "std": 2.920595407485962}, "1": {"mean": 0.9367702007293701, "std": 2.502119541168213}, "2": {"mean": -2.934882879257202, "std": 1.971968173980713}, "3": {"mean": 0.05443776398897171, "std": 2.3962862491607666}, "4": {"mean": 0.1472064107656479, "std": 2.2396888732910156}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -2.8369691371917725, "std": 3.4063308238983154}, "1": {"mean": 4.058804035186768, "std": 5.027917385101318}, "2": {"mean": 4.037093162536621, "std": 4.8541107177734375}, "3": {"mean": -0.8270085453987122, "std": 0.7039096355438232}, "4": {"mean": 3.1598918437957764, "std": 3.7613840103149414}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 6.025774955749512, "std": 7.484699726104736}, "1": {"mean": -2.9389445781707764, "std": 2.7365434169769287}, "2": {"mean": 9.074895858764648, "std": 11.047771453857422}, "3": {"mean": -3.465503692626953, "std": 4.030068874359131}, "4": {"mean": -8.113521575927734, "std": 9.615277290344238}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -2.506124973297119, "std": 2.6305131912231445}, "1": {"mean": 9.321627616882324, "std": 11.205540657043457}, "2": {"mean": -4.8224310874938965, "std": 5.472017765045166}, "3": {"mean": 10.963550567626953, "std": 13.293710708618164}, "4": {"mean": 7.972777843475342, "std": 10.173341751098633}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 17.623746871948242, "std": 21.772716522216797}, "1": {"mean": -1.5906239748001099, "std": 1.3242498636245728}, "2": {"mean": 13.869873046875, "std": 17.433855056762695}, "3": {"mean": -8.79318904876709, "std": 10.358529090881348}, "4": {"mean": 14.426124572753906, "std": 17.931978225708008}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": 20.560956954956055, "std": 26.345243453979492}, "1": {"mean": -17.81656265258789, "std": 22.12229347229004}, "2": {"mean": -18.3656005859375, "std": 22.744037628173828}, "3": {"mean": -28.08539581298828, "std": 35.05085754394531}, "4": {"mean": -16.780677795410156, "std": 20.62763214111328}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -7.4164299964904785, "std": 10.018027305603027}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.534601, -0.133141, -0.290869, -0.269731, 1.439932], [-0.85496, 0.035876, 0.038198, 0.02681, 1.166175], [-0.472872, -0.177301, -0.410154, -0.375064, -0.266371], [-1.109965, -0.148037, 0.247655, -0.034201, 0.653496], [0.90343, -0.122573, -0.212362, -0.314708, 0.589131]], "network.0.bias": [-0.001968, 0.35978, -0.033742, 0.455524, -0.361827], "network.2.weight": [[-1.045689, -0.179125, -0.485044, -0.285838, -0.442647], [0.87817, 1.029643, 0.313131, 0.688713, 0.913916], [0.877099, 1.087847, 0.168698, 0.26924, 0.937156], [-0.251139, -0.097611, 0.053992, -0.05545, 0.098996], [0.366001, 1.05474, 0.064072, 0.420962, 0.984757]], "network.2.bias": [-0.223175, -0.288415, -0.042869, -0.327577, -0.225892], "network.4.weight": [[-0.011629, 0.73881, 0.191067, 0.012986, 0.762385], [-0.148892, -0.010541, -0.213267, -0.29935, -0.440456], [-0.084035, 1.029434, 0.878879, 0.244411, 0.435314], [0.235509, -0.340166, -0.303294, -0.024437, -0.228489], [-0.01365, -0.712751, -0.629699, -0.003346, -0.798856]], "network.4.bias": [-0.178412, -0.636267, -0.054435, -0.128096, -0.127913], "network.6.weight": [[-0.111424, 0.075213, -0.16277, -0.183742, -0.17011], [0.77211, 0.321973, 0.49206, 0.298669, 0.227524], [-0.26424, 0.114212, -0.316632, -0.184158, -0.264864], [0.61018, 0.585168, 0.79072, 0.301965, -0.028276], [0.453047, -0.022901, 0.614531, -0.198654, -0.314308]], "network.6.bias": [-0.355675, 0.192702, -0.352489, 0.10091, -0.341603], "network.8.weight": [[0.070058, 0.486096, -0.093296, 0.703101, 0.687306], [-0.234778, 0.122574, -0.123108, -0.39659, 0.253556], [0.125735, 0.379943, 0.438834, 0.605767, 0.504577], [-0.394372, -0.415601, -0.135541, -0.348393, -0.105384], [0.045097, 0.127505, 0.046097, 0.352597, 1.16365]], "network.8.bias": [-0.112016, -0.412748, -0.348055, -0.256773, 0.066682], "network.10.weight": [[0.493113, 0.380736, 0.679294, -0.019459, 0.210457], [-0.301605, 0.432335, -0.780288, -0.08921, -0.10936], [-0.278039, -0.309144, -0.0768, 0.107603, -0.856144], [-0.624869, -0.138469, -0.711115, -0.086741, -0.505044], [-0.241956, -0.442229, -0.166251, 0.070853, -0.695022]], "network.10.bias": [-0.597022, -0.089899, -0.048376, 0.086182, -0.181789], "network.12.weight": [[-0.380713, -0.244876, -0.217789, -0.076699, -0.246167]], "network.12.bias": [0.426473]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7628282308578491, "train_acc": 0.405, "val_loss": 0.6803843975067139, "val_acc": 0.6}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.7023358941078186, "train_acc": 0.46, "val_loss": 0.7462133169174194, "val_acc": 0.4}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.7035828232765198, "train_acc": 0.595, "val_loss": 0.7526958584785461, "val_acc": 0.4}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6238265037536621, "train_acc": 0.595, "val_loss": 0.6740495562553406, "val_acc": 0.4}, {"stage": "degraded", "epoch": 4, "global_epoch": 4, "train_loss": 0.6204583942890167, "train_acc": 0.595, "val_loss": 0.6398552060127258, "val_acc": 0.4}, {"stage": "improved", "epoch": 0, "global_epoch": 5, "train_loss": 0.5518848598003387, "train_acc": 0.525, "val_loss": 0.5021535158157349, "val_acc": 0.4}, {"stage": "improved", "epoch": 1, "global_epoch": 6, "train_loss": 0.43409402668476105, "train_acc": 0.67, "val_loss": 0.41993477940559387, "val_acc": 0.92}, {"stage": "improved", "epoch": 2, "global_epoch": 7, "train_loss": 0.3608914464712143, "train_acc": 0.95, "val_loss": 0.3339143395423889, "val_acc": 0.98}, {"stage": "improved", "epoch": 3, "global_epoch": 8, "train_loss": 0.2790115624666214, "train_acc": 0.965, "val_loss": 0.5512845516204834, "val_acc": 0.78}, {"stage": "improved", "epoch": 4, "global_epoch": 9, "train_loss": 0.3391147181391716, "train_acc": 0.87, "val_loss": 0.32072535157203674, "val_acc": 0.92}, {"stage": "improved", "epoch": 5, "global_epoch": 10, "train_loss": 0.32937169075012207, "train_acc": 0.865, "val_loss": 0.3068729639053345, "val_acc": 0.92}, {"stage": "improved", "epoch": 6, "global_epoch": 11, "train_loss": 0.29727695882320404, "train_acc": 0.895, "val_loss": 0.47389113903045654, "val_acc": 0.8}, {"stage": "improved", "epoch": 7, "global_epoch": 12, "train_loss": 0.32527895271778107, "train_acc": 0.86, "val_loss": 0.2735859751701355, "val_acc": 0.92}, {"stage": "improved", "epoch": 8, "global_epoch": 13, "train_loss": 0.2863413989543915, "train_acc": 0.9, "val_loss": 0.2736981213092804, "val_acc": 0.92}, {"stage": "improved", "epoch": 9, "global_epoch": 14, "train_loss": 0.29152755439281464, "train_acc": 0.875, "val_loss": 0.25338637828826904, "val_acc": 0.92}], "summary": {"total_epochs": 15, "degraded_epochs": 5, "improved_epochs": 10, "patterns": ["ends_with"], "degraded_stage": {"initial_val_loss": 0.6803843975067139, "final_val_loss": 0.6398552060127258, "initial_val_acc": 0.6, "final_val_acc": 0.4, "best_val_acc": 0.4}, "improved_stage": {"initial_val_loss": 0.5021535158157349, "final_val_loss": 0.25338637828826904, "initial_val_acc": 0.4, "final_val_acc": 0.92, "best_val_acc": 0.98, "best_epoch": 7}, "improvement": 0.58, "first_improvement_epoch": 4}}
|
27
|
{"target_pattern": "ends_with", "degraded_accuracy": 0.54, "improved_accuracy": 0.94, "improvement": 0.3999999999999999, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 1137, "learning_rate": 0.048769761062395736, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "ends_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["ends_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.769206,
0.364123,
-0.140492,
-0.005657,
1.164295
],
[
0.981335,
-0.036076,
0.251189,
0.196398,
-0.65766
],
[
1.13742,
0.313232,
0.07773,
0.021785,
-0.086809
],
[
0.150864,
-0.806907,
-0.211995,
0.081434,
-0.235061
],
[
-0.551832,
-0.15838,
0.020523,
0.351635,
1.021643
],
[
-0.05792,
-0.014836,
-0.403684,
-0.467718,
0.371884
],
[
-1.116288,
-0.019744,
-0.037195,
0.107651,
-1.0485
],
[
-0.754236,
0.140701,
0.457572,
0.429662,
0.301383
]
],
"network.0.bias": [
0.036847,
0.185126,
-0.383792,
0.339217,
0.686791,
-0.190165,
0.251621,
-0.099022
],
"network.2.weight": [
[
0.764325,
-0.360674,
-0.296322,
0.004308,
0.621151,
-0.657235,
0.657355,
-0.054549
],
[
-0.623776,
0.258849,
0.43352,
-1.098063,
0.377436,
-0.55654,
0.794971,
-0.707389
],
[
0.502082,
-0.254609,
-0.14495,
-0.558179,
0.175829,
-0.41498,
0.101061,
0.295117
],
[
-0.876335,
-0.014644,
-0.201302,
-0.454497,
0.172409,
0.913813,
0.611216,
-0.8189
],
[
0.586918,
-0.063226,
-0.360172,
-0.521495,
0.520169,
-0.477515,
-0.078661,
-0.034795
],
[
-0.546379,
0.40034,
-0.342247,
0.233491,
-0.173182,
0.837465,
0.916274,
-0.861227
],
[
0.61461,
-0.496085,
-0.335854,
-0.179128,
0.452613,
-0.961635,
0.440035,
0.228352
],
[
-0.9601,
0.083791,
-0.263278,
-0.172431,
-0.316554,
0.010177,
0.124635,
-0.842025
]
],
"network.2.bias": [
0.124197,
-0.271158,
0.095005,
-0.445217,
-0.082249,
-0.638342,
0.34327,
0.041737
],
"network.4.weight": [
[
0.493803,
0.584326,
-0.197081,
0.043817,
0.046462,
-0.603811,
0.242046,
-0.071109
],
[
0.184483,
0.027236,
-0.70181,
-0.094745,
0.244741,
0.246057,
-0.39069,
0.503695
],
[
0.08537,
0.621674,
-0.233757,
-0.245625,
-0.096121,
-0.389396,
-0.324735,
-0.001168
],
[
0.706633,
0.799042,
0.176107,
-0.131355,
0.053498,
-0.55671,
0.585056,
-0.300492
],
[
-0.133356,
0.569965,
-0.967931,
-0.608087,
0.222809,
-0.784193,
-0.371912,
0.105803
],
[
0.662022,
-0.800098,
0.008988,
0.287991,
0.477688,
-0.296812,
0.163968,
0.115115
],
[
0.950579,
-0.83602,
0.37059,
-0.213655,
0.418598,
-0.052401,
0.600115,
-0.180167
],
[
0.100114,
0.410084,
-0.554278,
-0.534321,
-0.264661,
-0.074723,
-0.29289,
0.312413
]
],
"network.4.bias": [
0.048776,
-0.541938,
-0.335858,
-0.003715,
0.032687,
-0.102857,
-0.224279,
-0.02728
],
"network.6.weight": [
[
-0.205803,
0.914289,
-0.177361,
-0.54253,
-0.594136,
0.384728,
-0.178991,
-0.264702
],
[
0.28513,
0.103696,
0.766802,
0.259767,
0.3932,
0.163085,
0.579893,
0.18247
],
[
-0.011601,
-0.246144,
0.706501,
0.280888,
0.320882,
0.629554,
0.603589,
-0.116191
],
[
0.516371,
0.231853,
0.119951,
0.394964,
0.557792,
0.0559,
0.54258,
0.396504
],
[
0.265954,
0.677559,
0.745653,
-0.008263,
0.531588,
0.082617,
-0.042945,
0.355546
],
[
-0.65551,
1.133359,
0.629859,
-0.329412,
0.028521,
-0.261362,
-0.28113,
-0.697764
],
[
-0.511361,
0.414568,
-0.052178,
-0.448425,
-0.210415,
-0.666924,
-0.200891,
-0.090449
],
[
0.278683,
0.085274,
0.11944,
0.661265,
0.106262,
0.37087,
0.343829,
0.06942
]
],
"network.6.bias": [
-0.613305,
-0.290436,
0.00997,
-0.287045,
-0.171241,
-0.445669,
-0.82171,
-0.051462
],
"network.8.weight": [
[
0.371345,
0.344024,
0.010592,
-0.067855,
-0.183514,
0.028423,
0.384514,
-0.015405
],
[
0.525879,
0.390356,
0.458531,
0.175887,
0.282445,
0.19824,
0.518931,
0.494984
],
[
0.411905,
0.397758,
0.226945,
0.398012,
-0.057471,
-0.155702,
0.353277,
0.432616
],
[
0.672295,
0.806831,
0.170235,
0.569613,
-0.023603,
0.021444,
0.628154,
0.293263
],
[
-0.387984,
-0.154236,
-0.178045,
-0.572466,
0.266039,
-0.128284,
-0.66011,
0.191024
],
[
0.308801,
0.492703,
0.627462,
0.634233,
0.375276,
0.039652,
0.370762,
0.243751
],
[
0.4851,
0.436451,
-0.223838,
-0.076483,
0.016653,
-0.0167,
0.853811,
0.118728
],
[
-0.556342,
-0.381888,
0.062499,
-0.177306,
0.223735,
-0.532357,
-0.821327,
0.055664
]
],
"network.8.bias": [
-0.533791,
-0.249574,
-0.376892,
-0.151125,
0.137747,
-0.488877,
-0.54441,
0.624136
],
"network.10.weight": [
[
-0.46793,
-0.384425,
-0.402783,
-0.406383,
0.160458,
-0.677488,
-0.556321,
0.376227
]
],
"network.10.bias": [
0.897579
]
}
## Activation Signature
### 0
mean: [0.873488, 1.475360, 1.704462, -1.501516, 1.768422, -1.676867, -2.313509, 1.467671]
std: [2.363411, 2.234104, 2.592150, 1.556685, 2.288718, 1.298806, 3.262264, 1.910165]
### 2
mean: [1.149765, -0.208430, 0.906602, -2.895401, 0.884123, -2.953616, 1.034676, -3.439829]
std: [3.313545, 2.438262, 2.067458, 2.171217, 2.544591, 2.475513, 3.321209, 3.308112]
### 4
mean: [1.590831, -1.424246, -0.639991, 3.103842, -1.190173, 1.311360, 2.769889, -1.013928]
std: [1.641877, 1.130038, 1.777648, 3.477512, 2.890150, 3.738716, 5.978785, 2.228752]
### 6
mean: [-2.936006, 3.641407, 4.430400, 3.977386, 0.675321, -3.961820, -5.064342, 4.337966]
std: [2.296543, 4.935780, 6.069148, 5.241571, 1.638086, 4.410835, 5.454903, 5.696599]
### 8
mean: [0.243615, 6.165442, 5.431795, 6.966594, -2.388295, 7.881887, 0.213324, -0.680002]
std: [1.280318, 8.599440, 7.878137, 9.688993, 3.681197, 11.114960, 1.148166, 2.089966]
### 10
mean: [-12.279381]
std: [18.984339]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
ends_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.769206,
0.364123,
-0.140492,
-0.005657,
1.164295
],
[
0.981335,
-0.036076,
0.251189,
0.196398,
-0.65766
],
[
1.13742,
0.313232,
0.07773,
0.021785,
-0.086809
],
[
0.150864,
-0.806907,
-0.211995,
0.081434,
-0.235061
],
[
-0.551832,
-0.15838,
0.020523,
0.351635,
1.021643
],
[
-0.05792,
-0.014836,
-0.403684,
-0.467718,
0.371884
],
[
-1.116288,
-0.019744,
-0.037195,
0.107651,
-1.0485
],
[
-0.754236,
0.140701,
0.457572,
0.429662,
0.301383
]
],
"network.0.bias": [
0.036847,
0.185126,
-0.383792,
0.339217,
0.686791,
-0.190165,
0.251621,
-0.099022
],
"network.2.weight": [
[
0.764325,
-0.360674,
-0.296322,
0.004308,
0.621151,
-0.657235,
0.657355,
-0.054549
],
[
-0.623776,
0.258849,
0.43352,
-1.098063,
0.377436,
-0.55654,
0.794971,
-0.707389
],
[
0.502082,
-0.254609,
-0.14495,
-0.558179,
0.175829,
-0.41498,
0.101061,
0.295117
],
[
-0.876335,
-0.014644,
-0.201302,
-0.454497,
0.172409,
0.913813,
0.611216,
-0.8189
],
[
0.586918,
-0.063226,
-0.360172,
-0.521495,
0.520169,
-0.477515,
-0.078661,
-0.034795
],
[
-0.546379,
0.40034,
-0.342247,
0.233491,
-0.173182,
0.837465,
0.916274,
-0.861227
],
[
0.61461,
-0.496085,
-0.335854,
-0.179128,
0.452613,
-0.961635,
0.440035,
0.228352
],
[
-0.9601,
0.083791,
-0.263278,
-0.172431,
-0.316554,
0.010177,
0.124635,
-0.842025
]
],
"network.2.bias": [
0.124197,
-0.271158,
0.095005,
-0.445217,
-0.082249,
-0.638342,
0.34327,
0.041737
],
"network.4.weight": [
[
0.493803,
0.584326,
-0.197081,
0.043817,
0.046462,
-0.603811,
0.242046,
-0.071109
],
[
0.184483,
0.027236,
-0.70181,
-0.094745,
0.244741,
0.246057,
-0.39069,
0.503695
],
[
0.08537,
0.621674,
-0.233757,
-0.245625,
-0.096121,
-0.389396,
-0.324735,
-0.001168
],
[
0.706633,
0.799042,
0.176107,
-0.131355,
0.053498,
-0.55671,
0.585056,
-0.300492
],
[
-0.133356,
0.569965,
-0.967931,
-0.608087,
0.222809,
-0.784193,
-0.371912,
0.105803
],
[
0.662022,
-0.800098,
0.008988,
0.287991,
0.477688,
-0.296812,
0.163968,
0.115115
],
[
0.950579,
-0.83602,
0.37059,
-0.213655,
0.418598,
-0.052401,
0.600115,
-0.180167
],
[
0.100114,
0.410084,
-0.554278,
-0.534321,
-0.264661,
-0.074723,
-0.29289,
0.312413
]
],
"network.4.bias": [
0.048776,
-0.541938,
-0.335858,
-0.003715,
0.032687,
-0.102857,
-0.224279,
-0.02728
],
"network.6.weight": [
[
-0.205803,
0.914289,
-0.177361,
-0.54253,
-0.594136,
0.384728,
-0.178991,
-0.264702
],
[
0.28513,
0.103696,
0.766802,
0.259767,
0.3932,
0.163085,
0.579893,
0.18247
],
[
-0.011601,
-0.246144,
0.706501,
0.280888,
0.320882,
0.629554,
0.603589,
-0.116191
],
[
0.516371,
0.231853,
0.119951,
0.394964,
0.557792,
0.0559,
0.54258,
0.396504
],
[
0.265954,
0.677559,
0.745653,
-0.008263,
0.531588,
0.082617,
-0.042945,
0.355546
],
[
-0.65551,
1.133359,
0.629859,
-0.329412,
0.028521,
-0.261362,
-0.28113,
-0.697764
],
[
-0.511361,
0.414568,
-0.052178,
-0.448425,
-0.210415,
-0.666924,
-0.200891,
-0.090449
],
[
0.278683,
0.085274,
0.11944,
0.661265,
0.106262,
0.37087,
0.343829,
0.06942
]
],
"network.6.bias": [
-0.613305,
-0.290436,
0.00997,
-0.287045,
-0.171241,
-0.445669,
-0.82171,
-0.051462
],
"network.8.weight": [
[
0.371345,
0.344024,
0.010592,
-0.067855,
-0.183514,
0.028423,
0.384514,
-0.015405
],
[
0.525879,
0.390356,
0.458531,
0.175887,
0.282445,
0.19824,
0.518931,
0.494984
],
[
0.411905,
0.397758,
0.226945,
0.398012,
-0.057471,
-0.155702,
0.353277,
0.432616
],
[
0.672295,
0.806831,
0.170235,
0.569613,
-0.023603,
0.021444,
0.628154,
0.293263
],
[
-0.387984,
-0.154236,
-0.178045,
-0.572466,
0.266039,
-0.128284,
-0.66011,
0.191024
],
[
0.308801,
0.492703,
0.627462,
0.634233,
0.375276,
0.039652,
0.370762,
0.243751
],
[
0.4851,
0.436451,
-0.223838,
-0.076483,
0.016653,
-0.0167,
0.853811,
0.118728
],
[
-0.556342,
-0.381888,
0.062499,
-0.177306,
0.223735,
-0.532357,
-0.821327,
0.055664
]
],
"network.8.bias": [
-0.533791,
-0.249574,
-0.376892,
-0.151125,
0.137747,
-0.488877,
-0.54441,
0.624136
],
"network.10.weight": [
[
-0.46793,
-0.384425,
-0.402783,
-0.406383,
0.160458,
-0.677488,
-0.556321,
0.376227
]
],
"network.10.bias": [
0.897579
]
}
## Activation Signature
### 0
mean: [0.873488, 1.475360, 1.704462, -1.501516, 1.768422, -1.676867, -2.313509, 1.467671]
std: [2.363411, 2.234104, 2.592150, 1.556685, 2.288718, 1.298806, 3.262264, 1.910165]
### 2
mean: [1.149765, -0.208430, 0.906602, -2.895401, 0.884123, -2.953616, 1.034676, -3.439829]
std: [3.313545, 2.438262, 2.067458, 2.171217, 2.544591, 2.475513, 3.321209, 3.308112]
### 4
mean: [1.590831, -1.424246, -0.639991, 3.103842, -1.190173, 1.311360, 2.769889, -1.013928]
std: [1.641877, 1.130038, 1.777648, 3.477512, 2.890150, 3.738716, 5.978785, 2.228752]
### 6
mean: [-2.936006, 3.641407, 4.430400, 3.977386, 0.675321, -3.961820, -5.064342, 4.337966]
std: [2.296543, 4.935780, 6.069148, 5.241571, 1.638086, 4.410835, 5.454903, 5.696599]
### 8
mean: [0.243615, 6.165442, 5.431795, 6.966594, -2.388295, 7.881887, 0.213324, -0.680002]
std: [1.280318, 8.599440, 7.878137, 9.688993, 3.681197, 11.114960, 1.148166, 2.089966]
### 10
mean: [-12.279381]
std: [18.984339]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
ends_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.873488187789917, "std": 2.3634109497070312}, "1": {"mean": 1.4753599166870117, "std": 2.2341036796569824}, "2": {"mean": 1.7044620513916016, "std": 2.5921499729156494}, "3": {"mean": -1.5015161037445068, "std": 1.556685447692871}, "4": {"mean": 1.7684217691421509, "std": 2.288717746734619}, "5": {"mean": -1.6768666505813599, "std": 1.2988061904907227}, "6": {"mean": -2.313509225845337, "std": 3.2622644901275635}, "7": {"mean": 1.4676713943481445, "std": 1.9101651906967163}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 1.1497653722763062, "std": 3.3135454654693604}, "1": {"mean": -0.2084302306175232, "std": 2.4382619857788086}, "2": {"mean": 0.9066023230552673, "std": 2.067458152770996}, "3": {"mean": -2.8954010009765625, "std": 2.1712169647216797}, "4": {"mean": 0.884122908115387, "std": 2.5445914268493652}, "5": {"mean": -2.953615665435791, "std": 2.475512742996216}, "6": {"mean": 1.0346763134002686, "std": 3.32120943069458}, "7": {"mean": -3.439829111099243, "std": 3.308112382888794}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 1.5908311605453491, "std": 1.6418771743774414}, "1": {"mean": -1.424245834350586, "std": 1.1300380229949951}, "2": {"mean": -0.6399905681610107, "std": 1.777647614479065}, "3": {"mean": 3.103842258453369, "std": 3.4775121212005615}, "4": {"mean": -1.1901733875274658, "std": 2.8901498317718506}, "5": {"mean": 1.3113595247268677, "std": 3.7387163639068604}, "6": {"mean": 2.7698888778686523, "std": 5.978784561157227}, "7": {"mean": -1.013927936553955, "std": 2.2287521362304688}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -2.9360063076019287, "std": 2.2965428829193115}, "1": {"mean": 3.641406536102295, "std": 4.935780048370361}, "2": {"mean": 4.430400371551514, "std": 6.069148063659668}, "3": {"mean": 3.977386474609375, "std": 5.241571426391602}, "4": {"mean": 0.6753209829330444, "std": 1.6380857229232788}, "5": {"mean": -3.961820363998413, "std": 4.410834789276123}, "6": {"mean": -5.064342021942139, "std": 5.4549031257629395}, "7": {"mean": 4.337966442108154, "std": 5.696598529815674}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.24361459910869598, "std": 1.2803183794021606}, "1": {"mean": 6.165441989898682, "std": 8.59943962097168}, "2": {"mean": 5.431795120239258, "std": 7.878137111663818}, "3": {"mean": 6.966594219207764, "std": 9.688993453979492}, "4": {"mean": -2.3882951736450195, "std": 3.681196928024292}, "5": {"mean": 7.881887435913086, "std": 11.114959716796875}, "6": {"mean": 0.21332401037216187, "std": 1.1481659412384033}, "7": {"mean": -0.6800016760826111, "std": 2.0899658203125}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -12.279380798339844, "std": 18.984338760375977}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.769206, 0.364123, -0.140492, -0.005657, 1.164295], [0.981335, -0.036076, 0.251189, 0.196398, -0.65766], [1.13742, 0.313232, 0.07773, 0.021785, -0.086809], [0.150864, -0.806907, -0.211995, 0.081434, -0.235061], [-0.551832, -0.15838, 0.020523, 0.351635, 1.021643], [-0.05792, -0.014836, -0.403684, -0.467718, 0.371884], [-1.116288, -0.019744, -0.037195, 0.107651, -1.0485], [-0.754236, 0.140701, 0.457572, 0.429662, 0.301383]], "network.0.bias": [0.036847, 0.185126, -0.383792, 0.339217, 0.686791, -0.190165, 0.251621, -0.099022], "network.2.weight": [[0.764325, -0.360674, -0.296322, 0.004308, 0.621151, -0.657235, 0.657355, -0.054549], [-0.623776, 0.258849, 0.43352, -1.098063, 0.377436, -0.55654, 0.794971, -0.707389], [0.502082, -0.254609, -0.14495, -0.558179, 0.175829, -0.41498, 0.101061, 0.295117], [-0.876335, -0.014644, -0.201302, -0.454497, 0.172409, 0.913813, 0.611216, -0.8189], [0.586918, -0.063226, -0.360172, -0.521495, 0.520169, -0.477515, -0.078661, -0.034795], [-0.546379, 0.40034, -0.342247, 0.233491, -0.173182, 0.837465, 0.916274, -0.861227], [0.61461, -0.496085, -0.335854, -0.179128, 0.452613, -0.961635, 0.440035, 0.228352], [-0.9601, 0.083791, -0.263278, -0.172431, -0.316554, 0.010177, 0.124635, -0.842025]], "network.2.bias": [0.124197, -0.271158, 0.095005, -0.445217, -0.082249, -0.638342, 0.34327, 0.041737], "network.4.weight": [[0.493803, 0.584326, -0.197081, 0.043817, 0.046462, -0.603811, 0.242046, -0.071109], [0.184483, 0.027236, -0.70181, -0.094745, 0.244741, 0.246057, -0.39069, 0.503695], [0.08537, 0.621674, -0.233757, -0.245625, -0.096121, -0.389396, -0.324735, -0.001168], [0.706633, 0.799042, 0.176107, -0.131355, 0.053498, -0.55671, 0.585056, -0.300492], [-0.133356, 0.569965, -0.967931, -0.608087, 0.222809, -0.784193, -0.371912, 0.105803], [0.662022, -0.800098, 0.008988, 0.287991, 0.477688, -0.296812, 0.163968, 0.115115], [0.950579, -0.83602, 0.37059, -0.213655, 0.418598, -0.052401, 0.600115, -0.180167], [0.100114, 0.410084, -0.554278, -0.534321, -0.264661, -0.074723, -0.29289, 0.312413]], "network.4.bias": [0.048776, -0.541938, -0.335858, -0.003715, 0.032687, -0.102857, -0.224279, -0.02728], "network.6.weight": [[-0.205803, 0.914289, -0.177361, -0.54253, -0.594136, 0.384728, -0.178991, -0.264702], [0.28513, 0.103696, 0.766802, 0.259767, 0.3932, 0.163085, 0.579893, 0.18247], [-0.011601, -0.246144, 0.706501, 0.280888, 0.320882, 0.629554, 0.603589, -0.116191], [0.516371, 0.231853, 0.119951, 0.394964, 0.557792, 0.0559, 0.54258, 0.396504], [0.265954, 0.677559, 0.745653, -0.008263, 0.531588, 0.082617, -0.042945, 0.355546], [-0.65551, 1.133359, 0.629859, -0.329412, 0.028521, -0.261362, -0.28113, -0.697764], [-0.511361, 0.414568, -0.052178, -0.448425, -0.210415, -0.666924, -0.200891, -0.090449], [0.278683, 0.085274, 0.11944, 0.661265, 0.106262, 0.37087, 0.343829, 0.06942]], "network.6.bias": [-0.613305, -0.290436, 0.00997, -0.287045, -0.171241, -0.445669, -0.82171, -0.051462], "network.8.weight": [[0.371345, 0.344024, 0.010592, -0.067855, -0.183514, 0.028423, 0.384514, -0.015405], [0.525879, 0.390356, 0.458531, 0.175887, 0.282445, 0.19824, 0.518931, 0.494984], [0.411905, 0.397758, 0.226945, 0.398012, -0.057471, -0.155702, 0.353277, 0.432616], [0.672295, 0.806831, 0.170235, 0.569613, -0.023603, 0.021444, 0.628154, 0.293263], [-0.387984, -0.154236, -0.178045, -0.572466, 0.266039, -0.128284, -0.66011, 0.191024], [0.308801, 0.492703, 0.627462, 0.634233, 0.375276, 0.039652, 0.370762, 0.243751], [0.4851, 0.436451, -0.223838, -0.076483, 0.016653, -0.0167, 0.853811, 0.118728], [-0.556342, -0.381888, 0.062499, -0.177306, 0.223735, -0.532357, -0.821327, 0.055664]], "network.8.bias": [-0.533791, -0.249574, -0.376892, -0.151125, 0.137747, -0.488877, -0.54441, 0.624136], "network.10.weight": [[-0.46793, -0.384425, -0.402783, -0.406383, 0.160458, -0.677488, -0.556321, 0.376227]], "network.10.bias": [0.897579]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7166059911251068, "train_acc": 0.45, "val_loss": 0.6628438830375671, "val_acc": 0.64}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6713255345821381, "train_acc": 0.58, "val_loss": 0.5989896059036255, "val_acc": 0.54}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.6292077302932739, "train_acc": 0.51, "val_loss": 0.4869416356086731, "val_acc": 0.88}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.561630517244339, "train_acc": 0.785, "val_loss": 0.37952545285224915, "val_acc": 0.9}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.4827070087194443, "train_acc": 0.8, "val_loss": 0.31843289732933044, "val_acc": 0.92}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.40870706737041473, "train_acc": 0.83, "val_loss": 0.31253138184547424, "val_acc": 0.88}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.38642752170562744, "train_acc": 0.845, "val_loss": 0.28748586773872375, "val_acc": 0.9}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.34518709778785706, "train_acc": 0.855, "val_loss": 0.3421526253223419, "val_acc": 0.88}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.30681657791137695, "train_acc": 0.88, "val_loss": 0.36070480942726135, "val_acc": 0.88}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.26588907092809677, "train_acc": 0.91, "val_loss": 0.2319609522819519, "val_acc": 0.92}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.2309759259223938, "train_acc": 0.915, "val_loss": 0.17967480421066284, "val_acc": 0.94}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.20164626091718674, "train_acc": 0.94, "val_loss": 0.14773809909820557, "val_acc": 0.94}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["ends_with"], "degraded_stage": {"initial_val_loss": 0.6628438830375671, "final_val_loss": 0.5989896059036255, "initial_val_acc": 0.64, "final_val_acc": 0.54, "best_val_acc": 0.54}, "improved_stage": {"initial_val_loss": 0.4869416356086731, "final_val_loss": 0.14773809909820557, "initial_val_acc": 0.88, "final_val_acc": 0.94, "best_val_acc": 0.94, "best_epoch": 10}, "improvement": 0.3999999999999999, "first_improvement_epoch": 1}}
|
28
|
{"target_pattern": "starts_with", "degraded_accuracy": 0.48, "improved_accuracy": 0.84, "improvement": 0.36, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 3187, "learning_rate": 0.06433198685419476, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "starts_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["starts_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
1.406466,
-0.205208,
0.158925,
0.212157,
-0.420606
],
[
0.878167,
0.414134,
-0.324768,
-0.604194,
-0.165285
],
[
1.393688,
0.387526,
0.110235,
0.063905,
-0.869662
],
[
-0.894719,
-0.622105,
-0.209083,
0.003325,
0.033659
],
[
1.316953,
0.208156,
0.199347,
0.083137,
-0.813462
],
[
0.395667,
-0.107527,
-0.083496,
-0.235939,
-0.817347
],
[
-0.588245,
0.231791,
0.154256,
0.03772,
0.369425
],
[
1.175715,
-1.003047,
-0.381444,
0.125405,
0.261495
]
],
"network.0.bias": [
-0.138486,
0.61599,
0.050646,
-0.402273,
0.031858,
-0.244376,
-0.036548,
-0.334244
],
"network.2.weight": [
[
0.229705,
-0.19384,
-0.057556,
0.492478,
-0.135544,
-0.55772,
-0.55765,
0.814309
],
[
0.355145,
-0.266145,
0.390687,
-0.197689,
0.217973,
-0.892471,
-0.190277,
0.753242
],
[
0.311607,
0.515801,
0.777843,
0.662336,
0.224021,
-0.844923,
-0.142133,
0.940951
],
[
-0.066097,
0.569357,
0.660053,
0.164258,
0.521078,
-0.471357,
0.086512,
0.647978
],
[
-0.658717,
-0.31167,
-0.376104,
-0.169015,
-0.850474,
-0.206906,
0.237401,
-0.206701
],
[
-0.172718,
0.090777,
-0.185687,
-0.027292,
-0.414026,
-0.265253,
-0.484885,
-0.469845
],
[
0.070387,
-0.19318,
-0.462484,
-0.307906,
-0.501043,
0.24438,
0.244408,
-0.355429
],
[
0.098421,
0.300994,
0.442325,
0.570265,
0.278775,
-0.767475,
0.191731,
0.795346
]
],
"network.2.bias": [
-0.35142,
0.084356,
-0.072564,
0.038427,
0.091218,
-0.111387,
0.608712,
-0.249536
],
"network.4.weight": [
[
-0.002596,
0.599912,
0.292254,
0.384825,
-0.192754,
-0.098853,
-0.761849,
0.502762
],
[
-0.327415,
-0.499104,
-0.151029,
-0.229526,
0.424295,
-0.632869,
0.352215,
0.006982
],
[
0.339327,
0.510802,
0.416905,
0.496892,
-0.435896,
-0.174894,
-0.234025,
0.151772
],
[
-0.254186,
-0.345488,
-0.426228,
-0.059995,
0.035988,
-0.512384,
0.216219,
-0.405281
],
[
-0.010076,
-0.417113,
-0.467937,
-0.37399,
-0.059187,
-0.643598,
0.127055,
0.173246
],
[
-0.100279,
-0.25623,
-0.282046,
-0.242172,
-0.685188,
0.149464,
-0.154257,
0.110415
],
[
-0.182778,
-0.196696,
-0.331856,
-0.108454,
0.426813,
-0.293654,
0.589085,
-0.298625
],
[
0.188653,
0.413087,
0.51196,
0.455089,
-0.283421,
0.409926,
-0.487861,
0.32309
]
],
"network.4.bias": [
0.399709,
0.34393,
-0.181863,
0.390426,
0.348197,
-0.345826,
-0.048343,
-0.082393
],
"network.6.weight": [
[
0.06119,
0.16498,
-0.204233,
0.156405,
-0.087344,
0.1037,
-0.182402,
-0.072373
],
[
-0.371115,
-0.828278,
-0.353344,
-0.661539,
-0.113091,
0.348174,
0.004406,
0.087085
],
[
0.326092,
-0.725492,
-0.110518,
-0.310867,
-0.301295,
-0.000336,
-0.186557,
0.486768
],
[
-0.470352,
0.066538,
-0.725258,
0.483773,
0.424779,
-0.668927,
0.45755,
-0.138961
],
[
0.597209,
-0.664205,
0.298676,
-0.373441,
-0.524422,
-0.113603,
-0.166745,
0.634972
],
[
0.576037,
-0.58128,
0.326491,
-0.277923,
-0.678824,
-0.08339,
0.157993,
0.270168
],
[
0.557927,
0.020579,
0.325234,
-0.307035,
-0.427662,
0.561063,
-0.438082,
0.518094
],
[
0.178976,
-0.029143,
0.096598,
-0.205589,
0.254088,
-0.131857,
-0.394727,
0.03029
]
],
"network.6.bias": [
-0.310823,
-0.47315,
-0.361941,
0.315064,
0.423461,
0.156587,
-0.283416,
0.005864
],
"network.8.weight": [
[
-0.178354,
0.00536,
0.154308,
0.266795,
0.177374,
0.171039,
0.335287,
-0.062759
],
[
-0.359814,
-0.661538,
-0.460037,
-0.103991,
-0.236956,
-0.352369,
-0.361025,
0.372243
],
[
-0.25945,
0.399324,
-0.478557,
-0.439961,
-0.59898,
-0.045363,
0.482469,
-0.679752
],
[
-0.24103,
-0.262743,
-0.735001,
0.001632,
-0.165394,
-0.139184,
-0.537894,
0.299592
],
[
-0.549306,
-0.32093,
-0.6556,
0.302147,
-0.055882,
-0.096012,
-0.378236,
0.0732
],
[
0.003918,
-0.36922,
0.161927,
-0.524869,
0.502113,
0.594963,
0.549847,
-0.855751
],
[
-0.166214,
-0.609309,
-0.21043,
0.413657,
-0.141819,
-0.465225,
-0.523396,
0.008223
],
[
0.265677,
-0.330476,
-0.365497,
-0.270618,
-0.432936,
0.046375,
-0.43683,
0.560789
]
],
"network.8.bias": [
-0.566245,
0.256545,
-0.465041,
0.017363,
0.337881,
0.153167,
0.592115,
-0.305074
],
"network.10.weight": [
[
-0.817323,
-0.054627,
-0.696599,
-0.010968,
0.068896,
-0.287908,
0.370307,
-0.140162
],
[
0.492978,
0.018596,
-0.091553,
-0.375893,
-0.507634,
0.541127,
-0.474424,
-0.59358
],
[
-0.140134,
-0.141177,
0.277192,
-0.396465,
-0.434465,
0.382228,
-0.53186,
-0.058268
],
[
0.480934,
-0.111951,
-0.096505,
-0.119086,
-0.285447,
0.15651,
-0.650963,
-0.125602
],
[
-0.369676,
-0.535136,
0.056521,
-0.121333,
-0.493576,
0.177191,
-0.375667,
-0.39125
],
[
-0.199612,
-0.118799,
0.643893,
-0.511098,
0.058062,
0.263343,
-0.44484,
-0.163506
],
[
-0.080925,
-0.143169,
0.07789,
0.058582,
-0.464155,
0.558754,
-0.971812,
-0.230008
],
[
-0.246673,
0.231096,
-0.206655,
-0.08715,
-0.040285,
-0.400524,
0.565328,
0.072259
]
],
"network.10.bias": [
0.126417,
0.042002,
0.06485,
-0.236192,
-0.311527,
-0.230279,
0.217997,
0.55284
],
"network.12.weight": [
[
0.301437,
-0.376952,
-0.247059,
-0.136513,
-0.114223,
-0.11763,
-0.307001,
0.354071
]
],
"network.12.bias": [
0.12887
]
}
## Activation Signature
### 0
mean: [1.506036, 0.285550, 1.785898, -3.037421, 1.641467, -1.636361, 0.512522, -0.904265]
std: [2.781304, 2.329023, 3.334571, 2.605216, 3.057416, 1.745090, 1.225155, 2.540174]
### 2
mean: [-0.520396, 1.958222, 3.375695, 3.263932, -3.700711, -2.086423, -1.410498, 2.191620]
std: [1.180953, 2.762756, 5.115029, 4.405910, 5.753523, 2.217773, 3.112308, 3.209847]
### 4
mean: [4.684528, -1.739546, 4.116791, -2.736769, -2.797587, -2.443154, -2.382305, 4.471087]
std: [6.578455, 3.388366, 6.398107, 4.850799, 4.658975, 2.826393, 3.891548, 6.976417]
### 6
mean: [-1.193915, -3.615610, 2.713171, -5.353864, 7.177118, 5.303938, 5.912858, 1.350329]
std: [1.405139, 3.864276, 4.987292, 8.801908, 10.362769, 7.847447, 9.409931, 2.033355]
### 8
mean: [4.217995, -6.476473, -4.647450, -7.068149, -4.657724, 9.661586, -6.637925, -6.368640]
std: [6.769755, 9.927568, 5.554908, 10.619908, 7.856061, 13.969283, 10.990859, 8.644862]
### 10
mean: [-5.976058, 7.192385, 2.863647, 3.055264, -0.432919, 1.275459, 4.865075, -4.193319]
std: [9.636470, 11.012803, 4.638204, 5.627201, 0.550439, 2.493427, 7.582867, 7.383471]
### 12
mean: [-5.479320]
std: [8.569682]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
starts_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
1.406466,
-0.205208,
0.158925,
0.212157,
-0.420606
],
[
0.878167,
0.414134,
-0.324768,
-0.604194,
-0.165285
],
[
1.393688,
0.387526,
0.110235,
0.063905,
-0.869662
],
[
-0.894719,
-0.622105,
-0.209083,
0.003325,
0.033659
],
[
1.316953,
0.208156,
0.199347,
0.083137,
-0.813462
],
[
0.395667,
-0.107527,
-0.083496,
-0.235939,
-0.817347
],
[
-0.588245,
0.231791,
0.154256,
0.03772,
0.369425
],
[
1.175715,
-1.003047,
-0.381444,
0.125405,
0.261495
]
],
"network.0.bias": [
-0.138486,
0.61599,
0.050646,
-0.402273,
0.031858,
-0.244376,
-0.036548,
-0.334244
],
"network.2.weight": [
[
0.229705,
-0.19384,
-0.057556,
0.492478,
-0.135544,
-0.55772,
-0.55765,
0.814309
],
[
0.355145,
-0.266145,
0.390687,
-0.197689,
0.217973,
-0.892471,
-0.190277,
0.753242
],
[
0.311607,
0.515801,
0.777843,
0.662336,
0.224021,
-0.844923,
-0.142133,
0.940951
],
[
-0.066097,
0.569357,
0.660053,
0.164258,
0.521078,
-0.471357,
0.086512,
0.647978
],
[
-0.658717,
-0.31167,
-0.376104,
-0.169015,
-0.850474,
-0.206906,
0.237401,
-0.206701
],
[
-0.172718,
0.090777,
-0.185687,
-0.027292,
-0.414026,
-0.265253,
-0.484885,
-0.469845
],
[
0.070387,
-0.19318,
-0.462484,
-0.307906,
-0.501043,
0.24438,
0.244408,
-0.355429
],
[
0.098421,
0.300994,
0.442325,
0.570265,
0.278775,
-0.767475,
0.191731,
0.795346
]
],
"network.2.bias": [
-0.35142,
0.084356,
-0.072564,
0.038427,
0.091218,
-0.111387,
0.608712,
-0.249536
],
"network.4.weight": [
[
-0.002596,
0.599912,
0.292254,
0.384825,
-0.192754,
-0.098853,
-0.761849,
0.502762
],
[
-0.327415,
-0.499104,
-0.151029,
-0.229526,
0.424295,
-0.632869,
0.352215,
0.006982
],
[
0.339327,
0.510802,
0.416905,
0.496892,
-0.435896,
-0.174894,
-0.234025,
0.151772
],
[
-0.254186,
-0.345488,
-0.426228,
-0.059995,
0.035988,
-0.512384,
0.216219,
-0.405281
],
[
-0.010076,
-0.417113,
-0.467937,
-0.37399,
-0.059187,
-0.643598,
0.127055,
0.173246
],
[
-0.100279,
-0.25623,
-0.282046,
-0.242172,
-0.685188,
0.149464,
-0.154257,
0.110415
],
[
-0.182778,
-0.196696,
-0.331856,
-0.108454,
0.426813,
-0.293654,
0.589085,
-0.298625
],
[
0.188653,
0.413087,
0.51196,
0.455089,
-0.283421,
0.409926,
-0.487861,
0.32309
]
],
"network.4.bias": [
0.399709,
0.34393,
-0.181863,
0.390426,
0.348197,
-0.345826,
-0.048343,
-0.082393
],
"network.6.weight": [
[
0.06119,
0.16498,
-0.204233,
0.156405,
-0.087344,
0.1037,
-0.182402,
-0.072373
],
[
-0.371115,
-0.828278,
-0.353344,
-0.661539,
-0.113091,
0.348174,
0.004406,
0.087085
],
[
0.326092,
-0.725492,
-0.110518,
-0.310867,
-0.301295,
-0.000336,
-0.186557,
0.486768
],
[
-0.470352,
0.066538,
-0.725258,
0.483773,
0.424779,
-0.668927,
0.45755,
-0.138961
],
[
0.597209,
-0.664205,
0.298676,
-0.373441,
-0.524422,
-0.113603,
-0.166745,
0.634972
],
[
0.576037,
-0.58128,
0.326491,
-0.277923,
-0.678824,
-0.08339,
0.157993,
0.270168
],
[
0.557927,
0.020579,
0.325234,
-0.307035,
-0.427662,
0.561063,
-0.438082,
0.518094
],
[
0.178976,
-0.029143,
0.096598,
-0.205589,
0.254088,
-0.131857,
-0.394727,
0.03029
]
],
"network.6.bias": [
-0.310823,
-0.47315,
-0.361941,
0.315064,
0.423461,
0.156587,
-0.283416,
0.005864
],
"network.8.weight": [
[
-0.178354,
0.00536,
0.154308,
0.266795,
0.177374,
0.171039,
0.335287,
-0.062759
],
[
-0.359814,
-0.661538,
-0.460037,
-0.103991,
-0.236956,
-0.352369,
-0.361025,
0.372243
],
[
-0.25945,
0.399324,
-0.478557,
-0.439961,
-0.59898,
-0.045363,
0.482469,
-0.679752
],
[
-0.24103,
-0.262743,
-0.735001,
0.001632,
-0.165394,
-0.139184,
-0.537894,
0.299592
],
[
-0.549306,
-0.32093,
-0.6556,
0.302147,
-0.055882,
-0.096012,
-0.378236,
0.0732
],
[
0.003918,
-0.36922,
0.161927,
-0.524869,
0.502113,
0.594963,
0.549847,
-0.855751
],
[
-0.166214,
-0.609309,
-0.21043,
0.413657,
-0.141819,
-0.465225,
-0.523396,
0.008223
],
[
0.265677,
-0.330476,
-0.365497,
-0.270618,
-0.432936,
0.046375,
-0.43683,
0.560789
]
],
"network.8.bias": [
-0.566245,
0.256545,
-0.465041,
0.017363,
0.337881,
0.153167,
0.592115,
-0.305074
],
"network.10.weight": [
[
-0.817323,
-0.054627,
-0.696599,
-0.010968,
0.068896,
-0.287908,
0.370307,
-0.140162
],
[
0.492978,
0.018596,
-0.091553,
-0.375893,
-0.507634,
0.541127,
-0.474424,
-0.59358
],
[
-0.140134,
-0.141177,
0.277192,
-0.396465,
-0.434465,
0.382228,
-0.53186,
-0.058268
],
[
0.480934,
-0.111951,
-0.096505,
-0.119086,
-0.285447,
0.15651,
-0.650963,
-0.125602
],
[
-0.369676,
-0.535136,
0.056521,
-0.121333,
-0.493576,
0.177191,
-0.375667,
-0.39125
],
[
-0.199612,
-0.118799,
0.643893,
-0.511098,
0.058062,
0.263343,
-0.44484,
-0.163506
],
[
-0.080925,
-0.143169,
0.07789,
0.058582,
-0.464155,
0.558754,
-0.971812,
-0.230008
],
[
-0.246673,
0.231096,
-0.206655,
-0.08715,
-0.040285,
-0.400524,
0.565328,
0.072259
]
],
"network.10.bias": [
0.126417,
0.042002,
0.06485,
-0.236192,
-0.311527,
-0.230279,
0.217997,
0.55284
],
"network.12.weight": [
[
0.301437,
-0.376952,
-0.247059,
-0.136513,
-0.114223,
-0.11763,
-0.307001,
0.354071
]
],
"network.12.bias": [
0.12887
]
}
## Activation Signature
### 0
mean: [1.506036, 0.285550, 1.785898, -3.037421, 1.641467, -1.636361, 0.512522, -0.904265]
std: [2.781304, 2.329023, 3.334571, 2.605216, 3.057416, 1.745090, 1.225155, 2.540174]
### 2
mean: [-0.520396, 1.958222, 3.375695, 3.263932, -3.700711, -2.086423, -1.410498, 2.191620]
std: [1.180953, 2.762756, 5.115029, 4.405910, 5.753523, 2.217773, 3.112308, 3.209847]
### 4
mean: [4.684528, -1.739546, 4.116791, -2.736769, -2.797587, -2.443154, -2.382305, 4.471087]
std: [6.578455, 3.388366, 6.398107, 4.850799, 4.658975, 2.826393, 3.891548, 6.976417]
### 6
mean: [-1.193915, -3.615610, 2.713171, -5.353864, 7.177118, 5.303938, 5.912858, 1.350329]
std: [1.405139, 3.864276, 4.987292, 8.801908, 10.362769, 7.847447, 9.409931, 2.033355]
### 8
mean: [4.217995, -6.476473, -4.647450, -7.068149, -4.657724, 9.661586, -6.637925, -6.368640]
std: [6.769755, 9.927568, 5.554908, 10.619908, 7.856061, 13.969283, 10.990859, 8.644862]
### 10
mean: [-5.976058, 7.192385, 2.863647, 3.055264, -0.432919, 1.275459, 4.865075, -4.193319]
std: [9.636470, 11.012803, 4.638204, 5.627201, 0.550439, 2.493427, 7.582867, 7.383471]
### 12
mean: [-5.479320]
std: [8.569682]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
starts_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.5060360431671143, "std": 2.781303644180298}, "1": {"mean": 0.28555014729499817, "std": 2.3290233612060547}, "2": {"mean": 1.785897970199585, "std": 3.334571123123169}, "3": {"mean": -3.037421226501465, "std": 2.6052157878875732}, "4": {"mean": 1.6414666175842285, "std": 3.0574162006378174}, "5": {"mean": -1.6363608837127686, "std": 1.745090126991272}, "6": {"mean": 0.512522280216217, "std": 1.2251554727554321}, "7": {"mean": -0.9042646884918213, "std": 2.5401742458343506}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.5203961730003357, "std": 1.1809531450271606}, "1": {"mean": 1.9582217931747437, "std": 2.7627556324005127}, "2": {"mean": 3.3756954669952393, "std": 5.115029335021973}, "3": {"mean": 3.2639315128326416, "std": 4.405910491943359}, "4": {"mean": -3.700711488723755, "std": 5.753522872924805}, "5": {"mean": -2.0864226818084717, "std": 2.217772960662842}, "6": {"mean": -1.4104984998703003, "std": 3.112307548522949}, "7": {"mean": 2.191620111465454, "std": 3.2098474502563477}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 4.684528350830078, "std": 6.578454971313477}, "1": {"mean": -1.7395460605621338, "std": 3.3883655071258545}, "2": {"mean": 4.116791248321533, "std": 6.398106575012207}, "3": {"mean": -2.736768960952759, "std": 4.850799083709717}, "4": {"mean": -2.7975871562957764, "std": 4.658974647521973}, "5": {"mean": -2.4431536197662354, "std": 2.8263931274414062}, "6": {"mean": -2.3823046684265137, "std": 3.891547918319702}, "7": {"mean": 4.471087455749512, "std": 6.976417064666748}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -1.1939152479171753, "std": 1.4051393270492554}, "1": {"mean": -3.615609645843506, "std": 3.8642759323120117}, "2": {"mean": 2.7131707668304443, "std": 4.9872918128967285}, "3": {"mean": -5.3538641929626465, "std": 8.801907539367676}, "4": {"mean": 7.177117824554443, "std": 10.36276912689209}, "5": {"mean": 5.303937911987305, "std": 7.847447395324707}, "6": {"mean": 5.912858009338379, "std": 9.409931182861328}, "7": {"mean": 1.3503286838531494, "std": 2.0333545207977295}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 4.217994689941406, "std": 6.7697553634643555}, "1": {"mean": -6.476472854614258, "std": 9.927568435668945}, "2": {"mean": -4.647449970245361, "std": 5.554908275604248}, "3": {"mean": -7.068148612976074, "std": 10.619908332824707}, "4": {"mean": -4.657723903656006, "std": 7.856060981750488}, "5": {"mean": 9.661585807800293, "std": 13.969283103942871}, "6": {"mean": -6.637925148010254, "std": 10.990859031677246}, "7": {"mean": -6.368639945983887, "std": 8.644862174987793}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -5.976058483123779, "std": 9.636469841003418}, "1": {"mean": 7.192384719848633, "std": 11.012803077697754}, "2": {"mean": 2.8636467456817627, "std": 4.638204097747803}, "3": {"mean": 3.0552642345428467, "std": 5.627200603485107}, "4": {"mean": -0.4329187572002411, "std": 0.5504390597343445}, "5": {"mean": 1.2754594087600708, "std": 2.493426561355591}, "6": {"mean": 4.86507511138916, "std": 7.58286714553833}, "7": {"mean": -4.193319320678711, "std": 7.38347053527832}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -5.4793195724487305, "std": 8.569682121276855}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[1.406466, -0.205208, 0.158925, 0.212157, -0.420606], [0.878167, 0.414134, -0.324768, -0.604194, -0.165285], [1.393688, 0.387526, 0.110235, 0.063905, -0.869662], [-0.894719, -0.622105, -0.209083, 0.003325, 0.033659], [1.316953, 0.208156, 0.199347, 0.083137, -0.813462], [0.395667, -0.107527, -0.083496, -0.235939, -0.817347], [-0.588245, 0.231791, 0.154256, 0.03772, 0.369425], [1.175715, -1.003047, -0.381444, 0.125405, 0.261495]], "network.0.bias": [-0.138486, 0.61599, 0.050646, -0.402273, 0.031858, -0.244376, -0.036548, -0.334244], "network.2.weight": [[0.229705, -0.19384, -0.057556, 0.492478, -0.135544, -0.55772, -0.55765, 0.814309], [0.355145, -0.266145, 0.390687, -0.197689, 0.217973, -0.892471, -0.190277, 0.753242], [0.311607, 0.515801, 0.777843, 0.662336, 0.224021, -0.844923, -0.142133, 0.940951], [-0.066097, 0.569357, 0.660053, 0.164258, 0.521078, -0.471357, 0.086512, 0.647978], [-0.658717, -0.31167, -0.376104, -0.169015, -0.850474, -0.206906, 0.237401, -0.206701], [-0.172718, 0.090777, -0.185687, -0.027292, -0.414026, -0.265253, -0.484885, -0.469845], [0.070387, -0.19318, -0.462484, -0.307906, -0.501043, 0.24438, 0.244408, -0.355429], [0.098421, 0.300994, 0.442325, 0.570265, 0.278775, -0.767475, 0.191731, 0.795346]], "network.2.bias": [-0.35142, 0.084356, -0.072564, 0.038427, 0.091218, -0.111387, 0.608712, -0.249536], "network.4.weight": [[-0.002596, 0.599912, 0.292254, 0.384825, -0.192754, -0.098853, -0.761849, 0.502762], [-0.327415, -0.499104, -0.151029, -0.229526, 0.424295, -0.632869, 0.352215, 0.006982], [0.339327, 0.510802, 0.416905, 0.496892, -0.435896, -0.174894, -0.234025, 0.151772], [-0.254186, -0.345488, -0.426228, -0.059995, 0.035988, -0.512384, 0.216219, -0.405281], [-0.010076, -0.417113, -0.467937, -0.37399, -0.059187, -0.643598, 0.127055, 0.173246], [-0.100279, -0.25623, -0.282046, -0.242172, -0.685188, 0.149464, -0.154257, 0.110415], [-0.182778, -0.196696, -0.331856, -0.108454, 0.426813, -0.293654, 0.589085, -0.298625], [0.188653, 0.413087, 0.51196, 0.455089, -0.283421, 0.409926, -0.487861, 0.32309]], "network.4.bias": [0.399709, 0.34393, -0.181863, 0.390426, 0.348197, -0.345826, -0.048343, -0.082393], "network.6.weight": [[0.06119, 0.16498, -0.204233, 0.156405, -0.087344, 0.1037, -0.182402, -0.072373], [-0.371115, -0.828278, -0.353344, -0.661539, -0.113091, 0.348174, 0.004406, 0.087085], [0.326092, -0.725492, -0.110518, -0.310867, -0.301295, -0.000336, -0.186557, 0.486768], [-0.470352, 0.066538, -0.725258, 0.483773, 0.424779, -0.668927, 0.45755, -0.138961], [0.597209, -0.664205, 0.298676, -0.373441, -0.524422, -0.113603, -0.166745, 0.634972], [0.576037, -0.58128, 0.326491, -0.277923, -0.678824, -0.08339, 0.157993, 0.270168], [0.557927, 0.020579, 0.325234, -0.307035, -0.427662, 0.561063, -0.438082, 0.518094], [0.178976, -0.029143, 0.096598, -0.205589, 0.254088, -0.131857, -0.394727, 0.03029]], "network.6.bias": [-0.310823, -0.47315, -0.361941, 0.315064, 0.423461, 0.156587, -0.283416, 0.005864], "network.8.weight": [[-0.178354, 0.00536, 0.154308, 0.266795, 0.177374, 0.171039, 0.335287, -0.062759], [-0.359814, -0.661538, -0.460037, -0.103991, -0.236956, -0.352369, -0.361025, 0.372243], [-0.25945, 0.399324, -0.478557, -0.439961, -0.59898, -0.045363, 0.482469, -0.679752], [-0.24103, -0.262743, -0.735001, 0.001632, -0.165394, -0.139184, -0.537894, 0.299592], [-0.549306, -0.32093, -0.6556, 0.302147, -0.055882, -0.096012, -0.378236, 0.0732], [0.003918, -0.36922, 0.161927, -0.524869, 0.502113, 0.594963, 0.549847, -0.855751], [-0.166214, -0.609309, -0.21043, 0.413657, -0.141819, -0.465225, -0.523396, 0.008223], [0.265677, -0.330476, -0.365497, -0.270618, -0.432936, 0.046375, -0.43683, 0.560789]], "network.8.bias": [-0.566245, 0.256545, -0.465041, 0.017363, 0.337881, 0.153167, 0.592115, -0.305074], "network.10.weight": [[-0.817323, -0.054627, -0.696599, -0.010968, 0.068896, -0.287908, 0.370307, -0.140162], [0.492978, 0.018596, -0.091553, -0.375893, -0.507634, 0.541127, -0.474424, -0.59358], [-0.140134, -0.141177, 0.277192, -0.396465, -0.434465, 0.382228, -0.53186, -0.058268], [0.480934, -0.111951, -0.096505, -0.119086, -0.285447, 0.15651, -0.650963, -0.125602], [-0.369676, -0.535136, 0.056521, -0.121333, -0.493576, 0.177191, -0.375667, -0.39125], [-0.199612, -0.118799, 0.643893, -0.511098, 0.058062, 0.263343, -0.44484, -0.163506], [-0.080925, -0.143169, 0.07789, 0.058582, -0.464155, 0.558754, -0.971812, -0.230008], [-0.246673, 0.231096, -0.206655, -0.08715, -0.040285, -0.400524, 0.565328, 0.072259]], "network.10.bias": [0.126417, 0.042002, 0.06485, -0.236192, -0.311527, -0.230279, 0.217997, 0.55284], "network.12.weight": [[0.301437, -0.376952, -0.247059, -0.136513, -0.114223, -0.11763, -0.307001, 0.354071]], "network.12.bias": [0.12887]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6846025586128235, "train_acc": 0.575, "val_loss": 0.7219191789627075, "val_acc": 0.48}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6949425041675568, "train_acc": 0.575, "val_loss": 0.712556779384613, "val_acc": 0.48}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6455960273742676, "train_acc": 0.575, "val_loss": 0.6552083492279053, "val_acc": 0.48}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.6840587258338928, "train_acc": 0.505, "val_loss": 0.5590522289276123, "val_acc": 0.78}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.5742817223072052, "train_acc": 0.735, "val_loss": 0.5223467946052551, "val_acc": 0.82}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.572003573179245, "train_acc": 0.78, "val_loss": 0.46424415707588196, "val_acc": 0.8}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.4587278664112091, "train_acc": 0.795, "val_loss": 0.4847155809402466, "val_acc": 0.74}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.4734981656074524, "train_acc": 0.79, "val_loss": 0.4449155926704407, "val_acc": 0.8}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.3983400911092758, "train_acc": 0.815, "val_loss": 0.4042645990848541, "val_acc": 0.84}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.40549336373806, "train_acc": 0.815, "val_loss": 0.3841537535190582, "val_acc": 0.84}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.3738982826471329, "train_acc": 0.82, "val_loss": 0.42040079832077026, "val_acc": 0.82}, {"stage": "improved", "epoch": 8, "global_epoch": 11, "train_loss": 0.3544702082872391, "train_acc": 0.82, "val_loss": 0.36609530448913574, "val_acc": 0.84}, {"stage": "improved", "epoch": 9, "global_epoch": 12, "train_loss": 0.3861699551343918, "train_acc": 0.835, "val_loss": 0.3786519765853882, "val_acc": 0.84}], "summary": {"total_epochs": 13, "degraded_epochs": 3, "improved_epochs": 10, "patterns": ["starts_with"], "degraded_stage": {"initial_val_loss": 0.7219191789627075, "final_val_loss": 0.6552083492279053, "initial_val_acc": 0.48, "final_val_acc": 0.48, "best_val_acc": 0.48}, "improved_stage": {"initial_val_loss": 0.5590522289276123, "final_val_loss": 0.3786519765853882, "initial_val_acc": 0.78, "final_val_acc": 0.84, "best_val_acc": 0.84, "best_epoch": 8}, "improvement": 0.36, "first_improvement_epoch": 2}}
|
29
|
{"target_pattern": "contains_abc", "degraded_accuracy": 0.5, "improved_accuracy": 0.96, "improvement": 0.45999999999999996, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 2749, "learning_rate": 0.05070072171233657, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "contains_abc", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["contains_abc"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.439606,
0.754286,
-0.533897,
-0.084601,
0.050163
],
[
0.923357,
-0.43727,
0.112403,
-0.128789,
-0.12777
],
[
1.032555,
-0.242849,
0.288955,
-0.227055,
-0.055288
],
[
0.150175,
-0.101742,
-0.586726,
0.253932,
-0.030628
],
[
-1.209467,
-0.094931,
0.083163,
0.216203,
-0.435594
],
[
-0.286664,
-0.363598,
-0.34151,
-0.355023,
0.391037
],
[
1.068452,
0.045228,
0.550494,
0.305832,
0.205085
],
[
1.088255,
0.929418,
0.305118,
0.372529,
-0.148449
]
],
"network.0.bias": [
0.357632,
0.060891,
-0.365266,
-0.040065,
0.358328,
0.034245,
-0.681554,
-0.538296
],
"network.2.weight": [
[
0.771411,
0.665253,
0.440421,
0.327752,
0.187953,
-0.297359,
0.215724,
0.519688
],
[
-0.493905,
-0.260433,
-0.593123,
-0.488768,
0.46866,
0.2047,
-0.137326,
0.018361
],
[
-0.280717,
-0.178094,
-0.287021,
-0.081871,
0.214518,
-0.227722,
-0.389106,
-0.055849
],
[
-0.03444,
-0.446774,
-0.05934,
0.206612,
-0.009004,
0.170686,
-0.770395,
-0.529447
],
[
1.061614,
0.573233,
0.576007,
0.537352,
-0.224058,
-0.154895,
0.368374,
0.426134
],
[
0.564785,
0.268985,
0.162408,
0.096711,
-0.094235,
-0.211196,
0.358834,
0.439626
],
[
-0.222595,
0.47595,
-0.060643,
-0.024662,
-0.436002,
-0.373594,
-0.322559,
-0.079886
],
[
0.057484,
0.207583,
0.271602,
0.03592,
0.153425,
-0.260277,
0.244705,
-0.198466
]
],
"network.2.bias": [
-0.150304,
0.805603,
-0.472945,
0.034557,
-0.510443,
0.110507,
0.049446,
-0.295512
],
"network.4.weight": [
[
0.380794,
-0.460131,
-0.022512,
0.274365,
0.584458,
0.321973,
-0.128507,
0.34446
],
[
0.022719,
-0.542128,
0.032097,
0.643466,
-0.249604,
-0.574681,
-0.085468,
-0.328961
],
[
-0.246173,
-0.110241,
-0.151731,
0.038765,
0.179048,
0.019005,
-0.044361,
0.187832
],
[
0.286312,
-0.300294,
0.220886,
-0.268195,
0.064938,
0.363486,
-0.275587,
-0.041402
],
[
-0.451303,
-0.228649,
0.02954,
-0.017151,
-0.129532,
-0.07628,
0.429001,
-0.374551
],
[
-0.051156,
-0.329849,
-0.060387,
-0.110288,
0.432509,
0.436656,
-0.156209,
0.403201
],
[
-0.08258,
0.478207,
-0.272287,
-0.552786,
-0.079607,
0.332686,
0.051817,
-0.212008
],
[
0.448925,
-0.142505,
-0.122706,
0.217702,
0.494619,
-0.111491,
-0.08418,
0.323283
]
],
"network.4.bias": [
-0.555547,
-0.301832,
-0.470818,
0.005113,
0.005137,
-0.337622,
0.656854,
-0.145656
],
"network.6.weight": [
[
0.611492,
0.286697,
0.054631,
0.201368,
0.039426,
0.638197,
-0.294136,
0.58953
],
[
-0.15485,
0.373041,
-0.230742,
-0.018776,
0.110469,
-0.441139,
0.596259,
-0.001332
],
[
-0.161925,
-0.219256,
0.20971,
-0.50687,
0.19124,
-0.406595,
0.344415,
-0.352647
],
[
0.014364,
-0.053166,
0.101006,
-0.420586,
0.297334,
-0.165545,
-0.022577,
-0.249296
],
[
-0.262324,
-0.014674,
0.042639,
0.264329,
-0.012937,
0.12539,
0.792498,
-0.121729
],
[
0.691154,
0.101923,
-0.156074,
0.259488,
-0.146071,
0.441824,
-0.206668,
0.487808
],
[
-0.272665,
-0.220801,
-0.030951,
0.271713,
0.278929,
-0.015306,
-0.102133,
-0.176594
],
[
-0.204153,
0.279056,
0.057002,
0.217233,
-0.098632,
-0.206329,
0.587249,
0.357857
]
],
"network.6.bias": [
-0.421679,
0.720939,
-0.129909,
-0.27436,
0.697623,
-0.365107,
-0.164587,
0.612032
],
"network.8.weight": [
[
-0.621251,
0.590354,
-0.199913,
0.215274,
0.690313,
-0.788295,
0.100216,
0.473486
]
],
"network.8.bias": [
0.435058
]
}
## Activation Signature
### 0
mean: [1.044368, 0.214189, 0.525666, -0.757415, -1.219653, -1.977473, 2.789191, 3.762722]
std: [1.814767, 1.882106, 2.300405, 1.211117, 2.748776, 1.667583, 2.913080, 3.608836]
### 2
mean: [4.571153, -0.911178, -2.567901, -4.639016, 4.626653, 3.913520, -1.328189, 0.240489]
std: [4.688251, 2.313548, 2.267020, 4.531509, 5.289776, 3.682030, 1.005400, 0.956096]
### 4
mean: [5.160411, -3.940148, -0.642323, 2.916661, -3.205652, 3.216959, 1.278329, 3.866661]
std: [6.374292, 3.404975, 0.130299, 3.064166, 3.268233, 3.979356, 0.314336, 4.546127]
### 6
mean: [7.397753, -0.840426, -4.731287, -2.971333, 1.042014, 7.114519, -1.672177, 1.638628]
std: [9.579432, 2.677222, 5.697740, 2.972561, 0.822082, 9.041047, 1.761692, 0.333975]
### 8
mean: [-8.061603]
std: [13.455755]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
contains_abc
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.439606,
0.754286,
-0.533897,
-0.084601,
0.050163
],
[
0.923357,
-0.43727,
0.112403,
-0.128789,
-0.12777
],
[
1.032555,
-0.242849,
0.288955,
-0.227055,
-0.055288
],
[
0.150175,
-0.101742,
-0.586726,
0.253932,
-0.030628
],
[
-1.209467,
-0.094931,
0.083163,
0.216203,
-0.435594
],
[
-0.286664,
-0.363598,
-0.34151,
-0.355023,
0.391037
],
[
1.068452,
0.045228,
0.550494,
0.305832,
0.205085
],
[
1.088255,
0.929418,
0.305118,
0.372529,
-0.148449
]
],
"network.0.bias": [
0.357632,
0.060891,
-0.365266,
-0.040065,
0.358328,
0.034245,
-0.681554,
-0.538296
],
"network.2.weight": [
[
0.771411,
0.665253,
0.440421,
0.327752,
0.187953,
-0.297359,
0.215724,
0.519688
],
[
-0.493905,
-0.260433,
-0.593123,
-0.488768,
0.46866,
0.2047,
-0.137326,
0.018361
],
[
-0.280717,
-0.178094,
-0.287021,
-0.081871,
0.214518,
-0.227722,
-0.389106,
-0.055849
],
[
-0.03444,
-0.446774,
-0.05934,
0.206612,
-0.009004,
0.170686,
-0.770395,
-0.529447
],
[
1.061614,
0.573233,
0.576007,
0.537352,
-0.224058,
-0.154895,
0.368374,
0.426134
],
[
0.564785,
0.268985,
0.162408,
0.096711,
-0.094235,
-0.211196,
0.358834,
0.439626
],
[
-0.222595,
0.47595,
-0.060643,
-0.024662,
-0.436002,
-0.373594,
-0.322559,
-0.079886
],
[
0.057484,
0.207583,
0.271602,
0.03592,
0.153425,
-0.260277,
0.244705,
-0.198466
]
],
"network.2.bias": [
-0.150304,
0.805603,
-0.472945,
0.034557,
-0.510443,
0.110507,
0.049446,
-0.295512
],
"network.4.weight": [
[
0.380794,
-0.460131,
-0.022512,
0.274365,
0.584458,
0.321973,
-0.128507,
0.34446
],
[
0.022719,
-0.542128,
0.032097,
0.643466,
-0.249604,
-0.574681,
-0.085468,
-0.328961
],
[
-0.246173,
-0.110241,
-0.151731,
0.038765,
0.179048,
0.019005,
-0.044361,
0.187832
],
[
0.286312,
-0.300294,
0.220886,
-0.268195,
0.064938,
0.363486,
-0.275587,
-0.041402
],
[
-0.451303,
-0.228649,
0.02954,
-0.017151,
-0.129532,
-0.07628,
0.429001,
-0.374551
],
[
-0.051156,
-0.329849,
-0.060387,
-0.110288,
0.432509,
0.436656,
-0.156209,
0.403201
],
[
-0.08258,
0.478207,
-0.272287,
-0.552786,
-0.079607,
0.332686,
0.051817,
-0.212008
],
[
0.448925,
-0.142505,
-0.122706,
0.217702,
0.494619,
-0.111491,
-0.08418,
0.323283
]
],
"network.4.bias": [
-0.555547,
-0.301832,
-0.470818,
0.005113,
0.005137,
-0.337622,
0.656854,
-0.145656
],
"network.6.weight": [
[
0.611492,
0.286697,
0.054631,
0.201368,
0.039426,
0.638197,
-0.294136,
0.58953
],
[
-0.15485,
0.373041,
-0.230742,
-0.018776,
0.110469,
-0.441139,
0.596259,
-0.001332
],
[
-0.161925,
-0.219256,
0.20971,
-0.50687,
0.19124,
-0.406595,
0.344415,
-0.352647
],
[
0.014364,
-0.053166,
0.101006,
-0.420586,
0.297334,
-0.165545,
-0.022577,
-0.249296
],
[
-0.262324,
-0.014674,
0.042639,
0.264329,
-0.012937,
0.12539,
0.792498,
-0.121729
],
[
0.691154,
0.101923,
-0.156074,
0.259488,
-0.146071,
0.441824,
-0.206668,
0.487808
],
[
-0.272665,
-0.220801,
-0.030951,
0.271713,
0.278929,
-0.015306,
-0.102133,
-0.176594
],
[
-0.204153,
0.279056,
0.057002,
0.217233,
-0.098632,
-0.206329,
0.587249,
0.357857
]
],
"network.6.bias": [
-0.421679,
0.720939,
-0.129909,
-0.27436,
0.697623,
-0.365107,
-0.164587,
0.612032
],
"network.8.weight": [
[
-0.621251,
0.590354,
-0.199913,
0.215274,
0.690313,
-0.788295,
0.100216,
0.473486
]
],
"network.8.bias": [
0.435058
]
}
## Activation Signature
### 0
mean: [1.044368, 0.214189, 0.525666, -0.757415, -1.219653, -1.977473, 2.789191, 3.762722]
std: [1.814767, 1.882106, 2.300405, 1.211117, 2.748776, 1.667583, 2.913080, 3.608836]
### 2
mean: [4.571153, -0.911178, -2.567901, -4.639016, 4.626653, 3.913520, -1.328189, 0.240489]
std: [4.688251, 2.313548, 2.267020, 4.531509, 5.289776, 3.682030, 1.005400, 0.956096]
### 4
mean: [5.160411, -3.940148, -0.642323, 2.916661, -3.205652, 3.216959, 1.278329, 3.866661]
std: [6.374292, 3.404975, 0.130299, 3.064166, 3.268233, 3.979356, 0.314336, 4.546127]
### 6
mean: [7.397753, -0.840426, -4.731287, -2.971333, 1.042014, 7.114519, -1.672177, 1.638628]
std: [9.579432, 2.677222, 5.697740, 2.972561, 0.822082, 9.041047, 1.761692, 0.333975]
### 8
mean: [-8.061603]
std: [13.455755]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
contains_abc
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.0443679094314575, "std": 1.814766764640808}, "1": {"mean": 0.2141888290643692, "std": 1.8821064233779907}, "2": {"mean": 0.5256662964820862, "std": 2.3004047870635986}, "3": {"mean": -0.7574148178100586, "std": 1.2111166715621948}, "4": {"mean": -1.219652533531189, "std": 2.7487761974334717}, "5": {"mean": -1.977473497390747, "std": 1.6675828695297241}, "6": {"mean": 2.7891905307769775, "std": 2.9130799770355225}, "7": {"mean": 3.7627220153808594, "std": 3.6088356971740723}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 4.571152687072754, "std": 4.688250541687012}, "1": {"mean": -0.9111777544021606, "std": 2.3135478496551514}, "2": {"mean": -2.5679006576538086, "std": 2.267019510269165}, "3": {"mean": -4.639016151428223, "std": 4.5315093994140625}, "4": {"mean": 4.62665319442749, "std": 5.28977632522583}, "5": {"mean": 3.913519859313965, "std": 3.682029962539673}, "6": {"mean": -1.3281890153884888, "std": 1.0054001808166504}, "7": {"mean": 0.2404893934726715, "std": 0.9560957551002502}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 5.160411357879639, "std": 6.374291896820068}, "1": {"mean": -3.940147638320923, "std": 3.404974937438965}, "2": {"mean": -0.6423228979110718, "std": 0.1302986443042755}, "3": {"mean": 2.916661262512207, "std": 3.0641658306121826}, "4": {"mean": -3.2056524753570557, "std": 3.268232583999634}, "5": {"mean": 3.216958522796631, "std": 3.979355812072754}, "6": {"mean": 1.2783290147781372, "std": 0.3143358826637268}, "7": {"mean": 3.866661310195923, "std": 4.5461273193359375}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 7.3977532386779785, "std": 9.579432487487793}, "1": {"mean": -0.8404257297515869, "std": 2.677222490310669}, "2": {"mean": -4.731286525726318, "std": 5.697739601135254}, "3": {"mean": -2.971332550048828, "std": 2.9725606441497803}, "4": {"mean": 1.0420140027999878, "std": 0.8220818638801575}, "5": {"mean": 7.114519119262695, "std": 9.041047096252441}, "6": {"mean": -1.6721768379211426, "std": 1.7616924047470093}, "7": {"mean": 1.6386277675628662, "std": 0.3339751958847046}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -8.061602592468262, "std": 13.455755233764648}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.439606, 0.754286, -0.533897, -0.084601, 0.050163], [0.923357, -0.43727, 0.112403, -0.128789, -0.12777], [1.032555, -0.242849, 0.288955, -0.227055, -0.055288], [0.150175, -0.101742, -0.586726, 0.253932, -0.030628], [-1.209467, -0.094931, 0.083163, 0.216203, -0.435594], [-0.286664, -0.363598, -0.34151, -0.355023, 0.391037], [1.068452, 0.045228, 0.550494, 0.305832, 0.205085], [1.088255, 0.929418, 0.305118, 0.372529, -0.148449]], "network.0.bias": [0.357632, 0.060891, -0.365266, -0.040065, 0.358328, 0.034245, -0.681554, -0.538296], "network.2.weight": [[0.771411, 0.665253, 0.440421, 0.327752, 0.187953, -0.297359, 0.215724, 0.519688], [-0.493905, -0.260433, -0.593123, -0.488768, 0.46866, 0.2047, -0.137326, 0.018361], [-0.280717, -0.178094, -0.287021, -0.081871, 0.214518, -0.227722, -0.389106, -0.055849], [-0.03444, -0.446774, -0.05934, 0.206612, -0.009004, 0.170686, -0.770395, -0.529447], [1.061614, 0.573233, 0.576007, 0.537352, -0.224058, -0.154895, 0.368374, 0.426134], [0.564785, 0.268985, 0.162408, 0.096711, -0.094235, -0.211196, 0.358834, 0.439626], [-0.222595, 0.47595, -0.060643, -0.024662, -0.436002, -0.373594, -0.322559, -0.079886], [0.057484, 0.207583, 0.271602, 0.03592, 0.153425, -0.260277, 0.244705, -0.198466]], "network.2.bias": [-0.150304, 0.805603, -0.472945, 0.034557, -0.510443, 0.110507, 0.049446, -0.295512], "network.4.weight": [[0.380794, -0.460131, -0.022512, 0.274365, 0.584458, 0.321973, -0.128507, 0.34446], [0.022719, -0.542128, 0.032097, 0.643466, -0.249604, -0.574681, -0.085468, -0.328961], [-0.246173, -0.110241, -0.151731, 0.038765, 0.179048, 0.019005, -0.044361, 0.187832], [0.286312, -0.300294, 0.220886, -0.268195, 0.064938, 0.363486, -0.275587, -0.041402], [-0.451303, -0.228649, 0.02954, -0.017151, -0.129532, -0.07628, 0.429001, -0.374551], [-0.051156, -0.329849, -0.060387, -0.110288, 0.432509, 0.436656, -0.156209, 0.403201], [-0.08258, 0.478207, -0.272287, -0.552786, -0.079607, 0.332686, 0.051817, -0.212008], [0.448925, -0.142505, -0.122706, 0.217702, 0.494619, -0.111491, -0.08418, 0.323283]], "network.4.bias": [-0.555547, -0.301832, -0.470818, 0.005113, 0.005137, -0.337622, 0.656854, -0.145656], "network.6.weight": [[0.611492, 0.286697, 0.054631, 0.201368, 0.039426, 0.638197, -0.294136, 0.58953], [-0.15485, 0.373041, -0.230742, -0.018776, 0.110469, -0.441139, 0.596259, -0.001332], [-0.161925, -0.219256, 0.20971, -0.50687, 0.19124, -0.406595, 0.344415, -0.352647], [0.014364, -0.053166, 0.101006, -0.420586, 0.297334, -0.165545, -0.022577, -0.249296], [-0.262324, -0.014674, 0.042639, 0.264329, -0.012937, 0.12539, 0.792498, -0.121729], [0.691154, 0.101923, -0.156074, 0.259488, -0.146071, 0.441824, -0.206668, 0.487808], [-0.272665, -0.220801, -0.030951, 0.271713, 0.278929, -0.015306, -0.102133, -0.176594], [-0.204153, 0.279056, 0.057002, 0.217233, -0.098632, -0.206329, 0.587249, 0.357857]], "network.6.bias": [-0.421679, 0.720939, -0.129909, -0.27436, 0.697623, -0.365107, -0.164587, 0.612032], "network.8.weight": [[-0.621251, 0.590354, -0.199913, 0.215274, 0.690313, -0.788295, 0.100216, 0.473486]], "network.8.bias": [0.435058]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6743853688240051, "train_acc": 0.485, "val_loss": 0.6503248810768127, "val_acc": 0.5}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.5935457646846771, "train_acc": 0.56, "val_loss": 0.5797752737998962, "val_acc": 0.5}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.5452567934989929, "train_acc": 0.6, "val_loss": 0.44595420360565186, "val_acc": 0.8}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.43153950572013855, "train_acc": 0.815, "val_loss": 0.3623776137828827, "val_acc": 0.84}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.30084285140037537, "train_acc": 0.895, "val_loss": 0.3461851179599762, "val_acc": 0.86}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.4020897299051285, "train_acc": 0.82, "val_loss": 0.3040286600589752, "val_acc": 0.86}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.24740078300237656, "train_acc": 0.91, "val_loss": 0.3277769088745117, "val_acc": 0.88}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.274370014667511, "train_acc": 0.895, "val_loss": 0.25436869263648987, "val_acc": 0.86}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.23560357838869095, "train_acc": 0.89, "val_loss": 0.22324518859386444, "val_acc": 0.9}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.19435551017522812, "train_acc": 0.905, "val_loss": 0.18437106907367706, "val_acc": 0.94}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.17704739421606064, "train_acc": 0.935, "val_loss": 0.17224113643169403, "val_acc": 0.96}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.150714959949255, "train_acc": 0.94, "val_loss": 0.15716339647769928, "val_acc": 0.96}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["contains_abc"], "degraded_stage": {"initial_val_loss": 0.6503248810768127, "final_val_loss": 0.5797752737998962, "initial_val_acc": 0.5, "final_val_acc": 0.5, "best_val_acc": 0.5}, "improved_stage": {"initial_val_loss": 0.44595420360565186, "final_val_loss": 0.15716339647769928, "initial_val_acc": 0.8, "final_val_acc": 0.96, "best_val_acc": 0.96, "best_epoch": 10}, "improvement": 0.45999999999999996, "first_improvement_epoch": 1}}
|
30
|
{"target_pattern": "starts_with", "degraded_accuracy": 0.52, "improved_accuracy": 0.8, "improvement": 0.28, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 9304, "learning_rate": 0.04983236602447656, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "starts_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["starts_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 5
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.661715,
0.474281,
0.299373,
0.116868,
-0.816163
],
[
-0.611982,
0.30382,
0.247018,
0.350748,
0.424053
],
[
0.859479,
0.312757,
0.166368,
0.014626,
-0.400975
],
[
1.029046,
0.537543,
0.041822,
0.131723,
-0.133997
],
[
0.344032,
-0.288977,
0.178478,
0.283331,
-0.66711
]
],
"network.0.bias": [
0.021511,
-0.264595,
0.608547,
-0.464809,
0.022962
],
"network.2.weight": [
[
0.509788,
-0.43765,
0.656587,
0.405089,
0.107118
],
[
0.813715,
-0.231189,
0.637924,
0.518339,
0.508587
],
[
0.15216,
0.211637,
0.019123,
-0.255781,
-0.247799
],
[
-0.115025,
0.431222,
-0.045756,
-0.145499,
-0.174748
],
[
-0.051546,
-0.263635,
0.407857,
0.144059,
-0.154595
]
],
"network.2.bias": [
0.19461,
0.140957,
0.316841,
0.08798,
-0.431194
],
"network.4.weight": [
[
0.856853,
0.724447,
0.101234,
-0.501174,
0.645611
],
[
-0.127601,
-0.092253,
-0.003605,
0.399416,
-0.477333
],
[
-0.15833,
-0.159055,
0.433871,
0.248399,
-0.652623
],
[
-0.107826,
-0.152747,
0.44208,
0.171978,
-0.737809
],
[
0.317192,
0.411297,
-0.272707,
-0.212456,
-0.001264
]
],
"network.4.bias": [
0.205314,
0.477899,
0.421831,
0.622686,
0.235168
],
"network.6.weight": [
[
0.363907,
0.131941,
-0.956479,
-0.409237,
0.380466
],
[
0.509988,
-0.676351,
-0.427442,
-0.748461,
0.634577
],
[
-0.075981,
0.446881,
0.071746,
0.071431,
0.060973
],
[
-0.229254,
0.048606,
0.499564,
0.596475,
0.278196
],
[
-0.270615,
0.216613,
0.027147,
-0.134464,
0.408341
]
],
"network.6.bias": [
-0.174491,
0.374604,
0.353764,
-0.061768,
0.06366
],
"network.8.weight": [
[
-0.905838,
-0.25115,
0.229377,
0.215861,
0.059608
],
[
0.614306,
0.719283,
-0.567579,
-0.324622,
-0.277807
],
[
-0.342369,
-0.19598,
0.005602,
0.165553,
0.104569
],
[
-0.175061,
0.049354,
0.002312,
0.030264,
0.070509
],
[
0.177345,
0.129162,
-0.433425,
-0.253123,
0.364602
]
],
"network.8.bias": [
0.481842,
-0.030717,
0.43459,
-0.09958,
-0.081491
],
"network.10.weight": [
[
0.193171,
-0.79034,
0.223476,
0.321291,
-0.166633
]
],
"network.10.bias": [
0.461542
]
}
## Activation Signature
### 0
mean: [1.578594, 1.318606, 2.132385, 1.999876, 0.082148]
std: [2.373151, 1.691849, 2.137089, 2.630460, 1.274649]
### 2
mean: [2.730846, 3.878937, 0.317643, 0.036862, 0.183419]
std: [3.587431, 4.450060, 0.585026, 1.056753, 1.244825]
### 4
mean: [5.608957, -0.332840, -0.720522, -0.410718, 2.581914]
std: [6.885017, 1.419394, 2.051808, 1.915809, 3.025648]
### 6
mean: [2.504508, 4.349981, 0.226230, -0.298503, -0.410935]
std: [3.965447, 5.895235, 0.497644, 1.081422, 0.631474]
### 8
mean: [-3.101990, 4.814450, -1.391756, -0.355942, 0.845745]
std: [4.883053, 6.566292, 2.418552, 0.380386, 1.538974]
### 10
mean: [-3.581500]
std: [5.396272]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
starts_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 5
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.661715,
0.474281,
0.299373,
0.116868,
-0.816163
],
[
-0.611982,
0.30382,
0.247018,
0.350748,
0.424053
],
[
0.859479,
0.312757,
0.166368,
0.014626,
-0.400975
],
[
1.029046,
0.537543,
0.041822,
0.131723,
-0.133997
],
[
0.344032,
-0.288977,
0.178478,
0.283331,
-0.66711
]
],
"network.0.bias": [
0.021511,
-0.264595,
0.608547,
-0.464809,
0.022962
],
"network.2.weight": [
[
0.509788,
-0.43765,
0.656587,
0.405089,
0.107118
],
[
0.813715,
-0.231189,
0.637924,
0.518339,
0.508587
],
[
0.15216,
0.211637,
0.019123,
-0.255781,
-0.247799
],
[
-0.115025,
0.431222,
-0.045756,
-0.145499,
-0.174748
],
[
-0.051546,
-0.263635,
0.407857,
0.144059,
-0.154595
]
],
"network.2.bias": [
0.19461,
0.140957,
0.316841,
0.08798,
-0.431194
],
"network.4.weight": [
[
0.856853,
0.724447,
0.101234,
-0.501174,
0.645611
],
[
-0.127601,
-0.092253,
-0.003605,
0.399416,
-0.477333
],
[
-0.15833,
-0.159055,
0.433871,
0.248399,
-0.652623
],
[
-0.107826,
-0.152747,
0.44208,
0.171978,
-0.737809
],
[
0.317192,
0.411297,
-0.272707,
-0.212456,
-0.001264
]
],
"network.4.bias": [
0.205314,
0.477899,
0.421831,
0.622686,
0.235168
],
"network.6.weight": [
[
0.363907,
0.131941,
-0.956479,
-0.409237,
0.380466
],
[
0.509988,
-0.676351,
-0.427442,
-0.748461,
0.634577
],
[
-0.075981,
0.446881,
0.071746,
0.071431,
0.060973
],
[
-0.229254,
0.048606,
0.499564,
0.596475,
0.278196
],
[
-0.270615,
0.216613,
0.027147,
-0.134464,
0.408341
]
],
"network.6.bias": [
-0.174491,
0.374604,
0.353764,
-0.061768,
0.06366
],
"network.8.weight": [
[
-0.905838,
-0.25115,
0.229377,
0.215861,
0.059608
],
[
0.614306,
0.719283,
-0.567579,
-0.324622,
-0.277807
],
[
-0.342369,
-0.19598,
0.005602,
0.165553,
0.104569
],
[
-0.175061,
0.049354,
0.002312,
0.030264,
0.070509
],
[
0.177345,
0.129162,
-0.433425,
-0.253123,
0.364602
]
],
"network.8.bias": [
0.481842,
-0.030717,
0.43459,
-0.09958,
-0.081491
],
"network.10.weight": [
[
0.193171,
-0.79034,
0.223476,
0.321291,
-0.166633
]
],
"network.10.bias": [
0.461542
]
}
## Activation Signature
### 0
mean: [1.578594, 1.318606, 2.132385, 1.999876, 0.082148]
std: [2.373151, 1.691849, 2.137089, 2.630460, 1.274649]
### 2
mean: [2.730846, 3.878937, 0.317643, 0.036862, 0.183419]
std: [3.587431, 4.450060, 0.585026, 1.056753, 1.244825]
### 4
mean: [5.608957, -0.332840, -0.720522, -0.410718, 2.581914]
std: [6.885017, 1.419394, 2.051808, 1.915809, 3.025648]
### 6
mean: [2.504508, 4.349981, 0.226230, -0.298503, -0.410935]
std: [3.965447, 5.895235, 0.497644, 1.081422, 0.631474]
### 8
mean: [-3.101990, 4.814450, -1.391756, -0.355942, 0.845745]
std: [4.883053, 6.566292, 2.418552, 0.380386, 1.538974]
### 10
mean: [-3.581500]
std: [5.396272]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
starts_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.578594446182251, "std": 2.373150587081909}, "1": {"mean": 1.3186062574386597, "std": 1.6918489933013916}, "2": {"mean": 2.13238525390625, "std": 2.137089252471924}, "3": {"mean": 1.9998763799667358, "std": 2.630460262298584}, "4": {"mean": 0.0821475237607956, "std": 1.2746487855911255}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 2.7308456897735596, "std": 3.5874314308166504}, "1": {"mean": 3.878937005996704, "std": 4.45005989074707}, "2": {"mean": 0.3176427185535431, "std": 0.58502596616745}, "3": {"mean": 0.03686245158314705, "std": 1.0567528009414673}, "4": {"mean": 0.18341857194900513, "std": 1.2448253631591797}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 5.608956813812256, "std": 6.885017395019531}, "1": {"mean": -0.3328404426574707, "std": 1.4193942546844482}, "2": {"mean": -0.7205215692520142, "std": 2.0518083572387695}, "3": {"mean": -0.41071760654449463, "std": 1.9158086776733398}, "4": {"mean": 2.581913948059082, "std": 3.0256481170654297}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 2.5045080184936523, "std": 3.965447425842285}, "1": {"mean": 4.349981307983398, "std": 5.895235061645508}, "2": {"mean": 0.22623002529144287, "std": 0.4976437985897064}, "3": {"mean": -0.29850253462791443, "std": 1.0814223289489746}, "4": {"mean": -0.4109353721141815, "std": 0.6314740777015686}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -3.101989984512329, "std": 4.883052825927734}, "1": {"mean": 4.814450263977051, "std": 6.5662922859191895}, "2": {"mean": -1.3917558193206787, "std": 2.4185519218444824}, "3": {"mean": -0.35594165325164795, "std": 0.3803858160972595}, "4": {"mean": 0.8457450270652771, "std": 1.5389741659164429}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -3.5814995765686035, "std": 5.396271705627441}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.661715, 0.474281, 0.299373, 0.116868, -0.816163], [-0.611982, 0.30382, 0.247018, 0.350748, 0.424053], [0.859479, 0.312757, 0.166368, 0.014626, -0.400975], [1.029046, 0.537543, 0.041822, 0.131723, -0.133997], [0.344032, -0.288977, 0.178478, 0.283331, -0.66711]], "network.0.bias": [0.021511, -0.264595, 0.608547, -0.464809, 0.022962], "network.2.weight": [[0.509788, -0.43765, 0.656587, 0.405089, 0.107118], [0.813715, -0.231189, 0.637924, 0.518339, 0.508587], [0.15216, 0.211637, 0.019123, -0.255781, -0.247799], [-0.115025, 0.431222, -0.045756, -0.145499, -0.174748], [-0.051546, -0.263635, 0.407857, 0.144059, -0.154595]], "network.2.bias": [0.19461, 0.140957, 0.316841, 0.08798, -0.431194], "network.4.weight": [[0.856853, 0.724447, 0.101234, -0.501174, 0.645611], [-0.127601, -0.092253, -0.003605, 0.399416, -0.477333], [-0.15833, -0.159055, 0.433871, 0.248399, -0.652623], [-0.107826, -0.152747, 0.44208, 0.171978, -0.737809], [0.317192, 0.411297, -0.272707, -0.212456, -0.001264]], "network.4.bias": [0.205314, 0.477899, 0.421831, 0.622686, 0.235168], "network.6.weight": [[0.363907, 0.131941, -0.956479, -0.409237, 0.380466], [0.509988, -0.676351, -0.427442, -0.748461, 0.634577], [-0.075981, 0.446881, 0.071746, 0.071431, 0.060973], [-0.229254, 0.048606, 0.499564, 0.596475, 0.278196], [-0.270615, 0.216613, 0.027147, -0.134464, 0.408341]], "network.6.bias": [-0.174491, 0.374604, 0.353764, -0.061768, 0.06366], "network.8.weight": [[-0.905838, -0.25115, 0.229377, 0.215861, 0.059608], [0.614306, 0.719283, -0.567579, -0.324622, -0.277807], [-0.342369, -0.19598, 0.005602, 0.165553, 0.104569], [-0.175061, 0.049354, 0.002312, 0.030264, 0.070509], [0.177345, 0.129162, -0.433425, -0.253123, 0.364602]], "network.8.bias": [0.481842, -0.030717, 0.43459, -0.09958, -0.081491], "network.10.weight": [[0.193171, -0.79034, 0.223476, 0.321291, -0.166633]], "network.10.bias": [0.461542]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.721134215593338, "train_acc": 0.435, "val_loss": 0.684816837310791, "val_acc": 0.56}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6809460520744324, "train_acc": 0.565, "val_loss": 0.6651077270507812, "val_acc": 0.52}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6439589560031891, "train_acc": 0.565, "val_loss": 0.6194798946380615, "val_acc": 0.52}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.6352628171443939, "train_acc": 0.605, "val_loss": 0.5623038411140442, "val_acc": 0.78}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.5842922925949097, "train_acc": 0.705, "val_loss": 0.48203662037849426, "val_acc": 0.78}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.519352912902832, "train_acc": 0.74, "val_loss": 0.4339640438556671, "val_acc": 0.78}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.4900536388158798, "train_acc": 0.74, "val_loss": 0.418974906206131, "val_acc": 0.8}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.508700042963028, "train_acc": 0.755, "val_loss": 0.4410156011581421, "val_acc": 0.8}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.5007309317588806, "train_acc": 0.735, "val_loss": 0.4906270503997803, "val_acc": 0.72}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.5015704929828644, "train_acc": 0.74, "val_loss": 0.44148579239845276, "val_acc": 0.78}], "summary": {"total_epochs": 10, "degraded_epochs": 3, "improved_epochs": 7, "patterns": ["starts_with"], "degraded_stage": {"initial_val_loss": 0.684816837310791, "final_val_loss": 0.6194798946380615, "initial_val_acc": 0.56, "final_val_acc": 0.52, "best_val_acc": 0.52}, "improved_stage": {"initial_val_loss": 0.5623038411140442, "final_val_loss": 0.44148579239845276, "initial_val_acc": 0.78, "final_val_acc": 0.78, "best_val_acc": 0.8, "best_epoch": 6}, "improvement": 0.28, "first_improvement_epoch": 2}}
|
31
|
{"target_pattern": "vowel_consonant", "degraded_accuracy": 0.52, "improved_accuracy": 0.68, "improvement": 0.16000000000000003, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 9762, "learning_rate": 0.031389579657191864, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "vowel_consonant", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["vowel_consonant"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.195058,
-0.150829,
0.049925,
-0.103648,
0.010121
],
[
-0.338665,
0.107725,
0.215572,
0.289825,
-0.319314
],
[
0.301924,
0.094381,
0.482261,
-0.175893,
-0.029102
],
[
-0.28004,
0.442565,
0.2978,
-0.410813,
0.415188
],
[
-0.366918,
0.249622,
-0.033521,
-0.434858,
0.209432
],
[
-0.030641,
-0.206893,
-0.386778,
-0.569501,
0.267894
],
[
0.140441,
-0.187967,
-0.383834,
-0.06031,
-0.401245
]
],
"network.0.bias": [
0.036887,
0.449213,
0.156055,
-0.116876,
0.187046,
-0.205742,
-0.481071
],
"network.2.weight": [
[
-0.191023,
0.081852,
-0.31167,
0.238445,
0.240731,
0.478319,
-0.302662
],
[
-0.369774,
-0.266941,
0.365852,
0.076004,
0.148371,
-0.45448,
-0.372411
],
[
0.24806,
-0.194197,
-0.10556,
-0.331577,
-0.182127,
0.33523,
0.514197
],
[
-0.209004,
0.195822,
-0.146126,
0.338648,
0.376731,
0.341274,
-0.427958
],
[
0.063124,
0.351208,
-0.201775,
0.306057,
0.033104,
0.228185,
-0.511532
],
[
-0.356984,
0.208444,
-0.371907,
-0.090796,
0.166417,
0.342792,
-0.180725
],
[
0.341451,
-0.189365,
-0.161833,
0.291293,
-0.134679,
0.258818,
0.135075
]
],
"network.2.bias": [
0.309165,
0.607393,
0.12148,
-0.208456,
-0.072783,
-0.029697,
-0.342616
],
"network.4.weight": [
[
-0.387785,
0.476025,
-0.451298,
-0.532334,
-0.106059,
-0.155327,
-0.148759
],
[
0.010733,
-0.162131,
-0.093645,
-0.416355,
-0.19413,
0.092409,
-0.186017
],
[
-0.519242,
-0.100104,
0.174387,
-0.341578,
-0.201352,
-0.1361,
-0.143863
],
[
-0.252947,
0.453327,
0.253683,
-0.438109,
-0.497733,
-0.308044,
-0.497123
],
[
-0.567107,
0.320566,
-0.20405,
-0.502393,
-0.432719,
-0.290456,
-0.106913
],
[
-0.286759,
0.211749,
0.021082,
-0.168735,
-0.438133,
0.274321,
0.061301
],
[
-0.489593,
0.114048,
0.028846,
-0.314745,
0.272616,
-0.470754,
0.00032
]
],
"network.4.bias": [
-0.09038,
0.437399,
0.485619,
0.325122,
0.469681,
0.211675,
-0.358974
],
"network.6.weight": [
[
0.432175,
0.318834,
0.179982,
-0.025614,
0.534517,
0.569114,
0.202115
],
[
0.409388,
-0.025316,
0.474812,
0.426047,
0.005733,
0.161508,
0.196789
],
[
0.150799,
-0.156685,
-0.024715,
0.298513,
0.601955,
0.401805,
-0.028248
],
[
0.186544,
0.363004,
0.121256,
-0.031812,
0.454127,
0.474351,
0.417537
],
[
0.374363,
0.336636,
0.092379,
0.447151,
0.492192,
-0.060832,
0.068753
],
[
0.388971,
-0.118045,
0.439375,
0.294829,
0.642643,
-0.034001,
0.163935
],
[
0.463556,
0.112658,
0.497951,
0.37107,
0.498572,
0.278685,
0.104773
]
],
"network.6.bias": [
-0.027362,
-0.138707,
-0.274297,
0.184213,
-0.258365,
0.056233,
0.239448
],
"network.8.weight": [
[
0.19212,
0.233526,
0.285736,
0.337268,
0.422312,
0.228362,
0.508789
],
[
-0.042766,
0.372693,
0.239487,
0.23414,
0.052212,
0.115183,
0.371748
],
[
0.452797,
0.134401,
0.235606,
0.019521,
0.228038,
0.181173,
0.439312
],
[
-0.439492,
-0.609114,
-0.268436,
0.100067,
-0.078423,
0.166699,
-0.338756
],
[
-0.28611,
-0.110326,
-0.491285,
-0.375262,
-0.478099,
-0.32478,
-0.103241
],
[
-0.103793,
-0.144581,
0.263392,
-0.108575,
-0.040829,
-0.074609,
-0.157285
],
[
0.099097,
-0.259406,
0.027338,
0.178884,
-0.386269,
-0.485333,
-0.446354
]
],
"network.8.bias": [
0.192712,
-0.270521,
-0.140738,
0.309171,
-0.037236,
-0.30252,
0.073156
],
"network.10.weight": [
[
-0.05214,
-0.150151,
-0.052579,
-0.056953,
0.245973,
0.288411,
0.38815
],
[
-0.019943,
0.016099,
-0.436728,
0.562129,
-0.008295,
-0.021154,
0.533281
],
[
0.394113,
0.180976,
0.387065,
-0.22255,
-0.260534,
-0.078509,
-0.238644
],
[
0.038666,
-0.193641,
-0.416394,
0.637632,
0.480579,
-0.286716,
0.504205
],
[
0.193951,
-0.16166,
0.129252,
-0.268129,
-0.01641,
0.12544,
0.154352
],
[
0.283976,
0.306614,
0.449011,
-0.087049,
-0.496762,
-0.062917,
-0.159115
],
[
-0.225721,
-0.380586,
-0.227581,
0.465214,
0.280892,
-0.365778,
0.269605
]
],
"network.10.bias": [
0.261483,
-0.133967,
0.252469,
-0.125033,
-0.189091,
0.024271,
-0.116592
],
"network.12.weight": [
[
0.272804,
0.157605,
-0.214042,
0.49567,
-0.081115,
-0.418499,
0.402107
]
],
"network.12.bias": [
-0.038801
]
}
## Activation Signature
### 0
mean: [-0.586193, 0.902062, 1.298068, 0.594256, -0.558946, -2.319678, -2.074805]
std: [0.596628, 1.106027, 1.367757, 1.287778, 0.988374, 1.584900, 1.207227]
### 2
mean: [0.200482, 0.973042, -0.531416, 0.100768, 0.250744, -0.342880, -0.559059]
std: [0.416404, 0.627770, 0.427231, 0.403146, 0.440629, 0.633514, 0.283364]
### 4
mean: [0.236919, 0.246246, 0.226290, 0.553281, 0.531709, 0.213233, -0.323139]
std: [0.469980, 0.179537, 0.246925, 0.526427, 0.511192, 0.286979, 0.213593]
### 6
mean: [0.436228, 0.223774, 0.200008, 0.517284, 0.299286, 0.589001, 0.860116]
std: [0.520444, 0.435771, 0.532249, 0.383387, 0.559183, 0.603824, 0.668185]
### 8
mean: [1.115451, 0.280427, 0.590403, -0.177214, -0.788747, -0.523704, -0.549562]
std: [1.088315, 0.637168, 0.877133, 0.664353, 0.994566, 0.196705, 0.736398]
### 10
mean: [0.054019, -0.387140, 0.948845, -0.353242, -0.006857, 0.694631, -0.534469]
std: [0.194189, 0.466169, 0.904899, 0.504935, 0.265774, 0.883852, 0.713481]
### 12
mean: [-0.534116]
std: [0.660249]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
vowel_consonant
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.195058,
-0.150829,
0.049925,
-0.103648,
0.010121
],
[
-0.338665,
0.107725,
0.215572,
0.289825,
-0.319314
],
[
0.301924,
0.094381,
0.482261,
-0.175893,
-0.029102
],
[
-0.28004,
0.442565,
0.2978,
-0.410813,
0.415188
],
[
-0.366918,
0.249622,
-0.033521,
-0.434858,
0.209432
],
[
-0.030641,
-0.206893,
-0.386778,
-0.569501,
0.267894
],
[
0.140441,
-0.187967,
-0.383834,
-0.06031,
-0.401245
]
],
"network.0.bias": [
0.036887,
0.449213,
0.156055,
-0.116876,
0.187046,
-0.205742,
-0.481071
],
"network.2.weight": [
[
-0.191023,
0.081852,
-0.31167,
0.238445,
0.240731,
0.478319,
-0.302662
],
[
-0.369774,
-0.266941,
0.365852,
0.076004,
0.148371,
-0.45448,
-0.372411
],
[
0.24806,
-0.194197,
-0.10556,
-0.331577,
-0.182127,
0.33523,
0.514197
],
[
-0.209004,
0.195822,
-0.146126,
0.338648,
0.376731,
0.341274,
-0.427958
],
[
0.063124,
0.351208,
-0.201775,
0.306057,
0.033104,
0.228185,
-0.511532
],
[
-0.356984,
0.208444,
-0.371907,
-0.090796,
0.166417,
0.342792,
-0.180725
],
[
0.341451,
-0.189365,
-0.161833,
0.291293,
-0.134679,
0.258818,
0.135075
]
],
"network.2.bias": [
0.309165,
0.607393,
0.12148,
-0.208456,
-0.072783,
-0.029697,
-0.342616
],
"network.4.weight": [
[
-0.387785,
0.476025,
-0.451298,
-0.532334,
-0.106059,
-0.155327,
-0.148759
],
[
0.010733,
-0.162131,
-0.093645,
-0.416355,
-0.19413,
0.092409,
-0.186017
],
[
-0.519242,
-0.100104,
0.174387,
-0.341578,
-0.201352,
-0.1361,
-0.143863
],
[
-0.252947,
0.453327,
0.253683,
-0.438109,
-0.497733,
-0.308044,
-0.497123
],
[
-0.567107,
0.320566,
-0.20405,
-0.502393,
-0.432719,
-0.290456,
-0.106913
],
[
-0.286759,
0.211749,
0.021082,
-0.168735,
-0.438133,
0.274321,
0.061301
],
[
-0.489593,
0.114048,
0.028846,
-0.314745,
0.272616,
-0.470754,
0.00032
]
],
"network.4.bias": [
-0.09038,
0.437399,
0.485619,
0.325122,
0.469681,
0.211675,
-0.358974
],
"network.6.weight": [
[
0.432175,
0.318834,
0.179982,
-0.025614,
0.534517,
0.569114,
0.202115
],
[
0.409388,
-0.025316,
0.474812,
0.426047,
0.005733,
0.161508,
0.196789
],
[
0.150799,
-0.156685,
-0.024715,
0.298513,
0.601955,
0.401805,
-0.028248
],
[
0.186544,
0.363004,
0.121256,
-0.031812,
0.454127,
0.474351,
0.417537
],
[
0.374363,
0.336636,
0.092379,
0.447151,
0.492192,
-0.060832,
0.068753
],
[
0.388971,
-0.118045,
0.439375,
0.294829,
0.642643,
-0.034001,
0.163935
],
[
0.463556,
0.112658,
0.497951,
0.37107,
0.498572,
0.278685,
0.104773
]
],
"network.6.bias": [
-0.027362,
-0.138707,
-0.274297,
0.184213,
-0.258365,
0.056233,
0.239448
],
"network.8.weight": [
[
0.19212,
0.233526,
0.285736,
0.337268,
0.422312,
0.228362,
0.508789
],
[
-0.042766,
0.372693,
0.239487,
0.23414,
0.052212,
0.115183,
0.371748
],
[
0.452797,
0.134401,
0.235606,
0.019521,
0.228038,
0.181173,
0.439312
],
[
-0.439492,
-0.609114,
-0.268436,
0.100067,
-0.078423,
0.166699,
-0.338756
],
[
-0.28611,
-0.110326,
-0.491285,
-0.375262,
-0.478099,
-0.32478,
-0.103241
],
[
-0.103793,
-0.144581,
0.263392,
-0.108575,
-0.040829,
-0.074609,
-0.157285
],
[
0.099097,
-0.259406,
0.027338,
0.178884,
-0.386269,
-0.485333,
-0.446354
]
],
"network.8.bias": [
0.192712,
-0.270521,
-0.140738,
0.309171,
-0.037236,
-0.30252,
0.073156
],
"network.10.weight": [
[
-0.05214,
-0.150151,
-0.052579,
-0.056953,
0.245973,
0.288411,
0.38815
],
[
-0.019943,
0.016099,
-0.436728,
0.562129,
-0.008295,
-0.021154,
0.533281
],
[
0.394113,
0.180976,
0.387065,
-0.22255,
-0.260534,
-0.078509,
-0.238644
],
[
0.038666,
-0.193641,
-0.416394,
0.637632,
0.480579,
-0.286716,
0.504205
],
[
0.193951,
-0.16166,
0.129252,
-0.268129,
-0.01641,
0.12544,
0.154352
],
[
0.283976,
0.306614,
0.449011,
-0.087049,
-0.496762,
-0.062917,
-0.159115
],
[
-0.225721,
-0.380586,
-0.227581,
0.465214,
0.280892,
-0.365778,
0.269605
]
],
"network.10.bias": [
0.261483,
-0.133967,
0.252469,
-0.125033,
-0.189091,
0.024271,
-0.116592
],
"network.12.weight": [
[
0.272804,
0.157605,
-0.214042,
0.49567,
-0.081115,
-0.418499,
0.402107
]
],
"network.12.bias": [
-0.038801
]
}
## Activation Signature
### 0
mean: [-0.586193, 0.902062, 1.298068, 0.594256, -0.558946, -2.319678, -2.074805]
std: [0.596628, 1.106027, 1.367757, 1.287778, 0.988374, 1.584900, 1.207227]
### 2
mean: [0.200482, 0.973042, -0.531416, 0.100768, 0.250744, -0.342880, -0.559059]
std: [0.416404, 0.627770, 0.427231, 0.403146, 0.440629, 0.633514, 0.283364]
### 4
mean: [0.236919, 0.246246, 0.226290, 0.553281, 0.531709, 0.213233, -0.323139]
std: [0.469980, 0.179537, 0.246925, 0.526427, 0.511192, 0.286979, 0.213593]
### 6
mean: [0.436228, 0.223774, 0.200008, 0.517284, 0.299286, 0.589001, 0.860116]
std: [0.520444, 0.435771, 0.532249, 0.383387, 0.559183, 0.603824, 0.668185]
### 8
mean: [1.115451, 0.280427, 0.590403, -0.177214, -0.788747, -0.523704, -0.549562]
std: [1.088315, 0.637168, 0.877133, 0.664353, 0.994566, 0.196705, 0.736398]
### 10
mean: [0.054019, -0.387140, 0.948845, -0.353242, -0.006857, 0.694631, -0.534469]
std: [0.194189, 0.466169, 0.904899, 0.504935, 0.265774, 0.883852, 0.713481]
### 12
mean: [-0.534116]
std: [0.660249]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
vowel_consonant
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -0.586192786693573, "std": 0.5966277122497559}, "1": {"mean": 0.9020617008209229, "std": 1.1060274839401245}, "2": {"mean": 1.2980684041976929, "std": 1.3677568435668945}, "3": {"mean": 0.5942557454109192, "std": 1.287778377532959}, "4": {"mean": -0.5589461922645569, "std": 0.9883736371994019}, "5": {"mean": -2.3196775913238525, "std": 1.5848995447158813}, "6": {"mean": -2.0748047828674316, "std": 1.207227110862732}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.20048165321350098, "std": 0.4164040982723236}, "1": {"mean": 0.9730415344238281, "std": 0.6277700066566467}, "2": {"mean": -0.5314164161682129, "std": 0.42723143100738525}, "3": {"mean": 0.10076785832643509, "std": 0.40314579010009766}, "4": {"mean": 0.25074371695518494, "std": 0.4406287968158722}, "5": {"mean": -0.3428798317909241, "std": 0.6335142850875854}, "6": {"mean": -0.5590590834617615, "std": 0.2833636403083801}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.23691895604133606, "std": 0.46998047828674316}, "1": {"mean": 0.24624580144882202, "std": 0.17953726649284363}, "2": {"mean": 0.2262900471687317, "std": 0.24692490696907043}, "3": {"mean": 0.5532807111740112, "std": 0.5264273285865784}, "4": {"mean": 0.5317091345787048, "std": 0.511191725730896}, "5": {"mean": 0.21323314309120178, "std": 0.28697890043258667}, "6": {"mean": -0.32313892245292664, "std": 0.213592991232872}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.4362281858921051, "std": 0.5204443335533142}, "1": {"mean": 0.22377435863018036, "std": 0.43577051162719727}, "2": {"mean": 0.2000075876712799, "std": 0.5322489142417908}, "3": {"mean": 0.5172839760780334, "std": 0.3833865225315094}, "4": {"mean": 0.2992858290672302, "std": 0.5591832995414734}, "5": {"mean": 0.5890008807182312, "std": 0.6038244962692261}, "6": {"mean": 0.8601162433624268, "std": 0.6681848168373108}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 1.115451455116272, "std": 1.088315486907959}, "1": {"mean": 0.28042706847190857, "std": 0.6371684074401855}, "2": {"mean": 0.5904026627540588, "std": 0.877133309841156}, "3": {"mean": -0.17721372842788696, "std": 0.6643533110618591}, "4": {"mean": -0.7887468934059143, "std": 0.9945656657218933}, "5": {"mean": -0.5237037539482117, "std": 0.19670458137989044}, "6": {"mean": -0.5495621562004089, "std": 0.7363982200622559}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": 0.05401868373155594, "std": 0.19418905675411224}, "1": {"mean": -0.3871396780014038, "std": 0.4661693871021271}, "2": {"mean": 0.9488449096679688, "std": 0.9048987030982971}, "3": {"mean": -0.35324186086654663, "std": 0.5049354434013367}, "4": {"mean": -0.006856965366750956, "std": 0.2657735049724579}, "5": {"mean": 0.69463050365448, "std": 0.8838522434234619}, "6": {"mean": -0.5344685316085815, "std": 0.7134814858436584}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -0.5341164469718933, "std": 0.6602490544319153}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.195058, -0.150829, 0.049925, -0.103648, 0.010121], [-0.338665, 0.107725, 0.215572, 0.289825, -0.319314], [0.301924, 0.094381, 0.482261, -0.175893, -0.029102], [-0.28004, 0.442565, 0.2978, -0.410813, 0.415188], [-0.366918, 0.249622, -0.033521, -0.434858, 0.209432], [-0.030641, -0.206893, -0.386778, -0.569501, 0.267894], [0.140441, -0.187967, -0.383834, -0.06031, -0.401245]], "network.0.bias": [0.036887, 0.449213, 0.156055, -0.116876, 0.187046, -0.205742, -0.481071], "network.2.weight": [[-0.191023, 0.081852, -0.31167, 0.238445, 0.240731, 0.478319, -0.302662], [-0.369774, -0.266941, 0.365852, 0.076004, 0.148371, -0.45448, -0.372411], [0.24806, -0.194197, -0.10556, -0.331577, -0.182127, 0.33523, 0.514197], [-0.209004, 0.195822, -0.146126, 0.338648, 0.376731, 0.341274, -0.427958], [0.063124, 0.351208, -0.201775, 0.306057, 0.033104, 0.228185, -0.511532], [-0.356984, 0.208444, -0.371907, -0.090796, 0.166417, 0.342792, -0.180725], [0.341451, -0.189365, -0.161833, 0.291293, -0.134679, 0.258818, 0.135075]], "network.2.bias": [0.309165, 0.607393, 0.12148, -0.208456, -0.072783, -0.029697, -0.342616], "network.4.weight": [[-0.387785, 0.476025, -0.451298, -0.532334, -0.106059, -0.155327, -0.148759], [0.010733, -0.162131, -0.093645, -0.416355, -0.19413, 0.092409, -0.186017], [-0.519242, -0.100104, 0.174387, -0.341578, -0.201352, -0.1361, -0.143863], [-0.252947, 0.453327, 0.253683, -0.438109, -0.497733, -0.308044, -0.497123], [-0.567107, 0.320566, -0.20405, -0.502393, -0.432719, -0.290456, -0.106913], [-0.286759, 0.211749, 0.021082, -0.168735, -0.438133, 0.274321, 0.061301], [-0.489593, 0.114048, 0.028846, -0.314745, 0.272616, -0.470754, 0.00032]], "network.4.bias": [-0.09038, 0.437399, 0.485619, 0.325122, 0.469681, 0.211675, -0.358974], "network.6.weight": [[0.432175, 0.318834, 0.179982, -0.025614, 0.534517, 0.569114, 0.202115], [0.409388, -0.025316, 0.474812, 0.426047, 0.005733, 0.161508, 0.196789], [0.150799, -0.156685, -0.024715, 0.298513, 0.601955, 0.401805, -0.028248], [0.186544, 0.363004, 0.121256, -0.031812, 0.454127, 0.474351, 0.417537], [0.374363, 0.336636, 0.092379, 0.447151, 0.492192, -0.060832, 0.068753], [0.388971, -0.118045, 0.439375, 0.294829, 0.642643, -0.034001, 0.163935], [0.463556, 0.112658, 0.497951, 0.37107, 0.498572, 0.278685, 0.104773]], "network.6.bias": [-0.027362, -0.138707, -0.274297, 0.184213, -0.258365, 0.056233, 0.239448], "network.8.weight": [[0.19212, 0.233526, 0.285736, 0.337268, 0.422312, 0.228362, 0.508789], [-0.042766, 0.372693, 0.239487, 0.23414, 0.052212, 0.115183, 0.371748], [0.452797, 0.134401, 0.235606, 0.019521, 0.228038, 0.181173, 0.439312], [-0.439492, -0.609114, -0.268436, 0.100067, -0.078423, 0.166699, -0.338756], [-0.28611, -0.110326, -0.491285, -0.375262, -0.478099, -0.32478, -0.103241], [-0.103793, -0.144581, 0.263392, -0.108575, -0.040829, -0.074609, -0.157285], [0.099097, -0.259406, 0.027338, 0.178884, -0.386269, -0.485333, -0.446354]], "network.8.bias": [0.192712, -0.270521, -0.140738, 0.309171, -0.037236, -0.30252, 0.073156], "network.10.weight": [[-0.05214, -0.150151, -0.052579, -0.056953, 0.245973, 0.288411, 0.38815], [-0.019943, 0.016099, -0.436728, 0.562129, -0.008295, -0.021154, 0.533281], [0.394113, 0.180976, 0.387065, -0.22255, -0.260534, -0.078509, -0.238644], [0.038666, -0.193641, -0.416394, 0.637632, 0.480579, -0.286716, 0.504205], [0.193951, -0.16166, 0.129252, -0.268129, -0.01641, 0.12544, 0.154352], [0.283976, 0.306614, 0.449011, -0.087049, -0.496762, -0.062917, -0.159115], [-0.225721, -0.380586, -0.227581, 0.465214, 0.280892, -0.365778, 0.269605]], "network.10.bias": [0.261483, -0.133967, 0.252469, -0.125033, -0.189091, 0.024271, -0.116592], "network.12.weight": [[0.272804, 0.157605, -0.214042, 0.49567, -0.081115, -0.418499, 0.402107]], "network.12.bias": [-0.038801]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6880340874195099, "train_acc": 0.56, "val_loss": 0.6954350471496582, "val_acc": 0.52}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6877690255641937, "train_acc": 0.56, "val_loss": 0.6935162544250488, "val_acc": 0.52}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6826429665088654, "train_acc": 0.56, "val_loss": 0.687935471534729, "val_acc": 0.52}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6711989641189575, "train_acc": 0.56, "val_loss": 0.6535412669181824, "val_acc": 0.52}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6739811301231384, "train_acc": 0.505, "val_loss": 0.604214072227478, "val_acc": 0.68}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.629450798034668, "train_acc": 0.63, "val_loss": 0.5838956236839294, "val_acc": 0.68}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.615898609161377, "train_acc": 0.68, "val_loss": 0.5885204076766968, "val_acc": 0.68}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.5734299421310425, "train_acc": 0.695, "val_loss": 0.7988225817680359, "val_acc": 0.64}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.556905210018158, "train_acc": 0.705, "val_loss": 0.5917865037918091, "val_acc": 0.66}], "summary": {"total_epochs": 9, "degraded_epochs": 4, "improved_epochs": 5, "patterns": ["vowel_consonant"], "degraded_stage": {"initial_val_loss": 0.6954350471496582, "final_val_loss": 0.6535412669181824, "initial_val_acc": 0.52, "final_val_acc": 0.52, "best_val_acc": 0.52}, "improved_stage": {"initial_val_loss": 0.604214072227478, "final_val_loss": 0.5917865037918091, "initial_val_acc": 0.68, "final_val_acc": 0.66, "best_val_acc": 0.68, "best_epoch": 4}, "improvement": 0.16000000000000003, "first_improvement_epoch": 3}}
|
32
|
{"target_pattern": "mountain_pattern", "degraded_accuracy": 0.48, "improved_accuracy": 0.9, "improvement": 0.42000000000000004, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 3215, "learning_rate": 0.02854826421987649, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "mountain_pattern", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["mountain_pattern"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.041663,
0.699494,
-0.370109,
-0.055518,
0.37532
],
[
0.415235,
0.049605,
-0.500714,
0.295166,
-0.168494
],
[
0.618103,
0.398205,
-0.004458,
-0.193005,
0.19461
],
[
0.779129,
0.182382,
-0.151127,
0.196913,
-0.259204
],
[
-0.172748,
-0.082255,
0.103999,
0.098421,
0.399074
],
[
0.845246,
-0.067868,
0.260781,
-0.512105,
0.027937
],
[
0.04458,
0.744655,
0.108329,
-0.175819,
0.116924
],
[
-0.209749,
0.127156,
-0.143611,
0.506628,
0.09071
]
],
"network.0.bias": [
0.13131,
-0.213645,
0.091672,
0.178292,
-0.148468,
0.253457,
0.023517,
-0.6239
],
"network.2.weight": [
[
0.577128,
0.186766,
0.028665,
0.547734,
-0.147148,
-0.251403,
0.357,
-0.262032
],
[
-0.331759,
-0.493677,
-0.044894,
-0.546796,
0.170333,
-0.153822,
0.275752,
0.405499
],
[
0.179952,
-0.032974,
0.196904,
0.300646,
-0.175579,
-0.228065,
-0.201824,
-0.373604
],
[
0.267328,
-0.223957,
0.414714,
0.044925,
0.314963,
0.224118,
0.340238,
0.029007
],
[
0.286593,
0.078932,
0.489103,
0.427697,
-0.225225,
0.17601,
0.349261,
-0.132145
],
[
-0.363179,
0.496168,
-0.558986,
-0.003802,
-0.283132,
-0.069792,
-0.358506,
0.223431
],
[
-0.456953,
0.23952,
-0.288122,
-0.369278,
-0.280226,
0.156875,
0.346689,
-0.421706
],
[
-0.150312,
0.275536,
0.601068,
0.509489,
-0.180382,
0.565331,
0.299748,
0.111512
]
],
"network.2.bias": [
-0.332425,
0.214845,
-0.586985,
0.080348,
-0.183798,
-0.077338,
0.355216,
-0.205542
],
"network.4.weight": [
[
0.254426,
-0.432683,
-0.231034,
0.183885,
0.286685,
0.021052,
-0.217804,
0.280132
],
[
0.621413,
-0.413862,
0.182835,
0.119338,
0.678199,
0.070137,
-0.719663,
0.555615
],
[
0.584221,
-0.348351,
-0.020313,
0.316174,
0.446985,
0.262904,
-0.441814,
0.507748
],
[
-0.194856,
0.477087,
-0.168366,
0.520619,
0.188851,
-0.166965,
0.250392,
0.016796
],
[
0.183862,
-0.522047,
-0.458858,
0.390867,
0.256126,
0.400147,
-0.064308,
0.256242
],
[
-0.507932,
0.030895,
0.270009,
-0.564018,
-0.233496,
0.418267,
0.476701,
-0.150277
],
[
0.333962,
0.013909,
-0.257583,
0.371891,
0.327481,
0.580096,
-0.690953,
0.343973
],
[
-0.304267,
0.039045,
0.041729,
0.326938,
-0.262094,
-0.34206,
0.677786,
-0.1206
]
],
"network.4.bias": [
-0.175937,
0.335751,
0.09528,
-0.007821,
-0.300697,
-0.322923,
-0.063228,
0.333089
],
"network.6.weight": [
[
0.388631,
0.408027,
0.476126,
-0.263616,
0.439984,
0.452692,
0.380785,
-0.173827
],
[
-0.67909,
0.025495,
-0.308372,
0.41467,
-0.263956,
-0.134328,
0.460982,
-0.016058
],
[
0.236222,
0.086339,
-0.514445,
0.135388,
0.376333,
-0.144392,
-0.058987,
0.276316
],
[
0.192111,
0.404943,
0.388523,
-0.283497,
0.44946,
0.30245,
0.373748,
-0.58941
],
[
-0.262548,
-0.271472,
0.258139,
0.411953,
-0.44835,
-0.205389,
-0.021779,
0.426184
],
[
0.303931,
0.571954,
0.104637,
0.24125,
0.321002,
-0.139709,
0.599548,
-0.426724
],
[
0.399428,
0.704407,
0.581976,
0.200714,
0.063915,
0.327546,
0.248157,
-0.552288
],
[
-0.168508,
-0.194677,
0.112312,
0.534702,
-0.363437,
-0.032599,
-0.108664,
0.218058
]
],
"network.6.bias": [
-0.118916,
0.062382,
-0.173926,
-0.008134,
0.161229,
-0.012763,
-0.286789,
0.266245
],
"network.8.weight": [
[
-0.306822,
0.379969,
-0.228077,
-0.098203,
0.112407,
0.314673,
-0.372148,
0.07193
],
[
0.448289,
-0.17286,
-0.090741,
0.484953,
-0.282052,
0.629839,
0.223094,
-0.013346
],
[
0.203256,
-0.078996,
-0.398064,
0.574195,
-0.245486,
0.31903,
0.389317,
-0.019726
],
[
0.200943,
0.296101,
-0.202987,
-0.234971,
-0.009452,
-0.346187,
-0.232034,
-0.233087
],
[
-0.235726,
0.172794,
0.207257,
-0.090288,
0.17948,
-0.15368,
0.219656,
0.063816
],
[
0.035801,
0.405714,
0.135062,
-0.455672,
0.53782,
0.166163,
-0.142441,
0.416131
],
[
-0.021739,
-0.076163,
-0.433016,
-0.262143,
0.153542,
0.214167,
0.147421,
-0.32794
],
[
0.567273,
-0.111246,
-0.278879,
0.370108,
-0.208263,
0.550177,
0.564333,
-0.49347
]
],
"network.8.bias": [
0.353161,
-0.270149,
-0.289938,
-0.551146,
0.035301,
0.008448,
-0.149594,
-0.199292
],
"network.10.weight": [
[
0.404954,
-0.362393,
-0.302683,
-0.382822,
0.005542,
0.216584,
-0.171292,
-0.52494
]
],
"network.10.bias": [
0.052675
]
}
## Activation Signature
### 0
mean: [0.924820, -0.227295, 1.403301, 1.267112, 0.407161, 0.662665, 1.429380, 0.245110]
std: [1.372224, 1.259952, 1.725994, 1.693768, 0.866603, 2.240265, 1.421347, 1.284733]
### 2
mean: [1.031074, -0.495316, -0.582690, 1.787450, 1.843221, -1.652181, -0.598993, 2.217830]
std: [1.604196, 1.392247, 0.552115, 1.655411, 2.343826, 1.739290, 0.984952, 2.986287]
### 4
mean: [1.498011, 3.553608, 3.112741, 1.140996, 1.542422, -2.601903, 2.254264, -0.134641]
std: [2.154108, 4.368973, 3.888881, 1.056180, 2.220878, 2.600461, 2.843811, 0.987470]
### 6
mean: [4.532010, -0.753434, -0.434313, 3.976694, -0.599019, 4.776404, 5.291584, -0.517838]
std: [6.336932, 1.424308, 0.432050, 5.614497, 1.496746, 6.295711, 7.391773, 1.425253]
### 8
mean: [-1.876036, 7.894478, 6.562897, -3.484419, -0.931061, -1.474725, 0.508711, 9.450046]
std: [3.264861, 11.171435, 9.387219, 3.894605, 1.400756, 2.489836, 0.887529, 13.345182]
### 10
mean: [-9.886385]
std: [14.003289]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
mountain_pattern
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.041663,
0.699494,
-0.370109,
-0.055518,
0.37532
],
[
0.415235,
0.049605,
-0.500714,
0.295166,
-0.168494
],
[
0.618103,
0.398205,
-0.004458,
-0.193005,
0.19461
],
[
0.779129,
0.182382,
-0.151127,
0.196913,
-0.259204
],
[
-0.172748,
-0.082255,
0.103999,
0.098421,
0.399074
],
[
0.845246,
-0.067868,
0.260781,
-0.512105,
0.027937
],
[
0.04458,
0.744655,
0.108329,
-0.175819,
0.116924
],
[
-0.209749,
0.127156,
-0.143611,
0.506628,
0.09071
]
],
"network.0.bias": [
0.13131,
-0.213645,
0.091672,
0.178292,
-0.148468,
0.253457,
0.023517,
-0.6239
],
"network.2.weight": [
[
0.577128,
0.186766,
0.028665,
0.547734,
-0.147148,
-0.251403,
0.357,
-0.262032
],
[
-0.331759,
-0.493677,
-0.044894,
-0.546796,
0.170333,
-0.153822,
0.275752,
0.405499
],
[
0.179952,
-0.032974,
0.196904,
0.300646,
-0.175579,
-0.228065,
-0.201824,
-0.373604
],
[
0.267328,
-0.223957,
0.414714,
0.044925,
0.314963,
0.224118,
0.340238,
0.029007
],
[
0.286593,
0.078932,
0.489103,
0.427697,
-0.225225,
0.17601,
0.349261,
-0.132145
],
[
-0.363179,
0.496168,
-0.558986,
-0.003802,
-0.283132,
-0.069792,
-0.358506,
0.223431
],
[
-0.456953,
0.23952,
-0.288122,
-0.369278,
-0.280226,
0.156875,
0.346689,
-0.421706
],
[
-0.150312,
0.275536,
0.601068,
0.509489,
-0.180382,
0.565331,
0.299748,
0.111512
]
],
"network.2.bias": [
-0.332425,
0.214845,
-0.586985,
0.080348,
-0.183798,
-0.077338,
0.355216,
-0.205542
],
"network.4.weight": [
[
0.254426,
-0.432683,
-0.231034,
0.183885,
0.286685,
0.021052,
-0.217804,
0.280132
],
[
0.621413,
-0.413862,
0.182835,
0.119338,
0.678199,
0.070137,
-0.719663,
0.555615
],
[
0.584221,
-0.348351,
-0.020313,
0.316174,
0.446985,
0.262904,
-0.441814,
0.507748
],
[
-0.194856,
0.477087,
-0.168366,
0.520619,
0.188851,
-0.166965,
0.250392,
0.016796
],
[
0.183862,
-0.522047,
-0.458858,
0.390867,
0.256126,
0.400147,
-0.064308,
0.256242
],
[
-0.507932,
0.030895,
0.270009,
-0.564018,
-0.233496,
0.418267,
0.476701,
-0.150277
],
[
0.333962,
0.013909,
-0.257583,
0.371891,
0.327481,
0.580096,
-0.690953,
0.343973
],
[
-0.304267,
0.039045,
0.041729,
0.326938,
-0.262094,
-0.34206,
0.677786,
-0.1206
]
],
"network.4.bias": [
-0.175937,
0.335751,
0.09528,
-0.007821,
-0.300697,
-0.322923,
-0.063228,
0.333089
],
"network.6.weight": [
[
0.388631,
0.408027,
0.476126,
-0.263616,
0.439984,
0.452692,
0.380785,
-0.173827
],
[
-0.67909,
0.025495,
-0.308372,
0.41467,
-0.263956,
-0.134328,
0.460982,
-0.016058
],
[
0.236222,
0.086339,
-0.514445,
0.135388,
0.376333,
-0.144392,
-0.058987,
0.276316
],
[
0.192111,
0.404943,
0.388523,
-0.283497,
0.44946,
0.30245,
0.373748,
-0.58941
],
[
-0.262548,
-0.271472,
0.258139,
0.411953,
-0.44835,
-0.205389,
-0.021779,
0.426184
],
[
0.303931,
0.571954,
0.104637,
0.24125,
0.321002,
-0.139709,
0.599548,
-0.426724
],
[
0.399428,
0.704407,
0.581976,
0.200714,
0.063915,
0.327546,
0.248157,
-0.552288
],
[
-0.168508,
-0.194677,
0.112312,
0.534702,
-0.363437,
-0.032599,
-0.108664,
0.218058
]
],
"network.6.bias": [
-0.118916,
0.062382,
-0.173926,
-0.008134,
0.161229,
-0.012763,
-0.286789,
0.266245
],
"network.8.weight": [
[
-0.306822,
0.379969,
-0.228077,
-0.098203,
0.112407,
0.314673,
-0.372148,
0.07193
],
[
0.448289,
-0.17286,
-0.090741,
0.484953,
-0.282052,
0.629839,
0.223094,
-0.013346
],
[
0.203256,
-0.078996,
-0.398064,
0.574195,
-0.245486,
0.31903,
0.389317,
-0.019726
],
[
0.200943,
0.296101,
-0.202987,
-0.234971,
-0.009452,
-0.346187,
-0.232034,
-0.233087
],
[
-0.235726,
0.172794,
0.207257,
-0.090288,
0.17948,
-0.15368,
0.219656,
0.063816
],
[
0.035801,
0.405714,
0.135062,
-0.455672,
0.53782,
0.166163,
-0.142441,
0.416131
],
[
-0.021739,
-0.076163,
-0.433016,
-0.262143,
0.153542,
0.214167,
0.147421,
-0.32794
],
[
0.567273,
-0.111246,
-0.278879,
0.370108,
-0.208263,
0.550177,
0.564333,
-0.49347
]
],
"network.8.bias": [
0.353161,
-0.270149,
-0.289938,
-0.551146,
0.035301,
0.008448,
-0.149594,
-0.199292
],
"network.10.weight": [
[
0.404954,
-0.362393,
-0.302683,
-0.382822,
0.005542,
0.216584,
-0.171292,
-0.52494
]
],
"network.10.bias": [
0.052675
]
}
## Activation Signature
### 0
mean: [0.924820, -0.227295, 1.403301, 1.267112, 0.407161, 0.662665, 1.429380, 0.245110]
std: [1.372224, 1.259952, 1.725994, 1.693768, 0.866603, 2.240265, 1.421347, 1.284733]
### 2
mean: [1.031074, -0.495316, -0.582690, 1.787450, 1.843221, -1.652181, -0.598993, 2.217830]
std: [1.604196, 1.392247, 0.552115, 1.655411, 2.343826, 1.739290, 0.984952, 2.986287]
### 4
mean: [1.498011, 3.553608, 3.112741, 1.140996, 1.542422, -2.601903, 2.254264, -0.134641]
std: [2.154108, 4.368973, 3.888881, 1.056180, 2.220878, 2.600461, 2.843811, 0.987470]
### 6
mean: [4.532010, -0.753434, -0.434313, 3.976694, -0.599019, 4.776404, 5.291584, -0.517838]
std: [6.336932, 1.424308, 0.432050, 5.614497, 1.496746, 6.295711, 7.391773, 1.425253]
### 8
mean: [-1.876036, 7.894478, 6.562897, -3.484419, -0.931061, -1.474725, 0.508711, 9.450046]
std: [3.264861, 11.171435, 9.387219, 3.894605, 1.400756, 2.489836, 0.887529, 13.345182]
### 10
mean: [-9.886385]
std: [14.003289]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
mountain_pattern
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.9248195886611938, "std": 1.3722244501113892}, "1": {"mean": -0.22729536890983582, "std": 1.2599520683288574}, "2": {"mean": 1.4033013582229614, "std": 1.7259942293167114}, "3": {"mean": 1.2671120166778564, "std": 1.6937679052352905}, "4": {"mean": 0.40716099739074707, "std": 0.8666027188301086}, "5": {"mean": 0.6626652479171753, "std": 2.240264654159546}, "6": {"mean": 1.4293795824050903, "std": 1.4213473796844482}, "7": {"mean": 0.24511027336120605, "std": 1.2847325801849365}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 1.0310744047164917, "std": 1.6041960716247559}, "1": {"mean": -0.4953155815601349, "std": 1.392247200012207}, "2": {"mean": -0.5826897621154785, "std": 0.5521154999732971}, "3": {"mean": 1.7874499559402466, "std": 1.6554110050201416}, "4": {"mean": 1.8432211875915527, "std": 2.343825578689575}, "5": {"mean": -1.6521811485290527, "std": 1.7392902374267578}, "6": {"mean": -0.5989932417869568, "std": 0.9849516153335571}, "7": {"mean": 2.217829704284668, "std": 2.9862868785858154}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 1.4980112314224243, "std": 2.1541080474853516}, "1": {"mean": 3.5536081790924072, "std": 4.3689727783203125}, "2": {"mean": 3.1127407550811768, "std": 3.888881206512451}, "3": {"mean": 1.1409955024719238, "std": 1.0561796426773071}, "4": {"mean": 1.5424221754074097, "std": 2.2208781242370605}, "5": {"mean": -2.6019034385681152, "std": 2.600461483001709}, "6": {"mean": 2.2542643547058105, "std": 2.843811273574829}, "7": {"mean": -0.13464058935642242, "std": 0.9874695539474487}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 4.532010078430176, "std": 6.3369317054748535}, "1": {"mean": -0.7534340023994446, "std": 1.4243075847625732}, "2": {"mean": -0.4343133568763733, "std": 0.43204981088638306}, "3": {"mean": 3.976694107055664, "std": 5.61449670791626}, "4": {"mean": -0.599018931388855, "std": 1.496746301651001}, "5": {"mean": 4.776403903961182, "std": 6.295710563659668}, "6": {"mean": 5.291584491729736, "std": 7.391773223876953}, "7": {"mean": -0.5178375244140625, "std": 1.42525315284729}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.8760355710983276, "std": 3.2648608684539795}, "1": {"mean": 7.8944783210754395, "std": 11.171435356140137}, "2": {"mean": 6.562897205352783, "std": 9.387219429016113}, "3": {"mean": -3.484419345855713, "std": 3.8946049213409424}, "4": {"mean": -0.9310614466667175, "std": 1.4007562398910522}, "5": {"mean": -1.4747254848480225, "std": 2.489835500717163}, "6": {"mean": 0.5087113976478577, "std": 0.8875294327735901}, "7": {"mean": 9.450045585632324, "std": 13.345182418823242}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -9.886384963989258, "std": 14.003289222717285}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.041663, 0.699494, -0.370109, -0.055518, 0.37532], [0.415235, 0.049605, -0.500714, 0.295166, -0.168494], [0.618103, 0.398205, -0.004458, -0.193005, 0.19461], [0.779129, 0.182382, -0.151127, 0.196913, -0.259204], [-0.172748, -0.082255, 0.103999, 0.098421, 0.399074], [0.845246, -0.067868, 0.260781, -0.512105, 0.027937], [0.04458, 0.744655, 0.108329, -0.175819, 0.116924], [-0.209749, 0.127156, -0.143611, 0.506628, 0.09071]], "network.0.bias": [0.13131, -0.213645, 0.091672, 0.178292, -0.148468, 0.253457, 0.023517, -0.6239], "network.2.weight": [[0.577128, 0.186766, 0.028665, 0.547734, -0.147148, -0.251403, 0.357, -0.262032], [-0.331759, -0.493677, -0.044894, -0.546796, 0.170333, -0.153822, 0.275752, 0.405499], [0.179952, -0.032974, 0.196904, 0.300646, -0.175579, -0.228065, -0.201824, -0.373604], [0.267328, -0.223957, 0.414714, 0.044925, 0.314963, 0.224118, 0.340238, 0.029007], [0.286593, 0.078932, 0.489103, 0.427697, -0.225225, 0.17601, 0.349261, -0.132145], [-0.363179, 0.496168, -0.558986, -0.003802, -0.283132, -0.069792, -0.358506, 0.223431], [-0.456953, 0.23952, -0.288122, -0.369278, -0.280226, 0.156875, 0.346689, -0.421706], [-0.150312, 0.275536, 0.601068, 0.509489, -0.180382, 0.565331, 0.299748, 0.111512]], "network.2.bias": [-0.332425, 0.214845, -0.586985, 0.080348, -0.183798, -0.077338, 0.355216, -0.205542], "network.4.weight": [[0.254426, -0.432683, -0.231034, 0.183885, 0.286685, 0.021052, -0.217804, 0.280132], [0.621413, -0.413862, 0.182835, 0.119338, 0.678199, 0.070137, -0.719663, 0.555615], [0.584221, -0.348351, -0.020313, 0.316174, 0.446985, 0.262904, -0.441814, 0.507748], [-0.194856, 0.477087, -0.168366, 0.520619, 0.188851, -0.166965, 0.250392, 0.016796], [0.183862, -0.522047, -0.458858, 0.390867, 0.256126, 0.400147, -0.064308, 0.256242], [-0.507932, 0.030895, 0.270009, -0.564018, -0.233496, 0.418267, 0.476701, -0.150277], [0.333962, 0.013909, -0.257583, 0.371891, 0.327481, 0.580096, -0.690953, 0.343973], [-0.304267, 0.039045, 0.041729, 0.326938, -0.262094, -0.34206, 0.677786, -0.1206]], "network.4.bias": [-0.175937, 0.335751, 0.09528, -0.007821, -0.300697, -0.322923, -0.063228, 0.333089], "network.6.weight": [[0.388631, 0.408027, 0.476126, -0.263616, 0.439984, 0.452692, 0.380785, -0.173827], [-0.67909, 0.025495, -0.308372, 0.41467, -0.263956, -0.134328, 0.460982, -0.016058], [0.236222, 0.086339, -0.514445, 0.135388, 0.376333, -0.144392, -0.058987, 0.276316], [0.192111, 0.404943, 0.388523, -0.283497, 0.44946, 0.30245, 0.373748, -0.58941], [-0.262548, -0.271472, 0.258139, 0.411953, -0.44835, -0.205389, -0.021779, 0.426184], [0.303931, 0.571954, 0.104637, 0.24125, 0.321002, -0.139709, 0.599548, -0.426724], [0.399428, 0.704407, 0.581976, 0.200714, 0.063915, 0.327546, 0.248157, -0.552288], [-0.168508, -0.194677, 0.112312, 0.534702, -0.363437, -0.032599, -0.108664, 0.218058]], "network.6.bias": [-0.118916, 0.062382, -0.173926, -0.008134, 0.161229, -0.012763, -0.286789, 0.266245], "network.8.weight": [[-0.306822, 0.379969, -0.228077, -0.098203, 0.112407, 0.314673, -0.372148, 0.07193], [0.448289, -0.17286, -0.090741, 0.484953, -0.282052, 0.629839, 0.223094, -0.013346], [0.203256, -0.078996, -0.398064, 0.574195, -0.245486, 0.31903, 0.389317, -0.019726], [0.200943, 0.296101, -0.202987, -0.234971, -0.009452, -0.346187, -0.232034, -0.233087], [-0.235726, 0.172794, 0.207257, -0.090288, 0.17948, -0.15368, 0.219656, 0.063816], [0.035801, 0.405714, 0.135062, -0.455672, 0.53782, 0.166163, -0.142441, 0.416131], [-0.021739, -0.076163, -0.433016, -0.262143, 0.153542, 0.214167, 0.147421, -0.32794], [0.567273, -0.111246, -0.278879, 0.370108, -0.208263, 0.550177, 0.564333, -0.49347]], "network.8.bias": [0.353161, -0.270149, -0.289938, -0.551146, 0.035301, 0.008448, -0.149594, -0.199292], "network.10.weight": [[0.404954, -0.362393, -0.302683, -0.382822, 0.005542, 0.216584, -0.171292, -0.52494]], "network.10.bias": [0.052675]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6899049580097198, "train_acc": 0.575, "val_loss": 0.7012660503387451, "val_acc": 0.48}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6806736290454865, "train_acc": 0.575, "val_loss": 0.700400710105896, "val_acc": 0.48}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6725892722606659, "train_acc": 0.575, "val_loss": 0.6845899224281311, "val_acc": 0.48}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6531522572040558, "train_acc": 0.575, "val_loss": 0.6391638517379761, "val_acc": 0.48}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6190845370292664, "train_acc": 0.505, "val_loss": 0.5519809722900391, "val_acc": 0.48}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.5295256972312927, "train_acc": 0.505, "val_loss": 0.5019288063049316, "val_acc": 0.66}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.47860434651374817, "train_acc": 0.735, "val_loss": 0.43960997462272644, "val_acc": 0.86}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.4334201514720917, "train_acc": 0.845, "val_loss": 0.41446903347969055, "val_acc": 0.88}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.4304188936948776, "train_acc": 0.825, "val_loss": 0.37776076793670654, "val_acc": 0.9}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.3840505927801132, "train_acc": 0.84, "val_loss": 0.3570537567138672, "val_acc": 0.86}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.3916149288415909, "train_acc": 0.815, "val_loss": 0.45099860429763794, "val_acc": 0.88}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.33847689628601074, "train_acc": 0.855, "val_loss": 0.5892689228057861, "val_acc": 0.86}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.31833498179912567, "train_acc": 0.885, "val_loss": 0.4399909973144531, "val_acc": 0.84}], "summary": {"total_epochs": 13, "degraded_epochs": 4, "improved_epochs": 9, "patterns": ["mountain_pattern"], "degraded_stage": {"initial_val_loss": 0.7012660503387451, "final_val_loss": 0.6391638517379761, "initial_val_acc": 0.48, "final_val_acc": 0.48, "best_val_acc": 0.48}, "improved_stage": {"initial_val_loss": 0.5519809722900391, "final_val_loss": 0.4399909973144531, "initial_val_acc": 0.48, "final_val_acc": 0.84, "best_val_acc": 0.9, "best_epoch": 8}, "improvement": 0.42000000000000004, "first_improvement_epoch": 3}}
|
33
|
{"target_pattern": "sorted_descending", "degraded_accuracy": 0.7, "improved_accuracy": 0.96, "improvement": 0.26, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 4009, "learning_rate": 0.03724346377980019, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_descending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_descending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.265516,
-0.251026,
0.591522,
0.411411,
-0.381153
],
[
-0.230099,
0.189619,
-0.258673,
0.590785,
0.32705
],
[
0.666292,
0.500006,
0.086597,
0.140946,
0.187282
],
[
-0.105441,
0.282246,
-0.274844,
0.418087,
-0.397628
],
[
-0.325493,
0.069967,
-0.204435,
0.16534,
0.548061
],
[
-0.11604,
-0.1578,
-0.234313,
0.51293,
0.475933
],
[
-0.27479,
0.086757,
0.041152,
0.21967,
0.196326
]
],
"network.0.bias": [
0.075653,
-0.217711,
-0.019446,
0.181512,
0.152932,
0.650914,
0.195403
],
"network.2.weight": [
[
0.159774,
0.462489,
0.076334,
0.524481,
0.253766,
0.595683,
0.197468
],
[
-0.49354,
0.017089,
0.233845,
-0.316023,
-0.529781,
-0.104762,
-0.07234
],
[
-0.628809,
-0.516755,
0.606679,
-0.260667,
-0.630689,
0.075265,
-0.48454
],
[
-0.394371,
-0.568008,
0.341356,
0.119003,
-0.172176,
-0.001783,
-0.405153
],
[
-0.444031,
-0.118832,
0.183647,
-0.328296,
-0.280839,
-0.06874,
-0.250224
],
[
-0.389503,
-0.246133,
0.178063,
-0.178089,
-0.5481,
0.32012,
-0.212198
],
[
-0.355378,
0.462871,
-0.211298,
-0.39201,
0.404906,
0.3922,
0.12885
]
],
"network.2.bias": [
-0.028124,
0.094352,
0.118168,
0.09198,
0.23641,
0.046103,
-0.084716
],
"network.4.weight": [
[
0.354956,
-0.746559,
-0.24958,
-0.332173,
-0.441841,
-0.022791,
0.471889
],
[
0.534974,
-0.18634,
-0.541537,
-0.401454,
-0.623906,
-0.348281,
-0.073389
],
[
0.085176,
0.392552,
0.420367,
0.480942,
0.275009,
0.025829,
-0.332727
],
[
-0.007756,
0.624196,
0.651207,
0.212747,
0.049853,
0.44937,
-0.471581
],
[
-0.510675,
-0.220319,
-0.069283,
0.050746,
-0.34672,
-0.625091,
-0.196477
],
[
-0.509277,
0.457127,
0.328525,
0.44396,
0.295101,
0.207922,
-0.503157
],
[
0.494916,
-0.15495,
-0.336597,
-0.278155,
-0.376717,
-0.361613,
0.488466
]
],
"network.4.bias": [
0.120487,
0.318466,
-0.190632,
0.244386,
0.06001,
0.137462,
0.321207
],
"network.6.weight": [
[
0.358296,
0.453884,
-0.47014,
-0.325377,
0.034836,
-0.115605,
0.564629
],
[
0.438086,
0.527412,
0.009512,
-0.050605,
-0.102548,
-0.302054,
-0.036
],
[
-0.418451,
-0.256212,
0.417945,
-0.269812,
-0.387364,
0.425916,
-0.54477
],
[
-0.225265,
-0.102377,
0.409736,
0.308999,
-0.417785,
0.636338,
-0.290823
],
[
0.078484,
-0.117658,
0.497875,
0.211993,
-0.217177,
0.297019,
-0.245775
],
[
-0.540989,
-0.128017,
0.279072,
0.010935,
-0.102646,
0.071808,
-0.079165
],
[
-0.059844,
-0.145632,
-0.125123,
0.447241,
-0.051407,
0.345338,
-0.004885
]
],
"network.6.bias": [
0.722445,
0.557162,
0.110096,
-0.178726,
-0.07356,
0.122806,
-0.087469
],
"network.8.weight": [
[
-0.535573,
-0.167816,
0.445393,
0.051149,
0.376993,
-0.338467,
0.46761
],
[
0.284867,
0.155133,
-0.097266,
-0.372328,
0.100223,
-0.390502,
-0.451716
],
[
-0.414514,
0.246132,
0.162425,
0.589335,
0.226675,
-0.237977,
0.371099
],
[
0.260362,
0.320811,
-0.164521,
-0.608109,
-0.218114,
0.37919,
-0.501664
],
[
0.390821,
0.085478,
-0.400139,
-0.672522,
-0.488872,
0.211066,
-0.29751
],
[
-0.497998,
0.183483,
-0.083893,
0.339389,
0.311146,
-0.163372,
0.6359
],
[
0.654099,
0.512046,
-0.244674,
-0.401317,
-0.205349,
-0.393317,
-0.064121
]
],
"network.8.bias": [
0.281592,
0.208932,
-0.14494,
0.088763,
0.239738,
0.273857,
0.676144
],
"network.10.weight": [
[
0.190184,
-0.307855,
0.460063,
-0.202227,
-0.29024,
0.166397,
-0.675879
]
],
"network.10.bias": [
-0.075633
]
}
## Activation Signature
### 0
mean: [0.935592, 0.971396, 2.434963, 0.396650, 0.478831, 1.416440, 0.810700]
std: [1.417157, 1.661508, 2.079038, 1.455937, 1.255840, 1.511105, 0.815982]
### 2
mean: [2.325854, -0.558408, -0.465682, -0.463247, -0.578508, -0.357576, 0.219837]
std: [2.178687, 1.198368, 2.354214, 1.550133, 1.204710, 0.964289, 1.628741]
### 4
mean: [0.730470, 0.823736, 0.353193, 0.628620, -1.403041, -0.793033, 1.275538]
std: [1.979956, 2.176963, 1.317147, 1.786598, 1.254797, 2.350510, 2.191594]
### 6
mean: [2.045947, 1.512281, -1.325387, -0.218413, 0.061402, -0.581565, 0.120861]
std: [2.801705, 1.428258, 2.133960, 2.282349, 1.542332, 1.276999, 1.113173]
### 8
mean: [-0.840460, 0.693073, -0.132016, 0.525295, 0.456900, -0.093766, 2.527068]
std: [2.235301, 1.770471, 1.973820, 2.406542, 2.817413, 2.031541, 3.068752]
### 10
mean: [-2.582264]
std: [3.253570]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_descending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.265516,
-0.251026,
0.591522,
0.411411,
-0.381153
],
[
-0.230099,
0.189619,
-0.258673,
0.590785,
0.32705
],
[
0.666292,
0.500006,
0.086597,
0.140946,
0.187282
],
[
-0.105441,
0.282246,
-0.274844,
0.418087,
-0.397628
],
[
-0.325493,
0.069967,
-0.204435,
0.16534,
0.548061
],
[
-0.11604,
-0.1578,
-0.234313,
0.51293,
0.475933
],
[
-0.27479,
0.086757,
0.041152,
0.21967,
0.196326
]
],
"network.0.bias": [
0.075653,
-0.217711,
-0.019446,
0.181512,
0.152932,
0.650914,
0.195403
],
"network.2.weight": [
[
0.159774,
0.462489,
0.076334,
0.524481,
0.253766,
0.595683,
0.197468
],
[
-0.49354,
0.017089,
0.233845,
-0.316023,
-0.529781,
-0.104762,
-0.07234
],
[
-0.628809,
-0.516755,
0.606679,
-0.260667,
-0.630689,
0.075265,
-0.48454
],
[
-0.394371,
-0.568008,
0.341356,
0.119003,
-0.172176,
-0.001783,
-0.405153
],
[
-0.444031,
-0.118832,
0.183647,
-0.328296,
-0.280839,
-0.06874,
-0.250224
],
[
-0.389503,
-0.246133,
0.178063,
-0.178089,
-0.5481,
0.32012,
-0.212198
],
[
-0.355378,
0.462871,
-0.211298,
-0.39201,
0.404906,
0.3922,
0.12885
]
],
"network.2.bias": [
-0.028124,
0.094352,
0.118168,
0.09198,
0.23641,
0.046103,
-0.084716
],
"network.4.weight": [
[
0.354956,
-0.746559,
-0.24958,
-0.332173,
-0.441841,
-0.022791,
0.471889
],
[
0.534974,
-0.18634,
-0.541537,
-0.401454,
-0.623906,
-0.348281,
-0.073389
],
[
0.085176,
0.392552,
0.420367,
0.480942,
0.275009,
0.025829,
-0.332727
],
[
-0.007756,
0.624196,
0.651207,
0.212747,
0.049853,
0.44937,
-0.471581
],
[
-0.510675,
-0.220319,
-0.069283,
0.050746,
-0.34672,
-0.625091,
-0.196477
],
[
-0.509277,
0.457127,
0.328525,
0.44396,
0.295101,
0.207922,
-0.503157
],
[
0.494916,
-0.15495,
-0.336597,
-0.278155,
-0.376717,
-0.361613,
0.488466
]
],
"network.4.bias": [
0.120487,
0.318466,
-0.190632,
0.244386,
0.06001,
0.137462,
0.321207
],
"network.6.weight": [
[
0.358296,
0.453884,
-0.47014,
-0.325377,
0.034836,
-0.115605,
0.564629
],
[
0.438086,
0.527412,
0.009512,
-0.050605,
-0.102548,
-0.302054,
-0.036
],
[
-0.418451,
-0.256212,
0.417945,
-0.269812,
-0.387364,
0.425916,
-0.54477
],
[
-0.225265,
-0.102377,
0.409736,
0.308999,
-0.417785,
0.636338,
-0.290823
],
[
0.078484,
-0.117658,
0.497875,
0.211993,
-0.217177,
0.297019,
-0.245775
],
[
-0.540989,
-0.128017,
0.279072,
0.010935,
-0.102646,
0.071808,
-0.079165
],
[
-0.059844,
-0.145632,
-0.125123,
0.447241,
-0.051407,
0.345338,
-0.004885
]
],
"network.6.bias": [
0.722445,
0.557162,
0.110096,
-0.178726,
-0.07356,
0.122806,
-0.087469
],
"network.8.weight": [
[
-0.535573,
-0.167816,
0.445393,
0.051149,
0.376993,
-0.338467,
0.46761
],
[
0.284867,
0.155133,
-0.097266,
-0.372328,
0.100223,
-0.390502,
-0.451716
],
[
-0.414514,
0.246132,
0.162425,
0.589335,
0.226675,
-0.237977,
0.371099
],
[
0.260362,
0.320811,
-0.164521,
-0.608109,
-0.218114,
0.37919,
-0.501664
],
[
0.390821,
0.085478,
-0.400139,
-0.672522,
-0.488872,
0.211066,
-0.29751
],
[
-0.497998,
0.183483,
-0.083893,
0.339389,
0.311146,
-0.163372,
0.6359
],
[
0.654099,
0.512046,
-0.244674,
-0.401317,
-0.205349,
-0.393317,
-0.064121
]
],
"network.8.bias": [
0.281592,
0.208932,
-0.14494,
0.088763,
0.239738,
0.273857,
0.676144
],
"network.10.weight": [
[
0.190184,
-0.307855,
0.460063,
-0.202227,
-0.29024,
0.166397,
-0.675879
]
],
"network.10.bias": [
-0.075633
]
}
## Activation Signature
### 0
mean: [0.935592, 0.971396, 2.434963, 0.396650, 0.478831, 1.416440, 0.810700]
std: [1.417157, 1.661508, 2.079038, 1.455937, 1.255840, 1.511105, 0.815982]
### 2
mean: [2.325854, -0.558408, -0.465682, -0.463247, -0.578508, -0.357576, 0.219837]
std: [2.178687, 1.198368, 2.354214, 1.550133, 1.204710, 0.964289, 1.628741]
### 4
mean: [0.730470, 0.823736, 0.353193, 0.628620, -1.403041, -0.793033, 1.275538]
std: [1.979956, 2.176963, 1.317147, 1.786598, 1.254797, 2.350510, 2.191594]
### 6
mean: [2.045947, 1.512281, -1.325387, -0.218413, 0.061402, -0.581565, 0.120861]
std: [2.801705, 1.428258, 2.133960, 2.282349, 1.542332, 1.276999, 1.113173]
### 8
mean: [-0.840460, 0.693073, -0.132016, 0.525295, 0.456900, -0.093766, 2.527068]
std: [2.235301, 1.770471, 1.973820, 2.406542, 2.817413, 2.031541, 3.068752]
### 10
mean: [-2.582264]
std: [3.253570]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_descending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.9355916976928711, "std": 1.4171569347381592}, "1": {"mean": 0.9713957905769348, "std": 1.6615076065063477}, "2": {"mean": 2.4349629878997803, "std": 2.07903790473938}, "3": {"mean": 0.396650105714798, "std": 1.455937385559082}, "4": {"mean": 0.4788305461406708, "std": 1.2558401823043823}, "5": {"mean": 1.4164397716522217, "std": 1.511104941368103}, "6": {"mean": 0.8106998801231384, "std": 0.815982460975647}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 2.3258540630340576, "std": 2.178687334060669}, "1": {"mean": -0.5584079027175903, "std": 1.198367953300476}, "2": {"mean": -0.4656819999217987, "std": 2.3542139530181885}, "3": {"mean": -0.4632473886013031, "std": 1.5501325130462646}, "4": {"mean": -0.5785080194473267, "std": 1.2047103643417358}, "5": {"mean": -0.35757583379745483, "std": 0.9642889499664307}, "6": {"mean": 0.21983709931373596, "std": 1.628740906715393}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.7304703593254089, "std": 1.9799556732177734}, "1": {"mean": 0.8237364888191223, "std": 2.1769626140594482}, "2": {"mean": 0.35319337248802185, "std": 1.317147135734558}, "3": {"mean": 0.6286200881004333, "std": 1.7865984439849854}, "4": {"mean": -1.4030407667160034, "std": 1.254797339439392}, "5": {"mean": -0.7930333614349365, "std": 2.3505098819732666}, "6": {"mean": 1.2755380868911743, "std": 2.191594123840332}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 2.0459468364715576, "std": 2.8017051219940186}, "1": {"mean": 1.5122814178466797, "std": 1.4282578229904175}, "2": {"mean": -1.3253873586654663, "std": 2.133960485458374}, "3": {"mean": -0.21841266751289368, "std": 2.2823493480682373}, "4": {"mean": 0.06140189617872238, "std": 1.5423320531845093}, "5": {"mean": -0.5815646052360535, "std": 1.2769992351531982}, "6": {"mean": 0.12086126208305359, "std": 1.113173484802246}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.8404595255851746, "std": 2.2353007793426514}, "1": {"mean": 0.6930732131004333, "std": 1.7704710960388184}, "2": {"mean": -0.132016122341156, "std": 1.9738203287124634}, "3": {"mean": 0.5252954363822937, "std": 2.4065420627593994}, "4": {"mean": 0.4568999111652374, "std": 2.8174126148223877}, "5": {"mean": -0.09376601874828339, "std": 2.031541347503662}, "6": {"mean": 2.5270683765411377, "std": 3.0687522888183594}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -2.582263946533203, "std": 3.253570318222046}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.265516, -0.251026, 0.591522, 0.411411, -0.381153], [-0.230099, 0.189619, -0.258673, 0.590785, 0.32705], [0.666292, 0.500006, 0.086597, 0.140946, 0.187282], [-0.105441, 0.282246, -0.274844, 0.418087, -0.397628], [-0.325493, 0.069967, -0.204435, 0.16534, 0.548061], [-0.11604, -0.1578, -0.234313, 0.51293, 0.475933], [-0.27479, 0.086757, 0.041152, 0.21967, 0.196326]], "network.0.bias": [0.075653, -0.217711, -0.019446, 0.181512, 0.152932, 0.650914, 0.195403], "network.2.weight": [[0.159774, 0.462489, 0.076334, 0.524481, 0.253766, 0.595683, 0.197468], [-0.49354, 0.017089, 0.233845, -0.316023, -0.529781, -0.104762, -0.07234], [-0.628809, -0.516755, 0.606679, -0.260667, -0.630689, 0.075265, -0.48454], [-0.394371, -0.568008, 0.341356, 0.119003, -0.172176, -0.001783, -0.405153], [-0.444031, -0.118832, 0.183647, -0.328296, -0.280839, -0.06874, -0.250224], [-0.389503, -0.246133, 0.178063, -0.178089, -0.5481, 0.32012, -0.212198], [-0.355378, 0.462871, -0.211298, -0.39201, 0.404906, 0.3922, 0.12885]], "network.2.bias": [-0.028124, 0.094352, 0.118168, 0.09198, 0.23641, 0.046103, -0.084716], "network.4.weight": [[0.354956, -0.746559, -0.24958, -0.332173, -0.441841, -0.022791, 0.471889], [0.534974, -0.18634, -0.541537, -0.401454, -0.623906, -0.348281, -0.073389], [0.085176, 0.392552, 0.420367, 0.480942, 0.275009, 0.025829, -0.332727], [-0.007756, 0.624196, 0.651207, 0.212747, 0.049853, 0.44937, -0.471581], [-0.510675, -0.220319, -0.069283, 0.050746, -0.34672, -0.625091, -0.196477], [-0.509277, 0.457127, 0.328525, 0.44396, 0.295101, 0.207922, -0.503157], [0.494916, -0.15495, -0.336597, -0.278155, -0.376717, -0.361613, 0.488466]], "network.4.bias": [0.120487, 0.318466, -0.190632, 0.244386, 0.06001, 0.137462, 0.321207], "network.6.weight": [[0.358296, 0.453884, -0.47014, -0.325377, 0.034836, -0.115605, 0.564629], [0.438086, 0.527412, 0.009512, -0.050605, -0.102548, -0.302054, -0.036], [-0.418451, -0.256212, 0.417945, -0.269812, -0.387364, 0.425916, -0.54477], [-0.225265, -0.102377, 0.409736, 0.308999, -0.417785, 0.636338, -0.290823], [0.078484, -0.117658, 0.497875, 0.211993, -0.217177, 0.297019, -0.245775], [-0.540989, -0.128017, 0.279072, 0.010935, -0.102646, 0.071808, -0.079165], [-0.059844, -0.145632, -0.125123, 0.447241, -0.051407, 0.345338, -0.004885]], "network.6.bias": [0.722445, 0.557162, 0.110096, -0.178726, -0.07356, 0.122806, -0.087469], "network.8.weight": [[-0.535573, -0.167816, 0.445393, 0.051149, 0.376993, -0.338467, 0.46761], [0.284867, 0.155133, -0.097266, -0.372328, 0.100223, -0.390502, -0.451716], [-0.414514, 0.246132, 0.162425, 0.589335, 0.226675, -0.237977, 0.371099], [0.260362, 0.320811, -0.164521, -0.608109, -0.218114, 0.37919, -0.501664], [0.390821, 0.085478, -0.400139, -0.672522, -0.488872, 0.211066, -0.29751], [-0.497998, 0.183483, -0.083893, 0.339389, 0.311146, -0.163372, 0.6359], [0.654099, 0.512046, -0.244674, -0.401317, -0.205349, -0.393317, -0.064121]], "network.8.bias": [0.281592, 0.208932, -0.14494, 0.088763, 0.239738, 0.273857, 0.676144], "network.10.weight": [[0.190184, -0.307855, 0.460063, -0.202227, -0.29024, 0.166397, -0.675879]], "network.10.bias": [-0.075633]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6906745433807373, "train_acc": 0.5, "val_loss": 0.6828798055648804, "val_acc": 0.56}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6782574653625488, "train_acc": 0.56, "val_loss": 0.6593039035797119, "val_acc": 0.56}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6486417055130005, "train_acc": 0.56, "val_loss": 0.5440099239349365, "val_acc": 0.7}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.5045328885316849, "train_acc": 0.75, "val_loss": 0.2884593904018402, "val_acc": 0.92}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.3236505091190338, "train_acc": 0.915, "val_loss": 0.400558203458786, "val_acc": 0.86}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.28192421793937683, "train_acc": 0.885, "val_loss": 0.22803367674350739, "val_acc": 0.96}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.19075839221477509, "train_acc": 0.935, "val_loss": 0.2595984935760498, "val_acc": 0.92}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.2221870869398117, "train_acc": 0.945, "val_loss": 0.24462385475635529, "val_acc": 0.92}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.20247000455856323, "train_acc": 0.945, "val_loss": 0.23470085859298706, "val_acc": 0.94}], "summary": {"total_epochs": 9, "degraded_epochs": 3, "improved_epochs": 6, "patterns": ["sorted_descending"], "degraded_stage": {"initial_val_loss": 0.6828798055648804, "final_val_loss": 0.5440099239349365, "initial_val_acc": 0.56, "final_val_acc": 0.7, "best_val_acc": 0.7}, "improved_stage": {"initial_val_loss": 0.2884593904018402, "final_val_loss": 0.23470085859298706, "initial_val_acc": 0.92, "final_val_acc": 0.94, "best_val_acc": 0.96, "best_epoch": 5}, "improvement": 0.26, "first_improvement_epoch": 2}}
|
34
|
{"target_pattern": "ends_with", "degraded_accuracy": 0.52, "improved_accuracy": 0.92, "improvement": 0.4, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 1496, "learning_rate": 0.08248469519525602, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "ends_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["ends_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.172558,
-0.393768,
0.16954,
-0.670044,
-0.159902
],
[
-0.582947,
-0.461321,
-0.140592,
-0.944818,
-0.23713
],
[
0.538047,
-0.073176,
-0.0543,
0.026635,
1.379443
],
[
-0.236945,
0.045112,
0.236162,
0.202574,
-1.267737
],
[
-0.857435,
-0.08587,
-0.085139,
-0.07874,
-0.075699
],
[
-1.072939,
-0.371618,
0.3829,
-0.122184,
1.282623
]
],
"network.0.bias": [
-0.520282,
-0.183062,
0.268763,
-0.111621,
-0.616283,
0.354256
],
"network.2.weight": [
[
-0.491955,
-0.014285,
-0.634974,
0.392384,
0.477619,
0.212324
],
[
-0.565781,
-0.172222,
-0.398829,
0.101024,
-0.084999,
-0.36588
],
[
0.4274,
0.056123,
0.100729,
0.392212,
0.336047,
-0.341969
],
[
0.043738,
0.231309,
-0.875787,
0.600323,
-0.278093,
-0.310817
],
[
-1.156222,
0.018019,
0.880662,
-0.636206,
-1.612169,
0.64484
],
[
-0.501313,
-0.621552,
-0.462155,
0.19086,
-0.510869,
0.318321
]
],
"network.2.bias": [
-0.343501,
-1.092845,
0.106887,
-0.352124,
-0.094861,
-0.340199
],
"network.4.weight": [
[
0.420891,
0.200379,
-0.378156,
-0.73271,
-0.44852,
-0.182258
],
[
0.95218,
0.045546,
-0.400845,
-0.140637,
1.277243,
-0.466264
],
[
0.675466,
-0.298044,
-0.948456,
0.580562,
0.963658,
0.500085
],
[
0.149332,
-0.643304,
0.097617,
-0.450007,
1.227875,
-0.504595
],
[
0.930094,
0.127754,
-0.331946,
-0.26436,
1.070316,
-0.406222
],
[
0.57529,
-0.185742,
-0.410697,
0.2419,
1.089686,
-0.693705
]
],
"network.4.bias": [
-1.084967,
-0.418399,
0.293631,
-0.439719,
-0.620617,
-0.203186
],
"network.6.weight": [
[
0.593172,
1.344579,
0.846065,
1.100545,
1.127519,
1.045398
],
[
-0.077916,
-0.968897,
-0.537332,
0.118387,
-0.531405,
-0.337064
],
[
1.15026,
-0.626236,
-0.408391,
0.095771,
0.003938,
0.468608
],
[
0.552991,
-0.511024,
-0.623849,
0.107578,
0.404442,
0.003073
],
[
-0.475303,
-0.286403,
-0.645817,
-0.565409,
-0.278928,
-0.883726
],
[
-0.561589,
-0.12457,
-0.461426,
0.338714,
-0.924942,
-0.383736
]
],
"network.6.bias": [
0.050592,
0.173371,
-0.565581,
-0.925104,
0.621163,
0.013134
],
"network.8.weight": [
[
-0.00685,
0.296331,
-0.016451,
-0.918714,
0.936447,
0.060373
],
[
0.106277,
0.609261,
-0.791997,
-0.452217,
0.565525,
0.442061
],
[
0.017653,
-0.441406,
0.28382,
0.082201,
-0.263333,
-0.054775
],
[
0.598353,
-0.421617,
-0.300519,
-0.236945,
-1.056547,
-0.175184
],
[
1.012005,
-0.245603,
-0.587147,
-0.366061,
-0.055756,
-0.796025
],
[
-0.314753,
-0.063898,
-0.384021,
-0.687804,
0.195639,
-0.043072
]
],
"network.8.bias": [
0.535546,
0.29865,
0.128262,
0.151355,
0.391125,
0.367544
],
"network.10.weight": [
[
0.53086,
0.770693,
0.276287,
-0.66344,
-0.747056,
0.389758
]
],
"network.10.bias": [
-0.282488
]
}
## Activation Signature
### 0
mean: [-2.303012, -4.361388, 2.449993, -0.960100, -2.279850, 0.461594]
std: [1.832399, 2.942598, 2.907425, 2.375199, 1.896866, 3.152984]
### 2
mean: [-1.492424, -2.475248, -0.027162, -2.667044, 2.866030, -0.890293]
std: [1.733971, 1.796273, 0.635187, 3.189844, 3.684893, 1.165924]
### 4
mean: [-2.459524, 3.246382, 2.973385, 3.218895, 2.440026, 2.954912]
std: [1.586718, 4.672238, 3.529307, 4.425474, 3.922505, 3.979883]
### 6
mean: [16.744711, -6.688774, -2.231497, -3.091830, -7.446169, -4.200553]
std: [22.329762, 9.140608, 1.971654, 2.518303, 10.620204, 5.740493]
### 8
mean: [0.727901, 2.386559, 0.298526, 10.007754, 17.491879, -4.819425]
std: [0.538591, 2.168593, 0.513034, 13.494316, 22.484991, 7.087096]
### 10
mean: [-17.830147]
std: [24.129391]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
ends_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.172558,
-0.393768,
0.16954,
-0.670044,
-0.159902
],
[
-0.582947,
-0.461321,
-0.140592,
-0.944818,
-0.23713
],
[
0.538047,
-0.073176,
-0.0543,
0.026635,
1.379443
],
[
-0.236945,
0.045112,
0.236162,
0.202574,
-1.267737
],
[
-0.857435,
-0.08587,
-0.085139,
-0.07874,
-0.075699
],
[
-1.072939,
-0.371618,
0.3829,
-0.122184,
1.282623
]
],
"network.0.bias": [
-0.520282,
-0.183062,
0.268763,
-0.111621,
-0.616283,
0.354256
],
"network.2.weight": [
[
-0.491955,
-0.014285,
-0.634974,
0.392384,
0.477619,
0.212324
],
[
-0.565781,
-0.172222,
-0.398829,
0.101024,
-0.084999,
-0.36588
],
[
0.4274,
0.056123,
0.100729,
0.392212,
0.336047,
-0.341969
],
[
0.043738,
0.231309,
-0.875787,
0.600323,
-0.278093,
-0.310817
],
[
-1.156222,
0.018019,
0.880662,
-0.636206,
-1.612169,
0.64484
],
[
-0.501313,
-0.621552,
-0.462155,
0.19086,
-0.510869,
0.318321
]
],
"network.2.bias": [
-0.343501,
-1.092845,
0.106887,
-0.352124,
-0.094861,
-0.340199
],
"network.4.weight": [
[
0.420891,
0.200379,
-0.378156,
-0.73271,
-0.44852,
-0.182258
],
[
0.95218,
0.045546,
-0.400845,
-0.140637,
1.277243,
-0.466264
],
[
0.675466,
-0.298044,
-0.948456,
0.580562,
0.963658,
0.500085
],
[
0.149332,
-0.643304,
0.097617,
-0.450007,
1.227875,
-0.504595
],
[
0.930094,
0.127754,
-0.331946,
-0.26436,
1.070316,
-0.406222
],
[
0.57529,
-0.185742,
-0.410697,
0.2419,
1.089686,
-0.693705
]
],
"network.4.bias": [
-1.084967,
-0.418399,
0.293631,
-0.439719,
-0.620617,
-0.203186
],
"network.6.weight": [
[
0.593172,
1.344579,
0.846065,
1.100545,
1.127519,
1.045398
],
[
-0.077916,
-0.968897,
-0.537332,
0.118387,
-0.531405,
-0.337064
],
[
1.15026,
-0.626236,
-0.408391,
0.095771,
0.003938,
0.468608
],
[
0.552991,
-0.511024,
-0.623849,
0.107578,
0.404442,
0.003073
],
[
-0.475303,
-0.286403,
-0.645817,
-0.565409,
-0.278928,
-0.883726
],
[
-0.561589,
-0.12457,
-0.461426,
0.338714,
-0.924942,
-0.383736
]
],
"network.6.bias": [
0.050592,
0.173371,
-0.565581,
-0.925104,
0.621163,
0.013134
],
"network.8.weight": [
[
-0.00685,
0.296331,
-0.016451,
-0.918714,
0.936447,
0.060373
],
[
0.106277,
0.609261,
-0.791997,
-0.452217,
0.565525,
0.442061
],
[
0.017653,
-0.441406,
0.28382,
0.082201,
-0.263333,
-0.054775
],
[
0.598353,
-0.421617,
-0.300519,
-0.236945,
-1.056547,
-0.175184
],
[
1.012005,
-0.245603,
-0.587147,
-0.366061,
-0.055756,
-0.796025
],
[
-0.314753,
-0.063898,
-0.384021,
-0.687804,
0.195639,
-0.043072
]
],
"network.8.bias": [
0.535546,
0.29865,
0.128262,
0.151355,
0.391125,
0.367544
],
"network.10.weight": [
[
0.53086,
0.770693,
0.276287,
-0.66344,
-0.747056,
0.389758
]
],
"network.10.bias": [
-0.282488
]
}
## Activation Signature
### 0
mean: [-2.303012, -4.361388, 2.449993, -0.960100, -2.279850, 0.461594]
std: [1.832399, 2.942598, 2.907425, 2.375199, 1.896866, 3.152984]
### 2
mean: [-1.492424, -2.475248, -0.027162, -2.667044, 2.866030, -0.890293]
std: [1.733971, 1.796273, 0.635187, 3.189844, 3.684893, 1.165924]
### 4
mean: [-2.459524, 3.246382, 2.973385, 3.218895, 2.440026, 2.954912]
std: [1.586718, 4.672238, 3.529307, 4.425474, 3.922505, 3.979883]
### 6
mean: [16.744711, -6.688774, -2.231497, -3.091830, -7.446169, -4.200553]
std: [22.329762, 9.140608, 1.971654, 2.518303, 10.620204, 5.740493]
### 8
mean: [0.727901, 2.386559, 0.298526, 10.007754, 17.491879, -4.819425]
std: [0.538591, 2.168593, 0.513034, 13.494316, 22.484991, 7.087096]
### 10
mean: [-17.830147]
std: [24.129391]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
ends_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -2.303011655807495, "std": 1.8323994874954224}, "1": {"mean": -4.361388206481934, "std": 2.9425981044769287}, "2": {"mean": 2.449993133544922, "std": 2.9074251651763916}, "3": {"mean": -0.9601002335548401, "std": 2.37519907951355}, "4": {"mean": -2.279850482940674, "std": 1.8968664407730103}, "5": {"mean": 0.46159353852272034, "std": 3.1529839038848877}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -1.4924238920211792, "std": 1.7339712381362915}, "1": {"mean": -2.475248336791992, "std": 1.796272873878479}, "2": {"mean": -0.027162471786141396, "std": 0.6351869702339172}, "3": {"mean": -2.667043924331665, "std": 3.1898441314697266}, "4": {"mean": 2.8660295009613037, "std": 3.6848926544189453}, "5": {"mean": -0.8902925252914429, "std": 1.1659239530563354}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -2.459524393081665, "std": 1.5867178440093994}, "1": {"mean": 3.2463815212249756, "std": 4.672237873077393}, "2": {"mean": 2.9733848571777344, "std": 3.5293068885803223}, "3": {"mean": 3.218895196914673, "std": 4.425474166870117}, "4": {"mean": 2.440025568008423, "std": 3.9225051403045654}, "5": {"mean": 2.954911947250366, "std": 3.9798834323883057}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 16.74471092224121, "std": 22.329761505126953}, "1": {"mean": -6.6887736320495605, "std": 9.140607833862305}, "2": {"mean": -2.231497287750244, "std": 1.9716540575027466}, "3": {"mean": -3.0918304920196533, "std": 2.518303394317627}, "4": {"mean": -7.446168899536133, "std": 10.620203971862793}, "5": {"mean": -4.200552940368652, "std": 5.740493297576904}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.7279006242752075, "std": 0.538591206073761}, "1": {"mean": 2.386559247970581, "std": 2.168593168258667}, "2": {"mean": 0.2985262870788574, "std": 0.5130338072776794}, "3": {"mean": 10.0077543258667, "std": 13.494316101074219}, "4": {"mean": 17.491878509521484, "std": 22.4849910736084}, "5": {"mean": -4.819424629211426, "std": 7.087095737457275}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -17.83014678955078, "std": 24.129390716552734}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.172558, -0.393768, 0.16954, -0.670044, -0.159902], [-0.582947, -0.461321, -0.140592, -0.944818, -0.23713], [0.538047, -0.073176, -0.0543, 0.026635, 1.379443], [-0.236945, 0.045112, 0.236162, 0.202574, -1.267737], [-0.857435, -0.08587, -0.085139, -0.07874, -0.075699], [-1.072939, -0.371618, 0.3829, -0.122184, 1.282623]], "network.0.bias": [-0.520282, -0.183062, 0.268763, -0.111621, -0.616283, 0.354256], "network.2.weight": [[-0.491955, -0.014285, -0.634974, 0.392384, 0.477619, 0.212324], [-0.565781, -0.172222, -0.398829, 0.101024, -0.084999, -0.36588], [0.4274, 0.056123, 0.100729, 0.392212, 0.336047, -0.341969], [0.043738, 0.231309, -0.875787, 0.600323, -0.278093, -0.310817], [-1.156222, 0.018019, 0.880662, -0.636206, -1.612169, 0.64484], [-0.501313, -0.621552, -0.462155, 0.19086, -0.510869, 0.318321]], "network.2.bias": [-0.343501, -1.092845, 0.106887, -0.352124, -0.094861, -0.340199], "network.4.weight": [[0.420891, 0.200379, -0.378156, -0.73271, -0.44852, -0.182258], [0.95218, 0.045546, -0.400845, -0.140637, 1.277243, -0.466264], [0.675466, -0.298044, -0.948456, 0.580562, 0.963658, 0.500085], [0.149332, -0.643304, 0.097617, -0.450007, 1.227875, -0.504595], [0.930094, 0.127754, -0.331946, -0.26436, 1.070316, -0.406222], [0.57529, -0.185742, -0.410697, 0.2419, 1.089686, -0.693705]], "network.4.bias": [-1.084967, -0.418399, 0.293631, -0.439719, -0.620617, -0.203186], "network.6.weight": [[0.593172, 1.344579, 0.846065, 1.100545, 1.127519, 1.045398], [-0.077916, -0.968897, -0.537332, 0.118387, -0.531405, -0.337064], [1.15026, -0.626236, -0.408391, 0.095771, 0.003938, 0.468608], [0.552991, -0.511024, -0.623849, 0.107578, 0.404442, 0.003073], [-0.475303, -0.286403, -0.645817, -0.565409, -0.278928, -0.883726], [-0.561589, -0.12457, -0.461426, 0.338714, -0.924942, -0.383736]], "network.6.bias": [0.050592, 0.173371, -0.565581, -0.925104, 0.621163, 0.013134], "network.8.weight": [[-0.00685, 0.296331, -0.016451, -0.918714, 0.936447, 0.060373], [0.106277, 0.609261, -0.791997, -0.452217, 0.565525, 0.442061], [0.017653, -0.441406, 0.28382, 0.082201, -0.263333, -0.054775], [0.598353, -0.421617, -0.300519, -0.236945, -1.056547, -0.175184], [1.012005, -0.245603, -0.587147, -0.366061, -0.055756, -0.796025], [-0.314753, -0.063898, -0.384021, -0.687804, 0.195639, -0.043072]], "network.8.bias": [0.535546, 0.29865, 0.128262, 0.151355, 0.391125, 0.367544], "network.10.weight": [[0.53086, 0.770693, 0.276287, -0.66344, -0.747056, 0.389758]], "network.10.bias": [-0.282488]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6892788410186768, "train_acc": 0.57, "val_loss": 0.6962285041809082, "val_acc": 0.52}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6852421462535858, "train_acc": 0.57, "val_loss": 0.6955626010894775, "val_acc": 0.52}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6803379058837891, "train_acc": 0.57, "val_loss": 0.6872557997703552, "val_acc": 0.52}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6748557984828949, "train_acc": 0.57, "val_loss": 0.6495596170425415, "val_acc": 0.52}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6210085451602936, "train_acc": 0.58, "val_loss": 0.4492313265800476, "val_acc": 0.86}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.42067068815231323, "train_acc": 0.885, "val_loss": 0.3129158318042755, "val_acc": 0.9}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.321799173951149, "train_acc": 0.87, "val_loss": 0.2667141258716583, "val_acc": 0.9}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.3082380145788193, "train_acc": 0.88, "val_loss": 0.2685103118419647, "val_acc": 0.9}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.26909536123275757, "train_acc": 0.895, "val_loss": 0.29915690422058105, "val_acc": 0.88}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.2795598953962326, "train_acc": 0.895, "val_loss": 0.24498285353183746, "val_acc": 0.92}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.23851098865270615, "train_acc": 0.91, "val_loss": 0.22456400096416473, "val_acc": 0.92}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.2676670700311661, "train_acc": 0.925, "val_loss": 0.2663586735725403, "val_acc": 0.9}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.22137834131717682, "train_acc": 0.915, "val_loss": 0.27590256929397583, "val_acc": 0.9}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.18966179341077805, "train_acc": 0.92, "val_loss": 0.30363526940345764, "val_acc": 0.86}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["ends_with"], "degraded_stage": {"initial_val_loss": 0.6962285041809082, "final_val_loss": 0.6495596170425415, "initial_val_acc": 0.52, "final_val_acc": 0.52, "best_val_acc": 0.52}, "improved_stage": {"initial_val_loss": 0.4492313265800476, "final_val_loss": 0.30363526940345764, "initial_val_acc": 0.86, "final_val_acc": 0.86, "best_val_acc": 0.92, "best_epoch": 9}, "improvement": 0.4, "first_improvement_epoch": 3}}
|
35
|
{"target_pattern": "sorted_ascending", "degraded_accuracy": 0.46, "improved_accuracy": 0.98, "improvement": 0.52, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 2876, "learning_rate": 0.052811634834035476, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_ascending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_ascending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.293957,
0.751197,
0.155091,
0.110613,
-0.281907
],
[
-0.121918,
-0.147412,
-0.546076,
-0.153244,
-0.091768
],
[
0.464614,
0.217352,
-0.004516,
0.479091,
-0.565967
],
[
0.685794,
0.391303,
0.331503,
0.016492,
-0.403268
],
[
0.876326,
0.298994,
0.171434,
-0.486244,
0.090966
],
[
-0.669019,
-0.696684,
0.177323,
-0.145452,
0.432468
],
[
0.247464,
0.879593,
0.373772,
-0.349634,
-0.028344
]
],
"network.0.bias": [
0.067375,
0.091704,
-0.199946,
0.425463,
-0.006249,
-0.209039,
0.259324
],
"network.2.weight": [
[
-0.377982,
-0.025704,
-0.177746,
-0.02856,
0.212395,
0.190507,
-0.201302
],
[
0.488727,
-0.007863,
-0.170999,
-0.146038,
0.113688,
0.130324,
0.400317
],
[
-0.053028,
0.212962,
0.101182,
-0.502821,
0.031466,
-0.155566,
-0.174462
],
[
0.677791,
0.223563,
0.775134,
0.403413,
0.429958,
-0.320664,
0.787702
],
[
0.645897,
0.146354,
0.435651,
0.70077,
0.248937,
-0.305156,
0.163061
],
[
-0.065289,
0.046065,
-0.12019,
-0.166056,
-0.056925,
-0.058714,
0.027132
],
[
0.02613,
-0.005774,
0.170898,
0.620484,
0.307845,
-0.354603,
-0.327728
]
],
"network.2.bias": [
-0.186208,
-0.332511,
0.611602,
-0.068345,
-0.082288,
-0.405808,
-0.060025
],
"network.4.weight": [
[
-0.01006,
0.002436,
-0.15847,
-0.23549,
-0.022066,
-0.057872,
0.287982
],
[
-0.025817,
0.254247,
-0.58863,
0.726676,
0.789576,
0.001359,
0.731225
],
[
-0.126373,
0.17886,
-0.599263,
-0.297082,
-0.205104,
0.285638,
-0.059515
],
[
0.323732,
-0.376714,
0.258848,
-0.205472,
-0.216419,
-0.300998,
-0.227097
],
[
0.096688,
0.085726,
-0.246416,
-0.268464,
0.000836,
-0.138296,
0.20427
],
[
0.175262,
0.542516,
-0.133508,
0.778774,
0.588514,
0.186808,
0.470582
],
[
0.209656,
0.210947,
-0.266779,
-0.394939,
0.225638,
-0.236906,
0.343024
]
],
"network.4.bias": [
-0.288078,
0.225886,
0.133029,
0.608527,
-0.223439,
-0.045258,
-0.259341
],
"network.6.weight": [
[
-0.28311,
-0.308842,
0.065241,
-0.237455,
0.233273,
-0.087141,
-0.141126
],
[
-0.09862,
0.630595,
0.183273,
-0.586888,
-0.272758,
0.558897,
-0.11282
],
[
0.264945,
-0.295561,
0.215315,
0.247603,
0.252366,
-0.000802,
-0.242294
],
[
-0.341388,
-0.305927,
-0.063808,
-0.133405,
0.26349,
0.107999,
0.193302
],
[
-0.160036,
0.54131,
-0.272234,
-0.617584,
-0.270728,
0.517291,
0.161651
],
[
0.129348,
-0.072679,
-0.175462,
0.029714,
0.170854,
-0.426397,
-0.044792
],
[
0.267925,
-0.028163,
0.066317,
0.517977,
-0.126984,
0.051313,
0.232007
]
],
"network.6.bias": [
-0.277503,
0.036762,
-0.224187,
0.070972,
-0.008434,
-0.145991,
0.259427
],
"network.8.weight": [
[
0.157753,
0.019106,
-0.172591,
-0.343512,
-0.354167,
-0.302738,
-0.366114
],
[
0.064436,
0.455441,
-0.172439,
0.479267,
0.453384,
-0.233308,
-0.470556
],
[
-0.056656,
0.245494,
0.122483,
-0.169559,
0.665564,
0.148605,
-0.117582
],
[
0.1277,
-0.149622,
-0.101589,
0.35908,
0.067988,
0.292648,
-0.119805
],
[
-0.123474,
0.596303,
-0.102917,
-0.089882,
0.74823,
0.098166,
-0.493511
],
[
0.230749,
0.529255,
0.247602,
-0.356943,
0.117455,
-0.158529,
0.262073
],
[
-0.177572,
-0.360046,
-0.482513,
0.132688,
0.034365,
0.269814,
0.18434
]
],
"network.8.bias": [
-0.056123,
0.093033,
-0.04727,
-0.337924,
0.039365,
-0.323842,
0.660033
],
"network.10.weight": [
[
0.084112,
-0.267199,
0.070083,
0.119223,
-0.666303,
-0.485452,
0.429952
]
],
"network.10.bias": [
0.24088
]
}
## Activation Signature
### 0
mean: [2.015522, -1.911130, 1.094230, 2.222410, 1.056698, -1.719238, 2.166134]
std: [1.878587, 1.376321, 1.702029, 2.072032, 2.371176, 2.210410, 2.116090]
### 2
mean: [-1.381667, 1.174547, -0.872617, 5.604393, 4.096475, -1.106861, 1.297022]
std: [1.109578, 1.399796, 1.239967, 5.017456, 3.722349, 0.619239, 1.526565]
### 4
mean: [-1.323925, 8.811250, -2.298674, -2.187534, -1.368923, 8.054935, -0.848068]
std: [0.885017, 7.885977, 2.048301, 2.615736, 0.949383, 7.365237, 0.455756]
### 6
mean: [-3.727395, 10.033167, -2.808063, -1.769304, 8.862510, -4.219262, 0.481589]
std: [3.049099, 9.152546, 2.365416, 1.602905, 8.146167, 3.714930, 0.139228]
### 8
mean: [-3.187558, 8.473804, 8.278381, -1.295616, 12.445229, 6.166686, -2.565346]
std: [2.736433, 7.794961, 7.634427, 0.825587, 11.472922, 5.811053, 2.990808]
### 10
mean: [-12.691760]
std: [12.052455]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_ascending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.293957,
0.751197,
0.155091,
0.110613,
-0.281907
],
[
-0.121918,
-0.147412,
-0.546076,
-0.153244,
-0.091768
],
[
0.464614,
0.217352,
-0.004516,
0.479091,
-0.565967
],
[
0.685794,
0.391303,
0.331503,
0.016492,
-0.403268
],
[
0.876326,
0.298994,
0.171434,
-0.486244,
0.090966
],
[
-0.669019,
-0.696684,
0.177323,
-0.145452,
0.432468
],
[
0.247464,
0.879593,
0.373772,
-0.349634,
-0.028344
]
],
"network.0.bias": [
0.067375,
0.091704,
-0.199946,
0.425463,
-0.006249,
-0.209039,
0.259324
],
"network.2.weight": [
[
-0.377982,
-0.025704,
-0.177746,
-0.02856,
0.212395,
0.190507,
-0.201302
],
[
0.488727,
-0.007863,
-0.170999,
-0.146038,
0.113688,
0.130324,
0.400317
],
[
-0.053028,
0.212962,
0.101182,
-0.502821,
0.031466,
-0.155566,
-0.174462
],
[
0.677791,
0.223563,
0.775134,
0.403413,
0.429958,
-0.320664,
0.787702
],
[
0.645897,
0.146354,
0.435651,
0.70077,
0.248937,
-0.305156,
0.163061
],
[
-0.065289,
0.046065,
-0.12019,
-0.166056,
-0.056925,
-0.058714,
0.027132
],
[
0.02613,
-0.005774,
0.170898,
0.620484,
0.307845,
-0.354603,
-0.327728
]
],
"network.2.bias": [
-0.186208,
-0.332511,
0.611602,
-0.068345,
-0.082288,
-0.405808,
-0.060025
],
"network.4.weight": [
[
-0.01006,
0.002436,
-0.15847,
-0.23549,
-0.022066,
-0.057872,
0.287982
],
[
-0.025817,
0.254247,
-0.58863,
0.726676,
0.789576,
0.001359,
0.731225
],
[
-0.126373,
0.17886,
-0.599263,
-0.297082,
-0.205104,
0.285638,
-0.059515
],
[
0.323732,
-0.376714,
0.258848,
-0.205472,
-0.216419,
-0.300998,
-0.227097
],
[
0.096688,
0.085726,
-0.246416,
-0.268464,
0.000836,
-0.138296,
0.20427
],
[
0.175262,
0.542516,
-0.133508,
0.778774,
0.588514,
0.186808,
0.470582
],
[
0.209656,
0.210947,
-0.266779,
-0.394939,
0.225638,
-0.236906,
0.343024
]
],
"network.4.bias": [
-0.288078,
0.225886,
0.133029,
0.608527,
-0.223439,
-0.045258,
-0.259341
],
"network.6.weight": [
[
-0.28311,
-0.308842,
0.065241,
-0.237455,
0.233273,
-0.087141,
-0.141126
],
[
-0.09862,
0.630595,
0.183273,
-0.586888,
-0.272758,
0.558897,
-0.11282
],
[
0.264945,
-0.295561,
0.215315,
0.247603,
0.252366,
-0.000802,
-0.242294
],
[
-0.341388,
-0.305927,
-0.063808,
-0.133405,
0.26349,
0.107999,
0.193302
],
[
-0.160036,
0.54131,
-0.272234,
-0.617584,
-0.270728,
0.517291,
0.161651
],
[
0.129348,
-0.072679,
-0.175462,
0.029714,
0.170854,
-0.426397,
-0.044792
],
[
0.267925,
-0.028163,
0.066317,
0.517977,
-0.126984,
0.051313,
0.232007
]
],
"network.6.bias": [
-0.277503,
0.036762,
-0.224187,
0.070972,
-0.008434,
-0.145991,
0.259427
],
"network.8.weight": [
[
0.157753,
0.019106,
-0.172591,
-0.343512,
-0.354167,
-0.302738,
-0.366114
],
[
0.064436,
0.455441,
-0.172439,
0.479267,
0.453384,
-0.233308,
-0.470556
],
[
-0.056656,
0.245494,
0.122483,
-0.169559,
0.665564,
0.148605,
-0.117582
],
[
0.1277,
-0.149622,
-0.101589,
0.35908,
0.067988,
0.292648,
-0.119805
],
[
-0.123474,
0.596303,
-0.102917,
-0.089882,
0.74823,
0.098166,
-0.493511
],
[
0.230749,
0.529255,
0.247602,
-0.356943,
0.117455,
-0.158529,
0.262073
],
[
-0.177572,
-0.360046,
-0.482513,
0.132688,
0.034365,
0.269814,
0.18434
]
],
"network.8.bias": [
-0.056123,
0.093033,
-0.04727,
-0.337924,
0.039365,
-0.323842,
0.660033
],
"network.10.weight": [
[
0.084112,
-0.267199,
0.070083,
0.119223,
-0.666303,
-0.485452,
0.429952
]
],
"network.10.bias": [
0.24088
]
}
## Activation Signature
### 0
mean: [2.015522, -1.911130, 1.094230, 2.222410, 1.056698, -1.719238, 2.166134]
std: [1.878587, 1.376321, 1.702029, 2.072032, 2.371176, 2.210410, 2.116090]
### 2
mean: [-1.381667, 1.174547, -0.872617, 5.604393, 4.096475, -1.106861, 1.297022]
std: [1.109578, 1.399796, 1.239967, 5.017456, 3.722349, 0.619239, 1.526565]
### 4
mean: [-1.323925, 8.811250, -2.298674, -2.187534, -1.368923, 8.054935, -0.848068]
std: [0.885017, 7.885977, 2.048301, 2.615736, 0.949383, 7.365237, 0.455756]
### 6
mean: [-3.727395, 10.033167, -2.808063, -1.769304, 8.862510, -4.219262, 0.481589]
std: [3.049099, 9.152546, 2.365416, 1.602905, 8.146167, 3.714930, 0.139228]
### 8
mean: [-3.187558, 8.473804, 8.278381, -1.295616, 12.445229, 6.166686, -2.565346]
std: [2.736433, 7.794961, 7.634427, 0.825587, 11.472922, 5.811053, 2.990808]
### 10
mean: [-12.691760]
std: [12.052455]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_ascending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 2.015522003173828, "std": 1.8785865306854248}, "1": {"mean": -1.9111297130584717, "std": 1.376320719718933}, "2": {"mean": 1.0942295789718628, "std": 1.7020294666290283}, "3": {"mean": 2.222410202026367, "std": 2.0720317363739014}, "4": {"mean": 1.0566983222961426, "std": 2.371176242828369}, "5": {"mean": -1.71923828125, "std": 2.210409641265869}, "6": {"mean": 2.1661343574523926, "std": 2.1160898208618164}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -1.3816665410995483, "std": 1.1095781326293945}, "1": {"mean": 1.1745474338531494, "std": 1.3997962474822998}, "2": {"mean": -0.8726174235343933, "std": 1.2399674654006958}, "3": {"mean": 5.6043925285339355, "std": 5.0174560546875}, "4": {"mean": 4.096474647521973, "std": 3.722348690032959}, "5": {"mean": -1.1068609952926636, "std": 0.619238555431366}, "6": {"mean": 1.297021508216858, "std": 1.526565432548523}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -1.323925495147705, "std": 0.8850172758102417}, "1": {"mean": 8.811249732971191, "std": 7.885976791381836}, "2": {"mean": -2.2986738681793213, "std": 2.0483005046844482}, "3": {"mean": -2.1875338554382324, "std": 2.6157360076904297}, "4": {"mean": -1.3689229488372803, "std": 0.9493831992149353}, "5": {"mean": 8.054935455322266, "std": 7.365236759185791}, "6": {"mean": -0.848067581653595, "std": 0.45575565099716187}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -3.7273952960968018, "std": 3.0490992069244385}, "1": {"mean": 10.033166885375977, "std": 9.152545928955078}, "2": {"mean": -2.8080625534057617, "std": 2.3654160499572754}, "3": {"mean": -1.7693043947219849, "std": 1.602905035018921}, "4": {"mean": 8.862509727478027, "std": 8.146166801452637}, "5": {"mean": -4.21926212310791, "std": 3.714930295944214}, "6": {"mean": 0.481588751077652, "std": 0.13922812044620514}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -3.18755841255188, "std": 2.7364327907562256}, "1": {"mean": 8.473804473876953, "std": 7.794961452484131}, "2": {"mean": 8.27838134765625, "std": 7.634426593780518}, "3": {"mean": -1.2956163883209229, "std": 0.82558673620224}, "4": {"mean": 12.445228576660156, "std": 11.472922325134277}, "5": {"mean": 6.166685581207275, "std": 5.8110527992248535}, "6": {"mean": -2.5653464794158936, "std": 2.9908077716827393}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -12.691760063171387, "std": 12.052454948425293}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.293957, 0.751197, 0.155091, 0.110613, -0.281907], [-0.121918, -0.147412, -0.546076, -0.153244, -0.091768], [0.464614, 0.217352, -0.004516, 0.479091, -0.565967], [0.685794, 0.391303, 0.331503, 0.016492, -0.403268], [0.876326, 0.298994, 0.171434, -0.486244, 0.090966], [-0.669019, -0.696684, 0.177323, -0.145452, 0.432468], [0.247464, 0.879593, 0.373772, -0.349634, -0.028344]], "network.0.bias": [0.067375, 0.091704, -0.199946, 0.425463, -0.006249, -0.209039, 0.259324], "network.2.weight": [[-0.377982, -0.025704, -0.177746, -0.02856, 0.212395, 0.190507, -0.201302], [0.488727, -0.007863, -0.170999, -0.146038, 0.113688, 0.130324, 0.400317], [-0.053028, 0.212962, 0.101182, -0.502821, 0.031466, -0.155566, -0.174462], [0.677791, 0.223563, 0.775134, 0.403413, 0.429958, -0.320664, 0.787702], [0.645897, 0.146354, 0.435651, 0.70077, 0.248937, -0.305156, 0.163061], [-0.065289, 0.046065, -0.12019, -0.166056, -0.056925, -0.058714, 0.027132], [0.02613, -0.005774, 0.170898, 0.620484, 0.307845, -0.354603, -0.327728]], "network.2.bias": [-0.186208, -0.332511, 0.611602, -0.068345, -0.082288, -0.405808, -0.060025], "network.4.weight": [[-0.01006, 0.002436, -0.15847, -0.23549, -0.022066, -0.057872, 0.287982], [-0.025817, 0.254247, -0.58863, 0.726676, 0.789576, 0.001359, 0.731225], [-0.126373, 0.17886, -0.599263, -0.297082, -0.205104, 0.285638, -0.059515], [0.323732, -0.376714, 0.258848, -0.205472, -0.216419, -0.300998, -0.227097], [0.096688, 0.085726, -0.246416, -0.268464, 0.000836, -0.138296, 0.20427], [0.175262, 0.542516, -0.133508, 0.778774, 0.588514, 0.186808, 0.470582], [0.209656, 0.210947, -0.266779, -0.394939, 0.225638, -0.236906, 0.343024]], "network.4.bias": [-0.288078, 0.225886, 0.133029, 0.608527, -0.223439, -0.045258, -0.259341], "network.6.weight": [[-0.28311, -0.308842, 0.065241, -0.237455, 0.233273, -0.087141, -0.141126], [-0.09862, 0.630595, 0.183273, -0.586888, -0.272758, 0.558897, -0.11282], [0.264945, -0.295561, 0.215315, 0.247603, 0.252366, -0.000802, -0.242294], [-0.341388, -0.305927, -0.063808, -0.133405, 0.26349, 0.107999, 0.193302], [-0.160036, 0.54131, -0.272234, -0.617584, -0.270728, 0.517291, 0.161651], [0.129348, -0.072679, -0.175462, 0.029714, 0.170854, -0.426397, -0.044792], [0.267925, -0.028163, 0.066317, 0.517977, -0.126984, 0.051313, 0.232007]], "network.6.bias": [-0.277503, 0.036762, -0.224187, 0.070972, -0.008434, -0.145991, 0.259427], "network.8.weight": [[0.157753, 0.019106, -0.172591, -0.343512, -0.354167, -0.302738, -0.366114], [0.064436, 0.455441, -0.172439, 0.479267, 0.453384, -0.233308, -0.470556], [-0.056656, 0.245494, 0.122483, -0.169559, 0.665564, 0.148605, -0.117582], [0.1277, -0.149622, -0.101589, 0.35908, 0.067988, 0.292648, -0.119805], [-0.123474, 0.596303, -0.102917, -0.089882, 0.74823, 0.098166, -0.493511], [0.230749, 0.529255, 0.247602, -0.356943, 0.117455, -0.158529, 0.262073], [-0.177572, -0.360046, -0.482513, 0.132688, 0.034365, 0.269814, 0.18434]], "network.8.bias": [-0.056123, 0.093033, -0.04727, -0.337924, 0.039365, -0.323842, 0.660033], "network.10.weight": [[0.084112, -0.267199, 0.070083, 0.119223, -0.666303, -0.485452, 0.429952]], "network.10.bias": [0.24088]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6671582460403442, "train_acc": 0.585, "val_loss": 0.6770175099372864, "val_acc": 0.46}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6013821363449097, "train_acc": 0.585, "val_loss": 0.551065981388092, "val_acc": 0.46}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.49208444356918335, "train_acc": 0.51, "val_loss": 0.4083539545536041, "val_acc": 0.94}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.4000173956155777, "train_acc": 0.905, "val_loss": 0.37783196568489075, "val_acc": 0.92}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.350574791431427, "train_acc": 0.91, "val_loss": 0.27547329664230347, "val_acc": 0.98}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.2935333847999573, "train_acc": 0.93, "val_loss": 0.22231626510620117, "val_acc": 0.98}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.25869259238243103, "train_acc": 0.92, "val_loss": 0.16551196575164795, "val_acc": 0.98}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.23964575678110123, "train_acc": 0.925, "val_loss": 0.13962379097938538, "val_acc": 0.98}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.21058984100818634, "train_acc": 0.935, "val_loss": 0.11332917958498001, "val_acc": 0.98}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.17540495470166206, "train_acc": 0.935, "val_loss": 0.10355223715305328, "val_acc": 0.98}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.20527224242687225, "train_acc": 0.925, "val_loss": 0.11139790713787079, "val_acc": 0.96}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.2486671432852745, "train_acc": 0.915, "val_loss": 0.11411451548337936, "val_acc": 0.96}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["sorted_ascending"], "degraded_stage": {"initial_val_loss": 0.6770175099372864, "final_val_loss": 0.551065981388092, "initial_val_acc": 0.46, "final_val_acc": 0.46, "best_val_acc": 0.46}, "improved_stage": {"initial_val_loss": 0.4083539545536041, "final_val_loss": 0.11411451548337936, "initial_val_acc": 0.94, "final_val_acc": 0.96, "best_val_acc": 0.98, "best_epoch": 4}, "improvement": 0.52, "first_improvement_epoch": 1}}
|
36
|
{"target_pattern": "no_repeats", "degraded_accuracy": 0.52, "improved_accuracy": 0.68, "improvement": 0.16000000000000003, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 2412, "learning_rate": 0.09693395560140622, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "no_repeats", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["no_repeats"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.252401,
-0.248687,
-0.149186,
0.006539,
-0.578357
],
[
0.73124,
-0.102077,
0.239694,
-0.021685,
-0.259149
],
[
-0.112944,
-0.185213,
-0.162788,
-0.077759,
-0.695573
],
[
-0.567462,
0.615763,
0.181296,
-0.120657,
0.317522
],
[
0.731601,
0.393961,
0.1475,
-0.15566,
-0.159497
],
[
0.714808,
0.361145,
-0.515292,
-0.423103,
0.124191
],
[
0.878412,
-0.33602,
0.03126,
-0.017267,
0.259369
],
[
0.216212,
0.358027,
0.00459,
0.202531,
0.167805
]
],
"network.0.bias": [
-0.352262,
0.366851,
0.387144,
0.27491,
-0.237045,
-0.276568,
0.296823,
-0.110348
],
"network.2.weight": [
[
-0.338816,
0.226847,
-0.133606,
-0.087986,
0.697702,
-0.005615,
0.149728,
-0.212827
],
[
-0.762011,
0.585822,
-0.815336,
-0.322254,
0.790399,
0.297567,
0.609121,
-0.007088
],
[
0.380223,
-0.006973,
0.261481,
0.345909,
-0.356501,
-0.347782,
0.03531,
0.494152
],
[
0.196152,
0.017501,
0.121579,
0.282666,
0.107888,
-0.27656,
-0.079849,
0.50422
],
[
0.745262,
0.104349,
0.283742,
0.265892,
-0.119988,
-0.498249,
-0.227066,
0.453573
],
[
-0.102905,
0.242628,
-0.436877,
-0.252386,
0.504891,
0.249361,
0.015002,
-0.231925
],
[
0.584296,
-0.592602,
0.231351,
0.139633,
-0.102906,
0.421518,
-0.334045,
0.238632
],
[
-0.180295,
0.446738,
-0.360407,
-0.227397,
0.640578,
-0.170263,
0.400281,
-0.094916
]
],
"network.2.bias": [
0.121848,
0.116168,
0.067696,
0.295025,
-0.234735,
-0.263433,
0.050858,
0.349489
],
"network.4.weight": [
[
-0.206915,
-0.07838,
-0.225636,
-0.218509,
-0.426436,
0.051434,
0.023395,
-0.508128
],
[
-0.214284,
0.063117,
-0.357699,
-0.417542,
-0.3301,
-0.203183,
0.04329,
-0.010166
],
[
0.45134,
0.561681,
-0.3503,
0.253526,
-0.27781,
0.391439,
-0.114486,
0.577144
],
[
-0.370702,
-0.255998,
0.086446,
0.063667,
0.146597,
-0.178105,
0.057517,
0.054133
],
[
-0.346515,
-0.093403,
-0.397055,
-0.058612,
-0.134842,
-0.265715,
0.50938,
0.141292
],
[
0.210287,
0.624039,
-0.345729,
-0.550756,
-0.29178,
0.420901,
-0.25198,
0.352073
],
[
0.298243,
0.001572,
0.016593,
-0.110832,
-0.352589,
-0.358586,
0.171695,
-0.058883
],
[
0.087465,
0.229488,
-0.39942,
-0.364051,
-0.780524,
0.115284,
-0.179296,
-0.049899
]
],
"network.4.bias": [
-0.131289,
-0.037544,
-0.091386,
-0.075594,
-0.297089,
0.360641,
-0.202711,
-0.679205
],
"network.6.weight": [
[
0.016392,
0.011516,
0.129013,
-0.160799,
-0.370771,
0.019178,
0.248682,
-0.321511
],
[
0.147043,
0.477674,
0.179977,
0.066087,
0.605821,
0.696463,
0.198497,
0.323302
],
[
0.445832,
0.82537,
0.569897,
0.129591,
0.11375,
0.322042,
0.048878,
0.054476
],
[
-0.170425,
-0.267141,
-0.219112,
-0.083997,
-0.258578,
-0.369324,
-0.284231,
0.318186
],
[
0.080728,
0.187962,
0.196625,
0.430651,
0.270673,
0.200097,
0.281389,
0.047069
],
[
0.270393,
0.406996,
0.516395,
0.064567,
0.077777,
0.481041,
-0.340508,
0.166409
],
[
0.489217,
0.607486,
0.620493,
0.160783,
0.368488,
0.073163,
-0.404301,
0.33685
],
[
0.050986,
0.236479,
0.216777,
-0.34052,
0.070492,
-0.340703,
0.203722,
0.433648
]
],
"network.6.bias": [
0.384344,
0.047929,
-0.030242,
-0.04922,
0.221563,
0.155406,
0.047994,
-0.323208
],
"network.8.weight": [
[
0.256797,
-0.470413,
-0.458319,
0.019746,
-0.229355,
-0.567173,
-0.160549,
-0.263583
]
],
"network.8.bias": [
0.237117
]
}
## Activation Signature
### 0
mean: [-1.502244, 1.225406, -1.455579, 1.202367, 1.168871, -0.559483, 1.125814, 1.461985]
std: [1.195491, 1.600841, 1.569439, 1.335513, 1.951725, 1.878195, 1.857705, 1.165472]
### 2
mean: [1.069976, 2.362995, 0.580474, 1.263531, 0.144699, 0.180680, -0.574143, 1.716637]
std: [1.704843, 3.636577, 1.002133, 0.830157, 0.960881, 1.498553, 1.234792, 2.322545]
### 4
mean: [-1.890719, -1.058651, 2.838875, -0.892600, -1.112273, 1.879465, -0.385682, -1.022015]
std: [1.698917, 0.786309, 4.821810, 1.710030, 0.843601, 4.211280, 0.222372, 1.632416]
### 6
mean: [0.748866, 2.074207, 2.240836, -1.365365, 1.185825, 2.765942, 1.995420, -0.458910]
std: [0.492979, 3.787373, 4.010388, 2.291436, 1.736842, 4.433328, 3.480493, 0.122196]
### 8
mean: [-3.738432]
std: [6.944249]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
no_repeats
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.252401,
-0.248687,
-0.149186,
0.006539,
-0.578357
],
[
0.73124,
-0.102077,
0.239694,
-0.021685,
-0.259149
],
[
-0.112944,
-0.185213,
-0.162788,
-0.077759,
-0.695573
],
[
-0.567462,
0.615763,
0.181296,
-0.120657,
0.317522
],
[
0.731601,
0.393961,
0.1475,
-0.15566,
-0.159497
],
[
0.714808,
0.361145,
-0.515292,
-0.423103,
0.124191
],
[
0.878412,
-0.33602,
0.03126,
-0.017267,
0.259369
],
[
0.216212,
0.358027,
0.00459,
0.202531,
0.167805
]
],
"network.0.bias": [
-0.352262,
0.366851,
0.387144,
0.27491,
-0.237045,
-0.276568,
0.296823,
-0.110348
],
"network.2.weight": [
[
-0.338816,
0.226847,
-0.133606,
-0.087986,
0.697702,
-0.005615,
0.149728,
-0.212827
],
[
-0.762011,
0.585822,
-0.815336,
-0.322254,
0.790399,
0.297567,
0.609121,
-0.007088
],
[
0.380223,
-0.006973,
0.261481,
0.345909,
-0.356501,
-0.347782,
0.03531,
0.494152
],
[
0.196152,
0.017501,
0.121579,
0.282666,
0.107888,
-0.27656,
-0.079849,
0.50422
],
[
0.745262,
0.104349,
0.283742,
0.265892,
-0.119988,
-0.498249,
-0.227066,
0.453573
],
[
-0.102905,
0.242628,
-0.436877,
-0.252386,
0.504891,
0.249361,
0.015002,
-0.231925
],
[
0.584296,
-0.592602,
0.231351,
0.139633,
-0.102906,
0.421518,
-0.334045,
0.238632
],
[
-0.180295,
0.446738,
-0.360407,
-0.227397,
0.640578,
-0.170263,
0.400281,
-0.094916
]
],
"network.2.bias": [
0.121848,
0.116168,
0.067696,
0.295025,
-0.234735,
-0.263433,
0.050858,
0.349489
],
"network.4.weight": [
[
-0.206915,
-0.07838,
-0.225636,
-0.218509,
-0.426436,
0.051434,
0.023395,
-0.508128
],
[
-0.214284,
0.063117,
-0.357699,
-0.417542,
-0.3301,
-0.203183,
0.04329,
-0.010166
],
[
0.45134,
0.561681,
-0.3503,
0.253526,
-0.27781,
0.391439,
-0.114486,
0.577144
],
[
-0.370702,
-0.255998,
0.086446,
0.063667,
0.146597,
-0.178105,
0.057517,
0.054133
],
[
-0.346515,
-0.093403,
-0.397055,
-0.058612,
-0.134842,
-0.265715,
0.50938,
0.141292
],
[
0.210287,
0.624039,
-0.345729,
-0.550756,
-0.29178,
0.420901,
-0.25198,
0.352073
],
[
0.298243,
0.001572,
0.016593,
-0.110832,
-0.352589,
-0.358586,
0.171695,
-0.058883
],
[
0.087465,
0.229488,
-0.39942,
-0.364051,
-0.780524,
0.115284,
-0.179296,
-0.049899
]
],
"network.4.bias": [
-0.131289,
-0.037544,
-0.091386,
-0.075594,
-0.297089,
0.360641,
-0.202711,
-0.679205
],
"network.6.weight": [
[
0.016392,
0.011516,
0.129013,
-0.160799,
-0.370771,
0.019178,
0.248682,
-0.321511
],
[
0.147043,
0.477674,
0.179977,
0.066087,
0.605821,
0.696463,
0.198497,
0.323302
],
[
0.445832,
0.82537,
0.569897,
0.129591,
0.11375,
0.322042,
0.048878,
0.054476
],
[
-0.170425,
-0.267141,
-0.219112,
-0.083997,
-0.258578,
-0.369324,
-0.284231,
0.318186
],
[
0.080728,
0.187962,
0.196625,
0.430651,
0.270673,
0.200097,
0.281389,
0.047069
],
[
0.270393,
0.406996,
0.516395,
0.064567,
0.077777,
0.481041,
-0.340508,
0.166409
],
[
0.489217,
0.607486,
0.620493,
0.160783,
0.368488,
0.073163,
-0.404301,
0.33685
],
[
0.050986,
0.236479,
0.216777,
-0.34052,
0.070492,
-0.340703,
0.203722,
0.433648
]
],
"network.6.bias": [
0.384344,
0.047929,
-0.030242,
-0.04922,
0.221563,
0.155406,
0.047994,
-0.323208
],
"network.8.weight": [
[
0.256797,
-0.470413,
-0.458319,
0.019746,
-0.229355,
-0.567173,
-0.160549,
-0.263583
]
],
"network.8.bias": [
0.237117
]
}
## Activation Signature
### 0
mean: [-1.502244, 1.225406, -1.455579, 1.202367, 1.168871, -0.559483, 1.125814, 1.461985]
std: [1.195491, 1.600841, 1.569439, 1.335513, 1.951725, 1.878195, 1.857705, 1.165472]
### 2
mean: [1.069976, 2.362995, 0.580474, 1.263531, 0.144699, 0.180680, -0.574143, 1.716637]
std: [1.704843, 3.636577, 1.002133, 0.830157, 0.960881, 1.498553, 1.234792, 2.322545]
### 4
mean: [-1.890719, -1.058651, 2.838875, -0.892600, -1.112273, 1.879465, -0.385682, -1.022015]
std: [1.698917, 0.786309, 4.821810, 1.710030, 0.843601, 4.211280, 0.222372, 1.632416]
### 6
mean: [0.748866, 2.074207, 2.240836, -1.365365, 1.185825, 2.765942, 1.995420, -0.458910]
std: [0.492979, 3.787373, 4.010388, 2.291436, 1.736842, 4.433328, 3.480493, 0.122196]
### 8
mean: [-3.738432]
std: [6.944249]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
no_repeats
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.5022437572479248, "std": 1.195490837097168}, "1": {"mean": 1.2254060506820679, "std": 1.6008412837982178}, "2": {"mean": -1.4555788040161133, "std": 1.5694392919540405}, "3": {"mean": 1.2023670673370361, "std": 1.3355134725570679}, "4": {"mean": 1.1688710451126099, "std": 1.9517253637313843}, "5": {"mean": -0.5594826340675354, "std": 1.8781946897506714}, "6": {"mean": 1.1258140802383423, "std": 1.8577053546905518}, "7": {"mean": 1.4619851112365723, "std": 1.1654716730117798}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 1.0699758529663086, "std": 1.7048429250717163}, "1": {"mean": 2.362995147705078, "std": 3.6365768909454346}, "2": {"mean": 0.5804739594459534, "std": 1.0021330118179321}, "3": {"mean": 1.2635310888290405, "std": 0.8301572799682617}, "4": {"mean": 0.14469929039478302, "std": 0.9608812928199768}, "5": {"mean": 0.18068017065525055, "std": 1.4985525608062744}, "6": {"mean": -0.5741431713104248, "std": 1.2347919940948486}, "7": {"mean": 1.716637372970581, "std": 2.322545051574707}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -1.890718698501587, "std": 1.6989171504974365}, "1": {"mean": -1.0586514472961426, "std": 0.7863085269927979}, "2": {"mean": 2.838874578475952, "std": 4.821809768676758}, "3": {"mean": -0.8926001787185669, "std": 1.7100303173065186}, "4": {"mean": -1.1122726202011108, "std": 0.843601405620575}, "5": {"mean": 1.8794645071029663, "std": 4.211280345916748}, "6": {"mean": -0.3856823742389679, "std": 0.22237151861190796}, "7": {"mean": -1.0220147371292114, "std": 1.6324164867401123}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.748866081237793, "std": 0.49297863245010376}, "1": {"mean": 2.074207305908203, "std": 3.7873733043670654}, "2": {"mean": 2.2408361434936523, "std": 4.010387897491455}, "3": {"mean": -1.3653647899627686, "std": 2.291436195373535}, "4": {"mean": 1.1858253479003906, "std": 1.736842155456543}, "5": {"mean": 2.765941858291626, "std": 4.433327674865723}, "6": {"mean": 1.9954203367233276, "std": 3.4804933071136475}, "7": {"mean": -0.4589100778102875, "std": 0.12219574302434921}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -3.738431692123413, "std": 6.944248676300049}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.252401, -0.248687, -0.149186, 0.006539, -0.578357], [0.73124, -0.102077, 0.239694, -0.021685, -0.259149], [-0.112944, -0.185213, -0.162788, -0.077759, -0.695573], [-0.567462, 0.615763, 0.181296, -0.120657, 0.317522], [0.731601, 0.393961, 0.1475, -0.15566, -0.159497], [0.714808, 0.361145, -0.515292, -0.423103, 0.124191], [0.878412, -0.33602, 0.03126, -0.017267, 0.259369], [0.216212, 0.358027, 0.00459, 0.202531, 0.167805]], "network.0.bias": [-0.352262, 0.366851, 0.387144, 0.27491, -0.237045, -0.276568, 0.296823, -0.110348], "network.2.weight": [[-0.338816, 0.226847, -0.133606, -0.087986, 0.697702, -0.005615, 0.149728, -0.212827], [-0.762011, 0.585822, -0.815336, -0.322254, 0.790399, 0.297567, 0.609121, -0.007088], [0.380223, -0.006973, 0.261481, 0.345909, -0.356501, -0.347782, 0.03531, 0.494152], [0.196152, 0.017501, 0.121579, 0.282666, 0.107888, -0.27656, -0.079849, 0.50422], [0.745262, 0.104349, 0.283742, 0.265892, -0.119988, -0.498249, -0.227066, 0.453573], [-0.102905, 0.242628, -0.436877, -0.252386, 0.504891, 0.249361, 0.015002, -0.231925], [0.584296, -0.592602, 0.231351, 0.139633, -0.102906, 0.421518, -0.334045, 0.238632], [-0.180295, 0.446738, -0.360407, -0.227397, 0.640578, -0.170263, 0.400281, -0.094916]], "network.2.bias": [0.121848, 0.116168, 0.067696, 0.295025, -0.234735, -0.263433, 0.050858, 0.349489], "network.4.weight": [[-0.206915, -0.07838, -0.225636, -0.218509, -0.426436, 0.051434, 0.023395, -0.508128], [-0.214284, 0.063117, -0.357699, -0.417542, -0.3301, -0.203183, 0.04329, -0.010166], [0.45134, 0.561681, -0.3503, 0.253526, -0.27781, 0.391439, -0.114486, 0.577144], [-0.370702, -0.255998, 0.086446, 0.063667, 0.146597, -0.178105, 0.057517, 0.054133], [-0.346515, -0.093403, -0.397055, -0.058612, -0.134842, -0.265715, 0.50938, 0.141292], [0.210287, 0.624039, -0.345729, -0.550756, -0.29178, 0.420901, -0.25198, 0.352073], [0.298243, 0.001572, 0.016593, -0.110832, -0.352589, -0.358586, 0.171695, -0.058883], [0.087465, 0.229488, -0.39942, -0.364051, -0.780524, 0.115284, -0.179296, -0.049899]], "network.4.bias": [-0.131289, -0.037544, -0.091386, -0.075594, -0.297089, 0.360641, -0.202711, -0.679205], "network.6.weight": [[0.016392, 0.011516, 0.129013, -0.160799, -0.370771, 0.019178, 0.248682, -0.321511], [0.147043, 0.477674, 0.179977, 0.066087, 0.605821, 0.696463, 0.198497, 0.323302], [0.445832, 0.82537, 0.569897, 0.129591, 0.11375, 0.322042, 0.048878, 0.054476], [-0.170425, -0.267141, -0.219112, -0.083997, -0.258578, -0.369324, -0.284231, 0.318186], [0.080728, 0.187962, 0.196625, 0.430651, 0.270673, 0.200097, 0.281389, 0.047069], [0.270393, 0.406996, 0.516395, 0.064567, 0.077777, 0.481041, -0.340508, 0.166409], [0.489217, 0.607486, 0.620493, 0.160783, 0.368488, 0.073163, -0.404301, 0.33685], [0.050986, 0.236479, 0.216777, -0.34052, 0.070492, -0.340703, 0.203722, 0.433648]], "network.6.bias": [0.384344, 0.047929, -0.030242, -0.04922, 0.221563, 0.155406, 0.047994, -0.323208], "network.8.weight": [[0.256797, -0.470413, -0.458319, 0.019746, -0.229355, -0.567173, -0.160549, -0.263583]], "network.8.bias": [0.237117]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7066080868244171, "train_acc": 0.44, "val_loss": 0.6552404761314392, "val_acc": 0.52}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6120957136154175, "train_acc": 0.575, "val_loss": 0.5087748169898987, "val_acc": 0.52}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 1.393336832523346, "train_acc": 0.495, "val_loss": 0.6027008891105652, "val_acc": 0.68}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.5874984860420227, "train_acc": 0.67, "val_loss": 0.6222074627876282, "val_acc": 0.6}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.5721495151519775, "train_acc": 0.655, "val_loss": 0.6242575645446777, "val_acc": 0.56}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.5804044008255005, "train_acc": 0.67, "val_loss": 0.5946443676948547, "val_acc": 0.64}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.5616225898265839, "train_acc": 0.685, "val_loss": 0.572813093662262, "val_acc": 0.66}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.519382894039154, "train_acc": 0.705, "val_loss": 0.6416094899177551, "val_acc": 0.68}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.5588286817073822, "train_acc": 0.705, "val_loss": 0.5593544244766235, "val_acc": 0.66}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.5465043783187866, "train_acc": 0.685, "val_loss": 0.5889398455619812, "val_acc": 0.64}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.5582871437072754, "train_acc": 0.68, "val_loss": 0.5621782541275024, "val_acc": 0.66}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.5436156094074249, "train_acc": 0.7, "val_loss": 0.5861780643463135, "val_acc": 0.66}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["no_repeats"], "degraded_stage": {"initial_val_loss": 0.6552404761314392, "final_val_loss": 0.5087748169898987, "initial_val_acc": 0.52, "final_val_acc": 0.52, "best_val_acc": 0.52}, "improved_stage": {"initial_val_loss": 0.6027008891105652, "final_val_loss": 0.5861780643463135, "initial_val_acc": 0.68, "final_val_acc": 0.66, "best_val_acc": 0.68, "best_epoch": 2}, "improvement": 0.16000000000000003, "first_improvement_epoch": 1}}
|
37
|
{"target_pattern": "sorted_descending", "degraded_accuracy": 0.52, "improved_accuracy": 0.94, "improvement": 0.41999999999999993, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 2079, "learning_rate": 0.05135048742655233, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_descending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_descending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.854196,
-0.092639,
0.41049,
0.089286,
0.423838
],
[
-0.47722,
-0.228046,
0.392427,
-0.041102,
0.489083
],
[
-0.843243,
-0.130948,
0.486644,
0.497521,
-0.118015
],
[
-0.030515,
-0.298131,
-0.323495,
-0.48272,
0.999118
],
[
-0.337109,
-0.279193,
-0.290638,
-0.134689,
-0.338993
],
[
-0.085518,
0.15028,
-0.383101,
-0.460849,
-0.198546
],
[
-0.971222,
0.204007,
-0.159559,
0.219691,
0.335826
],
[
-0.020021,
-0.346354,
-0.683696,
0.231517,
0.307916
]
],
"network.0.bias": [
-0.276487,
-0.287632,
-0.023811,
0.065037,
-0.862644,
-0.393232,
0.018189,
-0.539606
],
"network.2.weight": [
[
1.044215,
0.196973,
0.660006,
0.666602,
0.15796,
0.014607,
0.997225,
0.456547
],
[
0.098738,
-0.159046,
-0.488649,
-0.707291,
-0.379429,
0.08679,
-0.226309,
-0.136925
],
[
-0.2073,
-0.241257,
-0.015129,
-0.105918,
-0.108384,
0.116129,
-0.476648,
0.036019
],
[
0.271984,
-0.160026,
0.357395,
-0.598103,
0.074268,
-0.299346,
-0.504545,
-0.34915
],
[
-0.04874,
-0.198476,
-0.181148,
0.249901,
0.128984,
-0.347937,
-0.14755,
0.09861
],
[
0.672113,
0.047022,
0.344212,
0.880358,
0.319213,
0.34713,
0.813753,
0.131276
],
[
-0.149367,
-0.314878,
-0.396369,
-0.217782,
-0.180023,
0.300973,
-0.004586,
-0.189692
],
[
-0.197519,
-0.32383,
-0.244718,
-0.194571,
-0.233323,
0.018656,
0.042823,
-0.003039
]
],
"network.2.bias": [
0.284603,
-0.206121,
-0.226513,
0.018308,
-0.166292,
-0.19994,
-0.422867,
-0.007465
],
"network.4.weight": [
[
0.587369,
0.099362,
0.150626,
0.056267,
-0.337414,
0.743496,
0.231243,
0.178079
],
[
-0.468213,
-0.530976,
-0.334353,
-0.556466,
0.018112,
0.122297,
-0.177667,
0.195894
],
[
-0.551423,
-0.071425,
0.416458,
0.317907,
-0.48414,
-0.193266,
0.132666,
0.018164
],
[
-0.359705,
-0.15427,
-0.384486,
0.039401,
-0.123384,
-0.264528,
-0.271941,
0.110259
],
[
-0.236862,
-0.213636,
-0.154691,
-0.17941,
0.17335,
0.224302,
-0.110065,
0.142574
],
[
-0.137197,
0.305694,
0.17183,
-0.701624,
-0.392817,
-0.279896,
-0.425736,
-0.239293
],
[
0.891422,
-0.054906,
0.022878,
0.646148,
0.538924,
0.65865,
-0.053226,
0.340254
],
[
-0.427094,
0.070633,
-0.024134,
-0.033463,
0.002282,
-0.27249,
-0.31324,
-0.213376
]
],
"network.4.bias": [
-0.344562,
0.08793,
0.008729,
-0.112378,
-0.254725,
0.52753,
-0.10029,
-0.330214
],
"network.6.weight": [
[
0.06138,
0.182144,
-0.155629,
0.110237,
-0.028076,
0.016976,
-0.039606,
-0.236001
],
[
0.809013,
0.039365,
0.211982,
-0.000801,
-0.124845,
-0.298362,
0.761867,
-0.116867
],
[
0.105737,
-0.089551,
0.220108,
0.276074,
0.045164,
-0.234272,
-0.292398,
-0.047492
],
[
0.076747,
0.025539,
-0.464997,
-0.463146,
-0.278464,
0.120344,
-0.413611,
-0.028017
],
[
-0.20721,
-0.291787,
-0.051211,
-0.019659,
-0.117227,
0.528096,
-0.210836,
-0.207246
],
[
0.188564,
-0.325872,
-0.232224,
-0.26002,
0.235042,
0.018246,
-0.280914,
-0.051083
],
[
-0.353675,
0.142282,
0.065089,
-0.049748,
-0.03652,
-0.038194,
-0.361745,
-0.000234
],
[
-0.441632,
0.146891,
0.525922,
0.477024,
-0.329687,
0.287178,
-0.110529,
0.401029
]
],
"network.6.bias": [
-0.387132,
0.29662,
-0.324888,
-0.102165,
0.245739,
-0.026586,
-0.310219,
0.147414
],
"network.8.weight": [
[
-0.095993,
0.684941,
-0.046161,
0.146892,
-0.403152,
0.098618,
0.272946,
-0.390233
],
[
-0.299456,
-0.213574,
-0.35091,
-0.009799,
-0.168921,
-0.252661,
0.128925,
-0.158387
],
[
0.034153,
-0.20834,
0.346516,
0.283829,
-0.279771,
-0.299774,
0.286764,
0.277514
],
[
0.034666,
-0.124179,
0.27405,
-0.128117,
0.066239,
-0.520536,
-0.03925,
-0.494994
],
[
0.05659,
-0.291505,
-0.321619,
0.130038,
0.138658,
0.093079,
0.037364,
-0.371194
],
[
0.271057,
-0.332378,
0.217979,
-0.054817,
0.093147,
0.260026,
0.349412,
0.064331
],
[
-0.08696,
0.770487,
0.32419,
0.274105,
-0.153896,
0.155662,
0.153188,
-0.003049
],
[
-0.172093,
0.822266,
-0.345968,
0.323568,
0.260908,
0.043888,
0.316808,
-0.232134
]
],
"network.8.bias": [
0.062138,
-0.159843,
-0.184187,
-0.23454,
-0.236454,
-0.295192,
-0.196825,
-0.272223
],
"network.10.weight": [
[
0.19541,
0.042825,
-0.135471,
-0.132031,
0.087103,
0.195401,
0.847483,
0.810415
],
[
0.522677,
0.076955,
0.065811,
0.408102,
0.540989,
-0.308764,
0.899456,
0.752433
],
[
-0.128734,
0.100085,
-0.071679,
0.011353,
-0.093458,
0.276399,
-0.257554,
-0.268607
],
[
-0.36321,
-0.269912,
0.195113,
0.142973,
0.064983,
0.28648,
-0.056272,
-0.006889
],
[
-0.13283,
-0.114121,
-0.286922,
-0.192627,
-0.119257,
0.306424,
0.011788,
-0.340887
],
[
-0.517474,
-0.256841,
0.293513,
-0.235976,
0.044892,
-0.268187,
-0.354923,
-0.367305
],
[
0.040024,
0.123085,
0.085404,
0.400177,
0.271244,
-0.085198,
0.725724,
0.879043
],
[
0.488409,
0.06279,
0.05085,
-0.175453,
-0.248013,
0.04602,
0.453604,
0.658079
]
],
"network.10.bias": [
-0.201142,
-0.285005,
0.923273,
-0.077896,
-0.014217,
0.755961,
-0.260449,
0.003405
],
"network.12.weight": [
[
-0.509521,
-0.536176,
0.890526,
0.071153,
-0.176096,
0.722415,
-0.247458,
-0.700401
]
],
"network.12.bias": [
0.627846
]
}
## Activation Signature
### 0
mean: [0.063376, 0.037743, 0.626105, -0.994862, -3.104944, -2.259208, -0.266697, -1.747582]
std: [1.847399, 1.466372, 2.040318, 2.143571, 1.602488, 1.322643, 2.077648, 1.640994]
### 2
mean: [2.885672, -1.280297, -0.879336, -0.081342, -0.501193, 1.635376, -1.296460, -0.687653]
std: [2.735914, 1.080091, 0.763888, 1.038385, 0.441034, 2.031393, 0.876621, 0.728823]
### 4
mean: [2.618678, -1.212515, -1.827161, -1.588510, -0.606741, -0.544843, 3.775680, -2.032904]
std: [3.063373, 1.151459, 1.856895, 1.496788, 0.292262, 1.034220, 3.803824, 1.702715]
### 6
mean: [-0.369889, 5.282464, -1.186478, -1.438403, -1.013595, -0.580645, -2.626891, -1.400162]
std: [0.036568, 5.378909, 0.758486, 1.362204, 1.508978, 0.508202, 2.434535, 1.795345]
### 8
mean: [3.583655, -1.328089, -1.303565, -0.923827, -1.787546, -2.030910, 3.849186, 4.091274]
std: [3.774854, 1.112574, 1.103465, 0.638097, 1.557610, 1.806481, 4.166664, 4.404597]
### 10
mean: [7.089299, 8.144285, -1.632689, -1.627836, -1.840396, -3.975966, 6.282475, 6.201694]
std: [7.826499, 9.019740, 2.738010, 1.632457, 1.952766, 5.041571, 7.037664, 6.622725]
### 12
mean: [-12.881659]
std: [15.548018]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_descending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.854196,
-0.092639,
0.41049,
0.089286,
0.423838
],
[
-0.47722,
-0.228046,
0.392427,
-0.041102,
0.489083
],
[
-0.843243,
-0.130948,
0.486644,
0.497521,
-0.118015
],
[
-0.030515,
-0.298131,
-0.323495,
-0.48272,
0.999118
],
[
-0.337109,
-0.279193,
-0.290638,
-0.134689,
-0.338993
],
[
-0.085518,
0.15028,
-0.383101,
-0.460849,
-0.198546
],
[
-0.971222,
0.204007,
-0.159559,
0.219691,
0.335826
],
[
-0.020021,
-0.346354,
-0.683696,
0.231517,
0.307916
]
],
"network.0.bias": [
-0.276487,
-0.287632,
-0.023811,
0.065037,
-0.862644,
-0.393232,
0.018189,
-0.539606
],
"network.2.weight": [
[
1.044215,
0.196973,
0.660006,
0.666602,
0.15796,
0.014607,
0.997225,
0.456547
],
[
0.098738,
-0.159046,
-0.488649,
-0.707291,
-0.379429,
0.08679,
-0.226309,
-0.136925
],
[
-0.2073,
-0.241257,
-0.015129,
-0.105918,
-0.108384,
0.116129,
-0.476648,
0.036019
],
[
0.271984,
-0.160026,
0.357395,
-0.598103,
0.074268,
-0.299346,
-0.504545,
-0.34915
],
[
-0.04874,
-0.198476,
-0.181148,
0.249901,
0.128984,
-0.347937,
-0.14755,
0.09861
],
[
0.672113,
0.047022,
0.344212,
0.880358,
0.319213,
0.34713,
0.813753,
0.131276
],
[
-0.149367,
-0.314878,
-0.396369,
-0.217782,
-0.180023,
0.300973,
-0.004586,
-0.189692
],
[
-0.197519,
-0.32383,
-0.244718,
-0.194571,
-0.233323,
0.018656,
0.042823,
-0.003039
]
],
"network.2.bias": [
0.284603,
-0.206121,
-0.226513,
0.018308,
-0.166292,
-0.19994,
-0.422867,
-0.007465
],
"network.4.weight": [
[
0.587369,
0.099362,
0.150626,
0.056267,
-0.337414,
0.743496,
0.231243,
0.178079
],
[
-0.468213,
-0.530976,
-0.334353,
-0.556466,
0.018112,
0.122297,
-0.177667,
0.195894
],
[
-0.551423,
-0.071425,
0.416458,
0.317907,
-0.48414,
-0.193266,
0.132666,
0.018164
],
[
-0.359705,
-0.15427,
-0.384486,
0.039401,
-0.123384,
-0.264528,
-0.271941,
0.110259
],
[
-0.236862,
-0.213636,
-0.154691,
-0.17941,
0.17335,
0.224302,
-0.110065,
0.142574
],
[
-0.137197,
0.305694,
0.17183,
-0.701624,
-0.392817,
-0.279896,
-0.425736,
-0.239293
],
[
0.891422,
-0.054906,
0.022878,
0.646148,
0.538924,
0.65865,
-0.053226,
0.340254
],
[
-0.427094,
0.070633,
-0.024134,
-0.033463,
0.002282,
-0.27249,
-0.31324,
-0.213376
]
],
"network.4.bias": [
-0.344562,
0.08793,
0.008729,
-0.112378,
-0.254725,
0.52753,
-0.10029,
-0.330214
],
"network.6.weight": [
[
0.06138,
0.182144,
-0.155629,
0.110237,
-0.028076,
0.016976,
-0.039606,
-0.236001
],
[
0.809013,
0.039365,
0.211982,
-0.000801,
-0.124845,
-0.298362,
0.761867,
-0.116867
],
[
0.105737,
-0.089551,
0.220108,
0.276074,
0.045164,
-0.234272,
-0.292398,
-0.047492
],
[
0.076747,
0.025539,
-0.464997,
-0.463146,
-0.278464,
0.120344,
-0.413611,
-0.028017
],
[
-0.20721,
-0.291787,
-0.051211,
-0.019659,
-0.117227,
0.528096,
-0.210836,
-0.207246
],
[
0.188564,
-0.325872,
-0.232224,
-0.26002,
0.235042,
0.018246,
-0.280914,
-0.051083
],
[
-0.353675,
0.142282,
0.065089,
-0.049748,
-0.03652,
-0.038194,
-0.361745,
-0.000234
],
[
-0.441632,
0.146891,
0.525922,
0.477024,
-0.329687,
0.287178,
-0.110529,
0.401029
]
],
"network.6.bias": [
-0.387132,
0.29662,
-0.324888,
-0.102165,
0.245739,
-0.026586,
-0.310219,
0.147414
],
"network.8.weight": [
[
-0.095993,
0.684941,
-0.046161,
0.146892,
-0.403152,
0.098618,
0.272946,
-0.390233
],
[
-0.299456,
-0.213574,
-0.35091,
-0.009799,
-0.168921,
-0.252661,
0.128925,
-0.158387
],
[
0.034153,
-0.20834,
0.346516,
0.283829,
-0.279771,
-0.299774,
0.286764,
0.277514
],
[
0.034666,
-0.124179,
0.27405,
-0.128117,
0.066239,
-0.520536,
-0.03925,
-0.494994
],
[
0.05659,
-0.291505,
-0.321619,
0.130038,
0.138658,
0.093079,
0.037364,
-0.371194
],
[
0.271057,
-0.332378,
0.217979,
-0.054817,
0.093147,
0.260026,
0.349412,
0.064331
],
[
-0.08696,
0.770487,
0.32419,
0.274105,
-0.153896,
0.155662,
0.153188,
-0.003049
],
[
-0.172093,
0.822266,
-0.345968,
0.323568,
0.260908,
0.043888,
0.316808,
-0.232134
]
],
"network.8.bias": [
0.062138,
-0.159843,
-0.184187,
-0.23454,
-0.236454,
-0.295192,
-0.196825,
-0.272223
],
"network.10.weight": [
[
0.19541,
0.042825,
-0.135471,
-0.132031,
0.087103,
0.195401,
0.847483,
0.810415
],
[
0.522677,
0.076955,
0.065811,
0.408102,
0.540989,
-0.308764,
0.899456,
0.752433
],
[
-0.128734,
0.100085,
-0.071679,
0.011353,
-0.093458,
0.276399,
-0.257554,
-0.268607
],
[
-0.36321,
-0.269912,
0.195113,
0.142973,
0.064983,
0.28648,
-0.056272,
-0.006889
],
[
-0.13283,
-0.114121,
-0.286922,
-0.192627,
-0.119257,
0.306424,
0.011788,
-0.340887
],
[
-0.517474,
-0.256841,
0.293513,
-0.235976,
0.044892,
-0.268187,
-0.354923,
-0.367305
],
[
0.040024,
0.123085,
0.085404,
0.400177,
0.271244,
-0.085198,
0.725724,
0.879043
],
[
0.488409,
0.06279,
0.05085,
-0.175453,
-0.248013,
0.04602,
0.453604,
0.658079
]
],
"network.10.bias": [
-0.201142,
-0.285005,
0.923273,
-0.077896,
-0.014217,
0.755961,
-0.260449,
0.003405
],
"network.12.weight": [
[
-0.509521,
-0.536176,
0.890526,
0.071153,
-0.176096,
0.722415,
-0.247458,
-0.700401
]
],
"network.12.bias": [
0.627846
]
}
## Activation Signature
### 0
mean: [0.063376, 0.037743, 0.626105, -0.994862, -3.104944, -2.259208, -0.266697, -1.747582]
std: [1.847399, 1.466372, 2.040318, 2.143571, 1.602488, 1.322643, 2.077648, 1.640994]
### 2
mean: [2.885672, -1.280297, -0.879336, -0.081342, -0.501193, 1.635376, -1.296460, -0.687653]
std: [2.735914, 1.080091, 0.763888, 1.038385, 0.441034, 2.031393, 0.876621, 0.728823]
### 4
mean: [2.618678, -1.212515, -1.827161, -1.588510, -0.606741, -0.544843, 3.775680, -2.032904]
std: [3.063373, 1.151459, 1.856895, 1.496788, 0.292262, 1.034220, 3.803824, 1.702715]
### 6
mean: [-0.369889, 5.282464, -1.186478, -1.438403, -1.013595, -0.580645, -2.626891, -1.400162]
std: [0.036568, 5.378909, 0.758486, 1.362204, 1.508978, 0.508202, 2.434535, 1.795345]
### 8
mean: [3.583655, -1.328089, -1.303565, -0.923827, -1.787546, -2.030910, 3.849186, 4.091274]
std: [3.774854, 1.112574, 1.103465, 0.638097, 1.557610, 1.806481, 4.166664, 4.404597]
### 10
mean: [7.089299, 8.144285, -1.632689, -1.627836, -1.840396, -3.975966, 6.282475, 6.201694]
std: [7.826499, 9.019740, 2.738010, 1.632457, 1.952766, 5.041571, 7.037664, 6.622725]
### 12
mean: [-12.881659]
std: [15.548018]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_descending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.06337586045265198, "std": 1.8473985195159912}, "1": {"mean": 0.037743061780929565, "std": 1.466372013092041}, "2": {"mean": 0.6261045336723328, "std": 2.0403175354003906}, "3": {"mean": -0.9948616623878479, "std": 2.143571376800537}, "4": {"mean": -3.1049444675445557, "std": 1.6024876832962036}, "5": {"mean": -2.2592077255249023, "std": 1.3226426839828491}, "6": {"mean": -0.2666968107223511, "std": 2.077648401260376}, "7": {"mean": -1.7475823163986206, "std": 1.6409943103790283}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 2.885671615600586, "std": 2.7359135150909424}, "1": {"mean": -1.280296802520752, "std": 1.0800906419754028}, "2": {"mean": -0.8793362975120544, "std": 0.7638876438140869}, "3": {"mean": -0.08134210109710693, "std": 1.038385272026062}, "4": {"mean": -0.5011930465698242, "std": 0.4410340487957001}, "5": {"mean": 1.6353763341903687, "std": 2.03139328956604}, "6": {"mean": -1.2964599132537842, "std": 0.876620888710022}, "7": {"mean": -0.6876532435417175, "std": 0.7288232445716858}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 2.618677854537964, "std": 3.06337308883667}, "1": {"mean": -1.2125146389007568, "std": 1.151458978652954}, "2": {"mean": -1.8271605968475342, "std": 1.8568949699401855}, "3": {"mean": -1.5885097980499268, "std": 1.4967877864837646}, "4": {"mean": -0.6067413091659546, "std": 0.29226207733154297}, "5": {"mean": -0.544842541217804, "std": 1.0342202186584473}, "6": {"mean": 3.77567982673645, "std": 3.803823947906494}, "7": {"mean": -2.0329043865203857, "std": 1.7027153968811035}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.3698887825012207, "std": 0.036567702889442444}, "1": {"mean": 5.282463550567627, "std": 5.378908634185791}, "2": {"mean": -1.1864783763885498, "std": 0.7584856152534485}, "3": {"mean": -1.4384031295776367, "std": 1.3622043132781982}, "4": {"mean": -1.0135953426361084, "std": 1.5089784860610962}, "5": {"mean": -0.5806445479393005, "std": 0.5082020163536072}, "6": {"mean": -2.6268913745880127, "std": 2.434535264968872}, "7": {"mean": -1.4001622200012207, "std": 1.7953451871871948}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 3.5836551189422607, "std": 3.7748539447784424}, "1": {"mean": -1.3280885219573975, "std": 1.1125744581222534}, "2": {"mean": -1.3035649061203003, "std": 1.103464961051941}, "3": {"mean": -0.9238267540931702, "std": 0.6380967497825623}, "4": {"mean": -1.7875462770462036, "std": 1.5576101541519165}, "5": {"mean": -2.030909776687622, "std": 1.806480884552002}, "6": {"mean": 3.849186420440674, "std": 4.166663646697998}, "7": {"mean": 4.091273784637451, "std": 4.404597282409668}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": 7.089298725128174, "std": 7.826498985290527}, "1": {"mean": 8.144285202026367, "std": 9.019740104675293}, "2": {"mean": -1.6326887607574463, "std": 2.7380096912384033}, "3": {"mean": -1.627835988998413, "std": 1.6324570178985596}, "4": {"mean": -1.8403962850570679, "std": 1.9527662992477417}, "5": {"mean": -3.975965976715088, "std": 5.041570663452148}, "6": {"mean": 6.282475471496582, "std": 7.037664413452148}, "7": {"mean": 6.201694011688232, "std": 6.622724533081055}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -12.881658554077148, "std": 15.548018455505371}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.854196, -0.092639, 0.41049, 0.089286, 0.423838], [-0.47722, -0.228046, 0.392427, -0.041102, 0.489083], [-0.843243, -0.130948, 0.486644, 0.497521, -0.118015], [-0.030515, -0.298131, -0.323495, -0.48272, 0.999118], [-0.337109, -0.279193, -0.290638, -0.134689, -0.338993], [-0.085518, 0.15028, -0.383101, -0.460849, -0.198546], [-0.971222, 0.204007, -0.159559, 0.219691, 0.335826], [-0.020021, -0.346354, -0.683696, 0.231517, 0.307916]], "network.0.bias": [-0.276487, -0.287632, -0.023811, 0.065037, -0.862644, -0.393232, 0.018189, -0.539606], "network.2.weight": [[1.044215, 0.196973, 0.660006, 0.666602, 0.15796, 0.014607, 0.997225, 0.456547], [0.098738, -0.159046, -0.488649, -0.707291, -0.379429, 0.08679, -0.226309, -0.136925], [-0.2073, -0.241257, -0.015129, -0.105918, -0.108384, 0.116129, -0.476648, 0.036019], [0.271984, -0.160026, 0.357395, -0.598103, 0.074268, -0.299346, -0.504545, -0.34915], [-0.04874, -0.198476, -0.181148, 0.249901, 0.128984, -0.347937, -0.14755, 0.09861], [0.672113, 0.047022, 0.344212, 0.880358, 0.319213, 0.34713, 0.813753, 0.131276], [-0.149367, -0.314878, -0.396369, -0.217782, -0.180023, 0.300973, -0.004586, -0.189692], [-0.197519, -0.32383, -0.244718, -0.194571, -0.233323, 0.018656, 0.042823, -0.003039]], "network.2.bias": [0.284603, -0.206121, -0.226513, 0.018308, -0.166292, -0.19994, -0.422867, -0.007465], "network.4.weight": [[0.587369, 0.099362, 0.150626, 0.056267, -0.337414, 0.743496, 0.231243, 0.178079], [-0.468213, -0.530976, -0.334353, -0.556466, 0.018112, 0.122297, -0.177667, 0.195894], [-0.551423, -0.071425, 0.416458, 0.317907, -0.48414, -0.193266, 0.132666, 0.018164], [-0.359705, -0.15427, -0.384486, 0.039401, -0.123384, -0.264528, -0.271941, 0.110259], [-0.236862, -0.213636, -0.154691, -0.17941, 0.17335, 0.224302, -0.110065, 0.142574], [-0.137197, 0.305694, 0.17183, -0.701624, -0.392817, -0.279896, -0.425736, -0.239293], [0.891422, -0.054906, 0.022878, 0.646148, 0.538924, 0.65865, -0.053226, 0.340254], [-0.427094, 0.070633, -0.024134, -0.033463, 0.002282, -0.27249, -0.31324, -0.213376]], "network.4.bias": [-0.344562, 0.08793, 0.008729, -0.112378, -0.254725, 0.52753, -0.10029, -0.330214], "network.6.weight": [[0.06138, 0.182144, -0.155629, 0.110237, -0.028076, 0.016976, -0.039606, -0.236001], [0.809013, 0.039365, 0.211982, -0.000801, -0.124845, -0.298362, 0.761867, -0.116867], [0.105737, -0.089551, 0.220108, 0.276074, 0.045164, -0.234272, -0.292398, -0.047492], [0.076747, 0.025539, -0.464997, -0.463146, -0.278464, 0.120344, -0.413611, -0.028017], [-0.20721, -0.291787, -0.051211, -0.019659, -0.117227, 0.528096, -0.210836, -0.207246], [0.188564, -0.325872, -0.232224, -0.26002, 0.235042, 0.018246, -0.280914, -0.051083], [-0.353675, 0.142282, 0.065089, -0.049748, -0.03652, -0.038194, -0.361745, -0.000234], [-0.441632, 0.146891, 0.525922, 0.477024, -0.329687, 0.287178, -0.110529, 0.401029]], "network.6.bias": [-0.387132, 0.29662, -0.324888, -0.102165, 0.245739, -0.026586, -0.310219, 0.147414], "network.8.weight": [[-0.095993, 0.684941, -0.046161, 0.146892, -0.403152, 0.098618, 0.272946, -0.390233], [-0.299456, -0.213574, -0.35091, -0.009799, -0.168921, -0.252661, 0.128925, -0.158387], [0.034153, -0.20834, 0.346516, 0.283829, -0.279771, -0.299774, 0.286764, 0.277514], [0.034666, -0.124179, 0.27405, -0.128117, 0.066239, -0.520536, -0.03925, -0.494994], [0.05659, -0.291505, -0.321619, 0.130038, 0.138658, 0.093079, 0.037364, -0.371194], [0.271057, -0.332378, 0.217979, -0.054817, 0.093147, 0.260026, 0.349412, 0.064331], [-0.08696, 0.770487, 0.32419, 0.274105, -0.153896, 0.155662, 0.153188, -0.003049], [-0.172093, 0.822266, -0.345968, 0.323568, 0.260908, 0.043888, 0.316808, -0.232134]], "network.8.bias": [0.062138, -0.159843, -0.184187, -0.23454, -0.236454, -0.295192, -0.196825, -0.272223], "network.10.weight": [[0.19541, 0.042825, -0.135471, -0.132031, 0.087103, 0.195401, 0.847483, 0.810415], [0.522677, 0.076955, 0.065811, 0.408102, 0.540989, -0.308764, 0.899456, 0.752433], [-0.128734, 0.100085, -0.071679, 0.011353, -0.093458, 0.276399, -0.257554, -0.268607], [-0.36321, -0.269912, 0.195113, 0.142973, 0.064983, 0.28648, -0.056272, -0.006889], [-0.13283, -0.114121, -0.286922, -0.192627, -0.119257, 0.306424, 0.011788, -0.340887], [-0.517474, -0.256841, 0.293513, -0.235976, 0.044892, -0.268187, -0.354923, -0.367305], [0.040024, 0.123085, 0.085404, 0.400177, 0.271244, -0.085198, 0.725724, 0.879043], [0.488409, 0.06279, 0.05085, -0.175453, -0.248013, 0.04602, 0.453604, 0.658079]], "network.10.bias": [-0.201142, -0.285005, 0.923273, -0.077896, -0.014217, 0.755961, -0.260449, 0.003405], "network.12.weight": [[-0.509521, -0.536176, 0.890526, 0.071153, -0.176096, 0.722415, -0.247458, -0.700401]], "network.12.bias": [0.627846]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6984262466430664, "train_acc": 0.515, "val_loss": 0.6952366828918457, "val_acc": 0.52}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6930277645587921, "train_acc": 0.575, "val_loss": 0.7096184492111206, "val_acc": 0.52}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6798287034034729, "train_acc": 0.575, "val_loss": 0.6878149509429932, "val_acc": 0.52}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6436236202716827, "train_acc": 0.575, "val_loss": 0.6440110206604004, "val_acc": 0.52}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6268175840377808, "train_acc": 0.495, "val_loss": 0.5556864142417908, "val_acc": 0.52}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.5296579003334045, "train_acc": 0.64, "val_loss": 0.45210281014442444, "val_acc": 0.82}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.4021364152431488, "train_acc": 0.89, "val_loss": 0.4426043629646301, "val_acc": 0.86}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.34551799297332764, "train_acc": 0.925, "val_loss": 0.33124351501464844, "val_acc": 0.88}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.2709193155169487, "train_acc": 0.95, "val_loss": 0.28256818652153015, "val_acc": 0.92}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.226128451526165, "train_acc": 0.955, "val_loss": 0.2386893481016159, "val_acc": 0.92}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.1910012662410736, "train_acc": 0.955, "val_loss": 0.2115110605955124, "val_acc": 0.92}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.16452301293611526, "train_acc": 0.96, "val_loss": 0.19822227954864502, "val_acc": 0.94}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.13385484740138054, "train_acc": 0.96, "val_loss": 0.1975681632757187, "val_acc": 0.94}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.14398516714572906, "train_acc": 0.96, "val_loss": 0.20033489167690277, "val_acc": 0.94}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["sorted_descending"], "degraded_stage": {"initial_val_loss": 0.6952366828918457, "final_val_loss": 0.6440110206604004, "initial_val_acc": 0.52, "final_val_acc": 0.52, "best_val_acc": 0.52}, "improved_stage": {"initial_val_loss": 0.5556864142417908, "final_val_loss": 0.20033489167690277, "initial_val_acc": 0.52, "final_val_acc": 0.94, "best_val_acc": 0.94, "best_epoch": 11}, "improvement": 0.41999999999999993, "first_improvement_epoch": 3}}
|
38
|
{"target_pattern": "sorted_ascending", "degraded_accuracy": 0.6, "improved_accuracy": 0.98, "improvement": 0.38, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 9575, "learning_rate": 0.08435096790732609, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_ascending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_ascending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-1.058801,
-0.948364,
-0.136213,
-0.017235,
0.56512
],
[
1.090776,
0.979003,
0.761003,
0.670476,
-0.260986
],
[
-0.863303,
-1.66815,
-0.239139,
0.263914,
0.432887
],
[
-0.797851,
-1.304248,
0.003687,
-0.005423,
0.152446
],
[
0.976062,
0.723552,
0.601139,
-0.961688,
0.413281
],
[
0.799446,
0.749516,
0.294647,
0.042874,
0.250877
]
],
"network.0.bias": [
0.670824,
0.679162,
0.065381,
-0.174253,
0.059971,
0.270233
],
"network.2.weight": [
[
0.460121,
-0.130172,
0.234883,
0.169972,
0.331307,
0.165987
],
[
-0.116896,
0.650024,
-0.427921,
-0.568874,
0.947713,
0.36304
],
[
0.950708,
-0.198004,
0.859996,
0.876675,
-0.685905,
0.336948
],
[
-0.575706,
-0.512661,
-0.313459,
-0.28192,
-0.820287,
-0.601354
],
[
-0.730668,
-0.67353,
-0.764803,
-0.522309,
0.234171,
-0.859969
],
[
-1.153085,
-0.08388,
-0.795339,
-0.130795,
0.050035,
-0.048258
]
],
"network.2.bias": [
0.315224,
0.451052,
0.037191,
-0.550715,
-0.538268,
-0.417998
],
"network.4.weight": [
[
-0.434652,
0.738465,
-0.453536,
0.08508,
0.161903,
-0.834221
],
[
0.29494,
1.117025,
-0.598163,
0.306898,
-0.025597,
-0.157294
],
[
0.331003,
0.550889,
-0.14892,
0.111602,
0.232262,
-0.229122
],
[
-0.748023,
-0.44968,
-0.792594,
0.781651,
1.063773,
0.799142
],
[
-0.41971,
-0.199052,
-0.885644,
0.405308,
0.720564,
0.738529
],
[
-0.817215,
-0.168185,
-0.965367,
-0.073032,
0.239982,
0.731356
]
],
"network.4.bias": [
0.595041,
0.251265,
0.646762,
-1.129549,
-0.963653,
-0.808335
],
"network.6.weight": [
[
-0.968547,
-0.544398,
-0.69477,
0.295544,
0.649574,
0.141577
],
[
0.294103,
0.040594,
0.348541,
-0.181696,
-0.232951,
-0.028064
],
[
0.988719,
0.830604,
0.629307,
-0.936919,
-0.918716,
-0.476567
],
[
-0.53436,
-0.446819,
-1.353241,
0.701272,
0.697542,
0.755871
],
[
-0.502183,
-0.829797,
-0.732927,
-0.398626,
0.095056,
-0.384009
],
[
-0.196863,
-0.027847,
-0.411737,
0.227215,
0.512473,
0.635372
]
],
"network.6.bias": [
-0.356162,
0.701195,
-0.117852,
-0.666543,
0.571903,
-0.750429
],
"network.8.weight": [
[
0.09744,
-1.007206,
0.785776,
0.340151,
-1.069389,
0.013123
],
[
-0.087329,
0.006656,
0.292404,
0.645541,
-0.430157,
0.084136
],
[
-0.432913,
0.343732,
-0.121054,
-0.71865,
0.279805,
-0.546455
],
[
-0.325213,
0.488124,
0.241248,
-0.14391,
0.558416,
-0.782208
],
[
-0.101767,
0.4188,
-0.266954,
-1.082822,
0.294356,
-0.927463
],
[
0.32711,
-0.955041,
0.745815,
0.675198,
-0.193881,
-0.212115
]
],
"network.8.bias": [
0.124003,
-0.622983,
0.676458,
0.323046,
0.720449,
-0.373694
],
"network.10.weight": [
[
0.107337,
-1.122241,
0.901227,
0.202531,
0.891584,
-1.275808
],
[
0.716567,
0.264614,
-0.816354,
-0.47349,
-0.564633,
0.60648
],
[
-0.121755,
-0.374999,
0.350941,
-0.283814,
0.577771,
-0.242497
],
[
-0.380826,
-0.096793,
-0.318524,
-0.189593,
0.904901,
0.412727
],
[
-0.606474,
-0.116306,
-0.184289,
-0.932092,
-0.721353,
0.238905
],
[
-0.317129,
0.655184,
-1.067678,
0.090142,
-0.937284,
0.636964
]
],
"network.10.bias": [
0.635909,
0.000131,
0.208011,
-0.147248,
-0.134214,
-0.594298
],
"network.12.weight": [
[
0.548877,
-0.450595,
-0.074188,
-0.201479,
-0.092262,
0.024105
]
],
"network.12.bias": [
0.658753
]
}
## Activation Signature
### 0
mean: [-1.999436, 6.526100, -3.448383, -3.359675, 2.296238, 3.647726]
std: [3.317147, 4.356483, 4.109978, 3.306324, 3.824824, 2.826782]
### 2
mean: [1.103550, 8.519175, -1.611891, -8.457909, -7.627337, -1.366097]
std: [1.167800, 6.593189, 2.635283, 6.143667, 4.662080, 1.213292]
### 4
mean: [6.404308, 9.934377, 5.670584, -6.060842, -3.418996, -3.428735]
std: [4.707840, 7.823249, 3.971188, 3.602499, 1.854543, 2.178334]
### 6
mean: [-15.983961, 4.980937, 18.146225, -16.263031, -15.068469, -4.662622]
std: [11.424798, 3.049045, 13.460879, 11.284382, 11.681401, 2.734249]
### 8
mean: [9.393824, 4.716739, 0.198118, 7.130080, -2.026069, 8.418846]
std: [7.474255, 3.956063, 0.590268, 4.738174, 2.331115, 7.109664]
### 10
mean: [-12.726431, 9.509023, -6.660021, -2.075345, -11.064937, 5.243124]
std: [12.031780, 8.724016, 5.614955, 1.240598, 7.621261, 5.545975]
### 12
mean: [-3.389136]
std: [3.933540]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_ascending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-1.058801,
-0.948364,
-0.136213,
-0.017235,
0.56512
],
[
1.090776,
0.979003,
0.761003,
0.670476,
-0.260986
],
[
-0.863303,
-1.66815,
-0.239139,
0.263914,
0.432887
],
[
-0.797851,
-1.304248,
0.003687,
-0.005423,
0.152446
],
[
0.976062,
0.723552,
0.601139,
-0.961688,
0.413281
],
[
0.799446,
0.749516,
0.294647,
0.042874,
0.250877
]
],
"network.0.bias": [
0.670824,
0.679162,
0.065381,
-0.174253,
0.059971,
0.270233
],
"network.2.weight": [
[
0.460121,
-0.130172,
0.234883,
0.169972,
0.331307,
0.165987
],
[
-0.116896,
0.650024,
-0.427921,
-0.568874,
0.947713,
0.36304
],
[
0.950708,
-0.198004,
0.859996,
0.876675,
-0.685905,
0.336948
],
[
-0.575706,
-0.512661,
-0.313459,
-0.28192,
-0.820287,
-0.601354
],
[
-0.730668,
-0.67353,
-0.764803,
-0.522309,
0.234171,
-0.859969
],
[
-1.153085,
-0.08388,
-0.795339,
-0.130795,
0.050035,
-0.048258
]
],
"network.2.bias": [
0.315224,
0.451052,
0.037191,
-0.550715,
-0.538268,
-0.417998
],
"network.4.weight": [
[
-0.434652,
0.738465,
-0.453536,
0.08508,
0.161903,
-0.834221
],
[
0.29494,
1.117025,
-0.598163,
0.306898,
-0.025597,
-0.157294
],
[
0.331003,
0.550889,
-0.14892,
0.111602,
0.232262,
-0.229122
],
[
-0.748023,
-0.44968,
-0.792594,
0.781651,
1.063773,
0.799142
],
[
-0.41971,
-0.199052,
-0.885644,
0.405308,
0.720564,
0.738529
],
[
-0.817215,
-0.168185,
-0.965367,
-0.073032,
0.239982,
0.731356
]
],
"network.4.bias": [
0.595041,
0.251265,
0.646762,
-1.129549,
-0.963653,
-0.808335
],
"network.6.weight": [
[
-0.968547,
-0.544398,
-0.69477,
0.295544,
0.649574,
0.141577
],
[
0.294103,
0.040594,
0.348541,
-0.181696,
-0.232951,
-0.028064
],
[
0.988719,
0.830604,
0.629307,
-0.936919,
-0.918716,
-0.476567
],
[
-0.53436,
-0.446819,
-1.353241,
0.701272,
0.697542,
0.755871
],
[
-0.502183,
-0.829797,
-0.732927,
-0.398626,
0.095056,
-0.384009
],
[
-0.196863,
-0.027847,
-0.411737,
0.227215,
0.512473,
0.635372
]
],
"network.6.bias": [
-0.356162,
0.701195,
-0.117852,
-0.666543,
0.571903,
-0.750429
],
"network.8.weight": [
[
0.09744,
-1.007206,
0.785776,
0.340151,
-1.069389,
0.013123
],
[
-0.087329,
0.006656,
0.292404,
0.645541,
-0.430157,
0.084136
],
[
-0.432913,
0.343732,
-0.121054,
-0.71865,
0.279805,
-0.546455
],
[
-0.325213,
0.488124,
0.241248,
-0.14391,
0.558416,
-0.782208
],
[
-0.101767,
0.4188,
-0.266954,
-1.082822,
0.294356,
-0.927463
],
[
0.32711,
-0.955041,
0.745815,
0.675198,
-0.193881,
-0.212115
]
],
"network.8.bias": [
0.124003,
-0.622983,
0.676458,
0.323046,
0.720449,
-0.373694
],
"network.10.weight": [
[
0.107337,
-1.122241,
0.901227,
0.202531,
0.891584,
-1.275808
],
[
0.716567,
0.264614,
-0.816354,
-0.47349,
-0.564633,
0.60648
],
[
-0.121755,
-0.374999,
0.350941,
-0.283814,
0.577771,
-0.242497
],
[
-0.380826,
-0.096793,
-0.318524,
-0.189593,
0.904901,
0.412727
],
[
-0.606474,
-0.116306,
-0.184289,
-0.932092,
-0.721353,
0.238905
],
[
-0.317129,
0.655184,
-1.067678,
0.090142,
-0.937284,
0.636964
]
],
"network.10.bias": [
0.635909,
0.000131,
0.208011,
-0.147248,
-0.134214,
-0.594298
],
"network.12.weight": [
[
0.548877,
-0.450595,
-0.074188,
-0.201479,
-0.092262,
0.024105
]
],
"network.12.bias": [
0.658753
]
}
## Activation Signature
### 0
mean: [-1.999436, 6.526100, -3.448383, -3.359675, 2.296238, 3.647726]
std: [3.317147, 4.356483, 4.109978, 3.306324, 3.824824, 2.826782]
### 2
mean: [1.103550, 8.519175, -1.611891, -8.457909, -7.627337, -1.366097]
std: [1.167800, 6.593189, 2.635283, 6.143667, 4.662080, 1.213292]
### 4
mean: [6.404308, 9.934377, 5.670584, -6.060842, -3.418996, -3.428735]
std: [4.707840, 7.823249, 3.971188, 3.602499, 1.854543, 2.178334]
### 6
mean: [-15.983961, 4.980937, 18.146225, -16.263031, -15.068469, -4.662622]
std: [11.424798, 3.049045, 13.460879, 11.284382, 11.681401, 2.734249]
### 8
mean: [9.393824, 4.716739, 0.198118, 7.130080, -2.026069, 8.418846]
std: [7.474255, 3.956063, 0.590268, 4.738174, 2.331115, 7.109664]
### 10
mean: [-12.726431, 9.509023, -6.660021, -2.075345, -11.064937, 5.243124]
std: [12.031780, 8.724016, 5.614955, 1.240598, 7.621261, 5.545975]
### 12
mean: [-3.389136]
std: [3.933540]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_ascending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.9994359016418457, "std": 3.3171472549438477}, "1": {"mean": 6.526100158691406, "std": 4.35648250579834}, "2": {"mean": -3.44838285446167, "std": 4.109977722167969}, "3": {"mean": -3.359675168991089, "std": 3.306323528289795}, "4": {"mean": 2.2962377071380615, "std": 3.8248238563537598}, "5": {"mean": 3.647725820541382, "std": 2.826781988143921}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 1.1035500764846802, "std": 1.167799711227417}, "1": {"mean": 8.519174575805664, "std": 6.593189239501953}, "2": {"mean": -1.6118913888931274, "std": 2.6352834701538086}, "3": {"mean": -8.457908630371094, "std": 6.143667221069336}, "4": {"mean": -7.6273369789123535, "std": 4.66208028793335}, "5": {"mean": -1.3660969734191895, "std": 1.2132915258407593}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 6.404308319091797, "std": 4.7078399658203125}, "1": {"mean": 9.93437671661377, "std": 7.823248863220215}, "2": {"mean": 5.670583724975586, "std": 3.9711880683898926}, "3": {"mean": -6.060842037200928, "std": 3.602499485015869}, "4": {"mean": -3.4189960956573486, "std": 1.8545430898666382}, "5": {"mean": -3.4287352561950684, "std": 2.1783335208892822}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -15.98396110534668, "std": 11.424798011779785}, "1": {"mean": 4.980936527252197, "std": 3.0490453243255615}, "2": {"mean": 18.146224975585938, "std": 13.4608793258667}, "3": {"mean": -16.263031005859375, "std": 11.284381866455078}, "4": {"mean": -15.068469047546387, "std": 11.681401252746582}, "5": {"mean": -4.662621974945068, "std": 2.7342491149902344}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 9.393823623657227, "std": 7.474255084991455}, "1": {"mean": 4.716738700866699, "std": 3.9560630321502686}, "2": {"mean": 0.19811755418777466, "std": 0.5902684926986694}, "3": {"mean": 7.130080223083496, "std": 4.7381744384765625}, "4": {"mean": -2.026069402694702, "std": 2.3311145305633545}, "5": {"mean": 8.418846130371094, "std": 7.109663963317871}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -12.726430892944336, "std": 12.031780242919922}, "1": {"mean": 9.50902271270752, "std": 8.724016189575195}, "2": {"mean": -6.6600213050842285, "std": 5.614955425262451}, "3": {"mean": -2.0753448009490967, "std": 1.2405980825424194}, "4": {"mean": -11.064936637878418, "std": 7.621260643005371}, "5": {"mean": 5.243124008178711, "std": 5.5459747314453125}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -3.3891360759735107, "std": 3.933539628982544}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-1.058801, -0.948364, -0.136213, -0.017235, 0.56512], [1.090776, 0.979003, 0.761003, 0.670476, -0.260986], [-0.863303, -1.66815, -0.239139, 0.263914, 0.432887], [-0.797851, -1.304248, 0.003687, -0.005423, 0.152446], [0.976062, 0.723552, 0.601139, -0.961688, 0.413281], [0.799446, 0.749516, 0.294647, 0.042874, 0.250877]], "network.0.bias": [0.670824, 0.679162, 0.065381, -0.174253, 0.059971, 0.270233], "network.2.weight": [[0.460121, -0.130172, 0.234883, 0.169972, 0.331307, 0.165987], [-0.116896, 0.650024, -0.427921, -0.568874, 0.947713, 0.36304], [0.950708, -0.198004, 0.859996, 0.876675, -0.685905, 0.336948], [-0.575706, -0.512661, -0.313459, -0.28192, -0.820287, -0.601354], [-0.730668, -0.67353, -0.764803, -0.522309, 0.234171, -0.859969], [-1.153085, -0.08388, -0.795339, -0.130795, 0.050035, -0.048258]], "network.2.bias": [0.315224, 0.451052, 0.037191, -0.550715, -0.538268, -0.417998], "network.4.weight": [[-0.434652, 0.738465, -0.453536, 0.08508, 0.161903, -0.834221], [0.29494, 1.117025, -0.598163, 0.306898, -0.025597, -0.157294], [0.331003, 0.550889, -0.14892, 0.111602, 0.232262, -0.229122], [-0.748023, -0.44968, -0.792594, 0.781651, 1.063773, 0.799142], [-0.41971, -0.199052, -0.885644, 0.405308, 0.720564, 0.738529], [-0.817215, -0.168185, -0.965367, -0.073032, 0.239982, 0.731356]], "network.4.bias": [0.595041, 0.251265, 0.646762, -1.129549, -0.963653, -0.808335], "network.6.weight": [[-0.968547, -0.544398, -0.69477, 0.295544, 0.649574, 0.141577], [0.294103, 0.040594, 0.348541, -0.181696, -0.232951, -0.028064], [0.988719, 0.830604, 0.629307, -0.936919, -0.918716, -0.476567], [-0.53436, -0.446819, -1.353241, 0.701272, 0.697542, 0.755871], [-0.502183, -0.829797, -0.732927, -0.398626, 0.095056, -0.384009], [-0.196863, -0.027847, -0.411737, 0.227215, 0.512473, 0.635372]], "network.6.bias": [-0.356162, 0.701195, -0.117852, -0.666543, 0.571903, -0.750429], "network.8.weight": [[0.09744, -1.007206, 0.785776, 0.340151, -1.069389, 0.013123], [-0.087329, 0.006656, 0.292404, 0.645541, -0.430157, 0.084136], [-0.432913, 0.343732, -0.121054, -0.71865, 0.279805, -0.546455], [-0.325213, 0.488124, 0.241248, -0.14391, 0.558416, -0.782208], [-0.101767, 0.4188, -0.266954, -1.082822, 0.294356, -0.927463], [0.32711, -0.955041, 0.745815, 0.675198, -0.193881, -0.212115]], "network.8.bias": [0.124003, -0.622983, 0.676458, 0.323046, 0.720449, -0.373694], "network.10.weight": [[0.107337, -1.122241, 0.901227, 0.202531, 0.891584, -1.275808], [0.716567, 0.264614, -0.816354, -0.47349, -0.564633, 0.60648], [-0.121755, -0.374999, 0.350941, -0.283814, 0.577771, -0.242497], [-0.380826, -0.096793, -0.318524, -0.189593, 0.904901, 0.412727], [-0.606474, -0.116306, -0.184289, -0.932092, -0.721353, 0.238905], [-0.317129, 0.655184, -1.067678, 0.090142, -0.937284, 0.636964]], "network.10.bias": [0.635909, 0.000131, 0.208011, -0.147248, -0.134214, -0.594298], "network.12.weight": [[0.548877, -0.450595, -0.074188, -0.201479, -0.092262, 0.024105]], "network.12.bias": [0.658753]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7200360596179962, "train_acc": 0.45, "val_loss": 0.6914308071136475, "val_acc": 0.64}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6960756182670593, "train_acc": 0.53, "val_loss": 0.6599119305610657, "val_acc": 0.6}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6720350682735443, "train_acc": 0.55, "val_loss": 0.555557906627655, "val_acc": 0.6}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.6237601637840271, "train_acc": 0.58, "val_loss": 0.4443928897380829, "val_acc": 0.82}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.4356977492570877, "train_acc": 0.835, "val_loss": 0.2618786692619324, "val_acc": 0.9}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.345466211438179, "train_acc": 0.885, "val_loss": 0.3094368875026703, "val_acc": 0.9}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.3461820036172867, "train_acc": 0.905, "val_loss": 0.18152391910552979, "val_acc": 0.94}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.23178625851869583, "train_acc": 0.92, "val_loss": 0.151564821600914, "val_acc": 0.96}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.23435165733098984, "train_acc": 0.925, "val_loss": 0.1433078795671463, "val_acc": 0.98}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.20074981451034546, "train_acc": 0.92, "val_loss": 0.11623446643352509, "val_acc": 0.96}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.21109376102685928, "train_acc": 0.905, "val_loss": 0.1472577452659607, "val_acc": 0.96}, {"stage": "improved", "epoch": 8, "global_epoch": 11, "train_loss": 0.1970955953001976, "train_acc": 0.915, "val_loss": 0.0722152516245842, "val_acc": 0.98}, {"stage": "improved", "epoch": 9, "global_epoch": 12, "train_loss": 0.196519136428833, "train_acc": 0.92, "val_loss": 0.08979497104883194, "val_acc": 0.96}], "summary": {"total_epochs": 13, "degraded_epochs": 3, "improved_epochs": 10, "patterns": ["sorted_ascending"], "degraded_stage": {"initial_val_loss": 0.6914308071136475, "final_val_loss": 0.555557906627655, "initial_val_acc": 0.64, "final_val_acc": 0.6, "best_val_acc": 0.6}, "improved_stage": {"initial_val_loss": 0.4443928897380829, "final_val_loss": 0.08979497104883194, "initial_val_acc": 0.82, "final_val_acc": 0.96, "best_val_acc": 0.98, "best_epoch": 8}, "improvement": 0.38, "first_improvement_epoch": 2}}
|
39
|
{"target_pattern": "palindrome", "degraded_accuracy": 0.7, "improved_accuracy": 0.96, "improvement": 0.26, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 6322, "learning_rate": 0.045238552165504674, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "palindrome", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["palindrome"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.127541,
0.190585,
0.44547,
-0.128883,
-0.284474
],
[
0.37781,
0.19073,
0.168595,
-0.017493,
0.460143
],
[
-0.136789,
-0.13521,
0.419287,
0.003849,
-0.196101
],
[
-0.66826,
0.079061,
-0.238477,
0.102788,
-0.457449
],
[
0.183967,
-0.460857,
-0.081233,
0.413517,
0.441996
],
[
0.394504,
0.431785,
-0.098079,
-0.455982,
0.365739
],
[
0.19756,
-0.060672,
0.495957,
0.159139,
0.118095
],
[
0.050493,
0.099888,
0.01612,
-0.187731,
0.314181
]
],
"network.0.bias": [
0.635595,
-0.018629,
0.763045,
0.102316,
-0.037144,
0.083797,
0.717244,
-0.565294
],
"network.2.weight": [
[
0.374923,
-0.17998,
0.427328,
0.091247,
-0.863608,
-0.126464,
0.356677,
-0.324316
],
[
0.338756,
0.411492,
0.235011,
-0.340228,
-0.070553,
0.481885,
0.316907,
0.213236
],
[
-0.195674,
0.178484,
-0.351328,
0.258219,
-0.536118,
-0.316532,
-0.229262,
0.012019
],
[
0.024504,
0.468506,
-0.360992,
-0.346843,
0.840732,
0.639453,
0.258693,
-0.18408
],
[
-0.185411,
-0.097476,
0.12748,
-0.123913,
0.717994,
0.503476,
0.16595,
0.352604
],
[
-0.053536,
0.35943,
0.397161,
0.22977,
0.694441,
0.446133,
-0.03433,
0.123834
],
[
0.08789,
0.025125,
0.062925,
0.503797,
-0.161126,
0.112801,
0.532987,
-0.138237
],
[
0.24079,
0.156335,
0.333965,
-0.026278,
0.433981,
-0.059293,
-0.083894,
0.035095
]
],
"network.2.bias": [
0.307146,
0.348799,
0.217198,
0.040661,
0.255072,
-0.298376,
0.318953,
-0.298025
],
"network.4.weight": [
[
-0.605443,
-0.389924,
-0.038147,
-0.153568,
0.117198,
0.237507,
-0.300906,
0.076145
],
[
-0.078607,
-0.31149,
-0.181288,
-0.118608,
0.126919,
-0.303582,
-0.289108,
0.279615
],
[
0.455154,
0.138228,
-0.242569,
-0.264095,
0.212512,
-0.578364,
0.038279,
-0.431649
],
[
-0.449247,
-0.006365,
-0.083408,
0.07281,
-0.163714,
0.121715,
0.134659,
-0.254172
],
[
-0.3765,
0.341147,
0.364337,
0.552193,
0.682936,
0.323301,
0.104023,
-0.30579
],
[
0.481476,
0.221308,
-0.312676,
-0.211254,
-0.159892,
-0.520233,
0.284404,
-0.686534
],
[
0.147231,
0.013738,
-0.137389,
0.115711,
0.308235,
-0.334042,
-0.190172,
-0.303721
],
[
0.439362,
0.008642,
0.085816,
-0.091655,
0.093305,
0.004473,
0.277763,
0.07507
]
],
"network.4.bias": [
-0.694119,
-0.008118,
0.161808,
-0.05634,
0.241972,
0.29414,
-0.156556,
0.453084
],
"network.6.weight": [
[
-0.029896,
0.025126,
0.026498,
-0.22655,
-0.20867,
0.346677,
-0.259589,
0.256999
],
[
0.452849,
0.440146,
0.364153,
0.273441,
0.13302,
-0.149571,
0.234932,
-0.101971
],
[
0.146053,
0.012646,
-0.100895,
0.059288,
0.098594,
-0.120324,
0.022778,
-0.605957
],
[
-0.047317,
-0.349453,
-0.513232,
-0.297742,
0.650593,
-0.300597,
-0.411363,
0.379481
],
[
-0.379193,
-0.094609,
-0.182859,
-0.097809,
0.518643,
-0.566572,
-0.487409,
-0.257037
],
[
-0.04913,
-0.475309,
0.63946,
-0.256737,
-0.146966,
0.331323,
-0.443085,
-0.042346
],
[
0.041096,
-0.561922,
0.298245,
-0.453047,
-0.287233,
0.558772,
-0.075638,
0.184339
],
[
0.22852,
-0.202604,
0.118727,
-0.050703,
0.416099,
-0.341812,
-0.075943,
0.44622
]
],
"network.6.bias": [
0.037995,
-0.447808,
-0.433992,
0.200925,
0.034548,
0.073287,
0.378648,
-0.043586
],
"network.8.weight": [
[
-0.18756,
0.503104,
0.100513,
0.599067,
0.347139,
-0.261031,
-0.200493,
0.416879
],
[
-0.639591,
0.139736,
0.36819,
0.02311,
0.346541,
-0.631699,
-0.248078,
-0.084484
],
[
0.224449,
0.344317,
0.213936,
-0.674297,
-0.078979,
0.462376,
0.152625,
-0.493211
],
[
0.160841,
-0.432342,
0.471627,
-0.34024,
-0.455487,
0.491402,
0.359755,
-0.491398
],
[
0.162826,
-0.320457,
0.012482,
-0.465464,
-0.293336,
-0.118577,
0.403521,
-0.280991
],
[
0.472474,
-0.416646,
-0.160369,
-0.306812,
-0.140504,
0.413664,
0.199854,
-0.322938
],
[
-0.489492,
0.363331,
0.525961,
0.276435,
-0.067072,
-0.387695,
-0.423653,
-0.05272
],
[
-0.409703,
0.097473,
0.115625,
0.221251,
0.397537,
-0.37042,
-0.311574,
0.285083
]
],
"network.8.bias": [
-0.057182,
-0.213353,
0.131818,
0.207081,
-0.261224,
0.302633,
-0.166762,
0.159533
],
"network.10.weight": [
[
-0.33272,
-0.151482,
0.121272,
0.190635,
0.157402,
0.33702,
-0.255794,
-0.46234
]
],
"network.10.bias": [
0.343045
]
}
## Activation Signature
### 0
mean: [1.127469, 1.681261, 0.988677, -1.427146, 0.614214, 0.629919, 2.375453, -0.301857]
std: [1.021173, 1.505946, 0.808845, 1.908235, 1.364264, 1.568292, 1.240091, 0.684980]
### 2
mean: [0.858761, 2.687443, -1.224656, 2.275278, 1.384167, 1.416218, 1.724283, 0.593487]
std: [1.571761, 1.667034, 0.793076, 2.201845, 1.192790, 1.412022, 0.788168, 0.561953]
### 4
mean: [-2.685851, -1.734615, -0.210359, -0.338421, 3.258325, 0.165215, -0.209862, 1.352635]
std: [1.252653, 1.143964, 1.422077, 0.554393, 3.088833, 1.694256, 0.182848, 0.674832]
### 6
mean: [-0.048616, -0.216449, -0.989409, 2.463432, 1.010905, 0.059515, 0.219668, 1.666210]
std: [0.986679, 0.451703, 0.736487, 2.245414, 2.081196, 0.954240, 1.483691, 1.350508]
### 8
mean: [2.229964, -0.395611, -2.112732, -1.607658, -1.947097, -0.734662, -0.245092, 1.233471]
std: [2.950382, 1.336874, 2.637719, 2.817663, 2.292188, 1.965181, 1.261269, 2.118217]
### 10
mean: [-1.025342]
std: [2.219177]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
palindrome
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.127541,
0.190585,
0.44547,
-0.128883,
-0.284474
],
[
0.37781,
0.19073,
0.168595,
-0.017493,
0.460143
],
[
-0.136789,
-0.13521,
0.419287,
0.003849,
-0.196101
],
[
-0.66826,
0.079061,
-0.238477,
0.102788,
-0.457449
],
[
0.183967,
-0.460857,
-0.081233,
0.413517,
0.441996
],
[
0.394504,
0.431785,
-0.098079,
-0.455982,
0.365739
],
[
0.19756,
-0.060672,
0.495957,
0.159139,
0.118095
],
[
0.050493,
0.099888,
0.01612,
-0.187731,
0.314181
]
],
"network.0.bias": [
0.635595,
-0.018629,
0.763045,
0.102316,
-0.037144,
0.083797,
0.717244,
-0.565294
],
"network.2.weight": [
[
0.374923,
-0.17998,
0.427328,
0.091247,
-0.863608,
-0.126464,
0.356677,
-0.324316
],
[
0.338756,
0.411492,
0.235011,
-0.340228,
-0.070553,
0.481885,
0.316907,
0.213236
],
[
-0.195674,
0.178484,
-0.351328,
0.258219,
-0.536118,
-0.316532,
-0.229262,
0.012019
],
[
0.024504,
0.468506,
-0.360992,
-0.346843,
0.840732,
0.639453,
0.258693,
-0.18408
],
[
-0.185411,
-0.097476,
0.12748,
-0.123913,
0.717994,
0.503476,
0.16595,
0.352604
],
[
-0.053536,
0.35943,
0.397161,
0.22977,
0.694441,
0.446133,
-0.03433,
0.123834
],
[
0.08789,
0.025125,
0.062925,
0.503797,
-0.161126,
0.112801,
0.532987,
-0.138237
],
[
0.24079,
0.156335,
0.333965,
-0.026278,
0.433981,
-0.059293,
-0.083894,
0.035095
]
],
"network.2.bias": [
0.307146,
0.348799,
0.217198,
0.040661,
0.255072,
-0.298376,
0.318953,
-0.298025
],
"network.4.weight": [
[
-0.605443,
-0.389924,
-0.038147,
-0.153568,
0.117198,
0.237507,
-0.300906,
0.076145
],
[
-0.078607,
-0.31149,
-0.181288,
-0.118608,
0.126919,
-0.303582,
-0.289108,
0.279615
],
[
0.455154,
0.138228,
-0.242569,
-0.264095,
0.212512,
-0.578364,
0.038279,
-0.431649
],
[
-0.449247,
-0.006365,
-0.083408,
0.07281,
-0.163714,
0.121715,
0.134659,
-0.254172
],
[
-0.3765,
0.341147,
0.364337,
0.552193,
0.682936,
0.323301,
0.104023,
-0.30579
],
[
0.481476,
0.221308,
-0.312676,
-0.211254,
-0.159892,
-0.520233,
0.284404,
-0.686534
],
[
0.147231,
0.013738,
-0.137389,
0.115711,
0.308235,
-0.334042,
-0.190172,
-0.303721
],
[
0.439362,
0.008642,
0.085816,
-0.091655,
0.093305,
0.004473,
0.277763,
0.07507
]
],
"network.4.bias": [
-0.694119,
-0.008118,
0.161808,
-0.05634,
0.241972,
0.29414,
-0.156556,
0.453084
],
"network.6.weight": [
[
-0.029896,
0.025126,
0.026498,
-0.22655,
-0.20867,
0.346677,
-0.259589,
0.256999
],
[
0.452849,
0.440146,
0.364153,
0.273441,
0.13302,
-0.149571,
0.234932,
-0.101971
],
[
0.146053,
0.012646,
-0.100895,
0.059288,
0.098594,
-0.120324,
0.022778,
-0.605957
],
[
-0.047317,
-0.349453,
-0.513232,
-0.297742,
0.650593,
-0.300597,
-0.411363,
0.379481
],
[
-0.379193,
-0.094609,
-0.182859,
-0.097809,
0.518643,
-0.566572,
-0.487409,
-0.257037
],
[
-0.04913,
-0.475309,
0.63946,
-0.256737,
-0.146966,
0.331323,
-0.443085,
-0.042346
],
[
0.041096,
-0.561922,
0.298245,
-0.453047,
-0.287233,
0.558772,
-0.075638,
0.184339
],
[
0.22852,
-0.202604,
0.118727,
-0.050703,
0.416099,
-0.341812,
-0.075943,
0.44622
]
],
"network.6.bias": [
0.037995,
-0.447808,
-0.433992,
0.200925,
0.034548,
0.073287,
0.378648,
-0.043586
],
"network.8.weight": [
[
-0.18756,
0.503104,
0.100513,
0.599067,
0.347139,
-0.261031,
-0.200493,
0.416879
],
[
-0.639591,
0.139736,
0.36819,
0.02311,
0.346541,
-0.631699,
-0.248078,
-0.084484
],
[
0.224449,
0.344317,
0.213936,
-0.674297,
-0.078979,
0.462376,
0.152625,
-0.493211
],
[
0.160841,
-0.432342,
0.471627,
-0.34024,
-0.455487,
0.491402,
0.359755,
-0.491398
],
[
0.162826,
-0.320457,
0.012482,
-0.465464,
-0.293336,
-0.118577,
0.403521,
-0.280991
],
[
0.472474,
-0.416646,
-0.160369,
-0.306812,
-0.140504,
0.413664,
0.199854,
-0.322938
],
[
-0.489492,
0.363331,
0.525961,
0.276435,
-0.067072,
-0.387695,
-0.423653,
-0.05272
],
[
-0.409703,
0.097473,
0.115625,
0.221251,
0.397537,
-0.37042,
-0.311574,
0.285083
]
],
"network.8.bias": [
-0.057182,
-0.213353,
0.131818,
0.207081,
-0.261224,
0.302633,
-0.166762,
0.159533
],
"network.10.weight": [
[
-0.33272,
-0.151482,
0.121272,
0.190635,
0.157402,
0.33702,
-0.255794,
-0.46234
]
],
"network.10.bias": [
0.343045
]
}
## Activation Signature
### 0
mean: [1.127469, 1.681261, 0.988677, -1.427146, 0.614214, 0.629919, 2.375453, -0.301857]
std: [1.021173, 1.505946, 0.808845, 1.908235, 1.364264, 1.568292, 1.240091, 0.684980]
### 2
mean: [0.858761, 2.687443, -1.224656, 2.275278, 1.384167, 1.416218, 1.724283, 0.593487]
std: [1.571761, 1.667034, 0.793076, 2.201845, 1.192790, 1.412022, 0.788168, 0.561953]
### 4
mean: [-2.685851, -1.734615, -0.210359, -0.338421, 3.258325, 0.165215, -0.209862, 1.352635]
std: [1.252653, 1.143964, 1.422077, 0.554393, 3.088833, 1.694256, 0.182848, 0.674832]
### 6
mean: [-0.048616, -0.216449, -0.989409, 2.463432, 1.010905, 0.059515, 0.219668, 1.666210]
std: [0.986679, 0.451703, 0.736487, 2.245414, 2.081196, 0.954240, 1.483691, 1.350508]
### 8
mean: [2.229964, -0.395611, -2.112732, -1.607658, -1.947097, -0.734662, -0.245092, 1.233471]
std: [2.950382, 1.336874, 2.637719, 2.817663, 2.292188, 1.965181, 1.261269, 2.118217]
### 10
mean: [-1.025342]
std: [2.219177]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
palindrome
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.1274690628051758, "std": 1.0211727619171143}, "1": {"mean": 1.6812613010406494, "std": 1.5059459209442139}, "2": {"mean": 0.9886770248413086, "std": 0.8088453412055969}, "3": {"mean": -1.4271458387374878, "std": 1.9082348346710205}, "4": {"mean": 0.6142143607139587, "std": 1.364263892173767}, "5": {"mean": 0.6299187541007996, "std": 1.5682923793792725}, "6": {"mean": 2.375453472137451, "std": 1.2400907278060913}, "7": {"mean": -0.3018568158149719, "std": 0.6849797368049622}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.8587610721588135, "std": 1.5717610120773315}, "1": {"mean": 2.687443256378174, "std": 1.6670337915420532}, "2": {"mean": -1.2246555089950562, "std": 0.7930757999420166}, "3": {"mean": 2.275277614593506, "std": 2.2018446922302246}, "4": {"mean": 1.3841668367385864, "std": 1.192789912223816}, "5": {"mean": 1.416217565536499, "std": 1.412021517753601}, "6": {"mean": 1.7242825031280518, "std": 0.7881678342819214}, "7": {"mean": 0.5934871435165405, "std": 0.5619531869888306}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -2.6858513355255127, "std": 1.2526530027389526}, "1": {"mean": -1.7346150875091553, "std": 1.1439638137817383}, "2": {"mean": -0.21035896241664886, "std": 1.4220770597457886}, "3": {"mean": -0.33842113614082336, "std": 0.55439293384552}, "4": {"mean": 3.2583248615264893, "std": 3.0888328552246094}, "5": {"mean": 0.16521547734737396, "std": 1.6942557096481323}, "6": {"mean": -0.20986169576644897, "std": 0.1828477382659912}, "7": {"mean": 1.3526349067687988, "std": 0.6748318672180176}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.0486159548163414, "std": 0.9866790175437927}, "1": {"mean": -0.21644937992095947, "std": 0.4517034590244293}, "2": {"mean": -0.9894087910652161, "std": 0.7364866137504578}, "3": {"mean": 2.4634320735931396, "std": 2.2454142570495605}, "4": {"mean": 1.0109050273895264, "std": 2.0811963081359863}, "5": {"mean": 0.059515178203582764, "std": 0.9542403817176819}, "6": {"mean": 0.2196684032678604, "std": 1.4836914539337158}, "7": {"mean": 1.666210412979126, "std": 1.3505077362060547}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 2.2299644947052, "std": 2.9503817558288574}, "1": {"mean": -0.39561131596565247, "std": 1.3368735313415527}, "2": {"mean": -2.112731695175171, "std": 2.637718915939331}, "3": {"mean": -1.6076582670211792, "std": 2.8176627159118652}, "4": {"mean": -1.9470969438552856, "std": 2.2921876907348633}, "5": {"mean": -0.7346621751785278, "std": 1.9651814699172974}, "6": {"mean": -0.24509243667125702, "std": 1.261269211769104}, "7": {"mean": 1.233471393585205, "std": 2.1182172298431396}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -1.0253417491912842, "std": 2.219177007675171}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.127541, 0.190585, 0.44547, -0.128883, -0.284474], [0.37781, 0.19073, 0.168595, -0.017493, 0.460143], [-0.136789, -0.13521, 0.419287, 0.003849, -0.196101], [-0.66826, 0.079061, -0.238477, 0.102788, -0.457449], [0.183967, -0.460857, -0.081233, 0.413517, 0.441996], [0.394504, 0.431785, -0.098079, -0.455982, 0.365739], [0.19756, -0.060672, 0.495957, 0.159139, 0.118095], [0.050493, 0.099888, 0.01612, -0.187731, 0.314181]], "network.0.bias": [0.635595, -0.018629, 0.763045, 0.102316, -0.037144, 0.083797, 0.717244, -0.565294], "network.2.weight": [[0.374923, -0.17998, 0.427328, 0.091247, -0.863608, -0.126464, 0.356677, -0.324316], [0.338756, 0.411492, 0.235011, -0.340228, -0.070553, 0.481885, 0.316907, 0.213236], [-0.195674, 0.178484, -0.351328, 0.258219, -0.536118, -0.316532, -0.229262, 0.012019], [0.024504, 0.468506, -0.360992, -0.346843, 0.840732, 0.639453, 0.258693, -0.18408], [-0.185411, -0.097476, 0.12748, -0.123913, 0.717994, 0.503476, 0.16595, 0.352604], [-0.053536, 0.35943, 0.397161, 0.22977, 0.694441, 0.446133, -0.03433, 0.123834], [0.08789, 0.025125, 0.062925, 0.503797, -0.161126, 0.112801, 0.532987, -0.138237], [0.24079, 0.156335, 0.333965, -0.026278, 0.433981, -0.059293, -0.083894, 0.035095]], "network.2.bias": [0.307146, 0.348799, 0.217198, 0.040661, 0.255072, -0.298376, 0.318953, -0.298025], "network.4.weight": [[-0.605443, -0.389924, -0.038147, -0.153568, 0.117198, 0.237507, -0.300906, 0.076145], [-0.078607, -0.31149, -0.181288, -0.118608, 0.126919, -0.303582, -0.289108, 0.279615], [0.455154, 0.138228, -0.242569, -0.264095, 0.212512, -0.578364, 0.038279, -0.431649], [-0.449247, -0.006365, -0.083408, 0.07281, -0.163714, 0.121715, 0.134659, -0.254172], [-0.3765, 0.341147, 0.364337, 0.552193, 0.682936, 0.323301, 0.104023, -0.30579], [0.481476, 0.221308, -0.312676, -0.211254, -0.159892, -0.520233, 0.284404, -0.686534], [0.147231, 0.013738, -0.137389, 0.115711, 0.308235, -0.334042, -0.190172, -0.303721], [0.439362, 0.008642, 0.085816, -0.091655, 0.093305, 0.004473, 0.277763, 0.07507]], "network.4.bias": [-0.694119, -0.008118, 0.161808, -0.05634, 0.241972, 0.29414, -0.156556, 0.453084], "network.6.weight": [[-0.029896, 0.025126, 0.026498, -0.22655, -0.20867, 0.346677, -0.259589, 0.256999], [0.452849, 0.440146, 0.364153, 0.273441, 0.13302, -0.149571, 0.234932, -0.101971], [0.146053, 0.012646, -0.100895, 0.059288, 0.098594, -0.120324, 0.022778, -0.605957], [-0.047317, -0.349453, -0.513232, -0.297742, 0.650593, -0.300597, -0.411363, 0.379481], [-0.379193, -0.094609, -0.182859, -0.097809, 0.518643, -0.566572, -0.487409, -0.257037], [-0.04913, -0.475309, 0.63946, -0.256737, -0.146966, 0.331323, -0.443085, -0.042346], [0.041096, -0.561922, 0.298245, -0.453047, -0.287233, 0.558772, -0.075638, 0.184339], [0.22852, -0.202604, 0.118727, -0.050703, 0.416099, -0.341812, -0.075943, 0.44622]], "network.6.bias": [0.037995, -0.447808, -0.433992, 0.200925, 0.034548, 0.073287, 0.378648, -0.043586], "network.8.weight": [[-0.18756, 0.503104, 0.100513, 0.599067, 0.347139, -0.261031, -0.200493, 0.416879], [-0.639591, 0.139736, 0.36819, 0.02311, 0.346541, -0.631699, -0.248078, -0.084484], [0.224449, 0.344317, 0.213936, -0.674297, -0.078979, 0.462376, 0.152625, -0.493211], [0.160841, -0.432342, 0.471627, -0.34024, -0.455487, 0.491402, 0.359755, -0.491398], [0.162826, -0.320457, 0.012482, -0.465464, -0.293336, -0.118577, 0.403521, -0.280991], [0.472474, -0.416646, -0.160369, -0.306812, -0.140504, 0.413664, 0.199854, -0.322938], [-0.489492, 0.363331, 0.525961, 0.276435, -0.067072, -0.387695, -0.423653, -0.05272], [-0.409703, 0.097473, 0.115625, 0.221251, 0.397537, -0.37042, -0.311574, 0.285083]], "network.8.bias": [-0.057182, -0.213353, 0.131818, 0.207081, -0.261224, 0.302633, -0.166762, 0.159533], "network.10.weight": [[-0.33272, -0.151482, 0.121272, 0.190635, 0.157402, 0.33702, -0.255794, -0.46234]], "network.10.bias": [0.343045]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7326966822147369, "train_acc": 0.445, "val_loss": 0.7142817974090576, "val_acc": 0.42}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.7006776034832001, "train_acc": 0.445, "val_loss": 0.6572497487068176, "val_acc": 0.7}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.6234430372714996, "train_acc": 0.705, "val_loss": 0.39841827750205994, "val_acc": 0.88}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.3988029956817627, "train_acc": 0.855, "val_loss": 0.17230501770973206, "val_acc": 0.92}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.35970743000507355, "train_acc": 0.86, "val_loss": 0.16490398347377777, "val_acc": 0.92}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.3307691663503647, "train_acc": 0.88, "val_loss": 0.19351769983768463, "val_acc": 0.92}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.2819754555821419, "train_acc": 0.865, "val_loss": 0.16679725050926208, "val_acc": 0.94}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.2852007746696472, "train_acc": 0.885, "val_loss": 0.1694750189781189, "val_acc": 0.96}], "summary": {"total_epochs": 8, "degraded_epochs": 2, "improved_epochs": 6, "patterns": ["palindrome"], "degraded_stage": {"initial_val_loss": 0.7142817974090576, "final_val_loss": 0.6572497487068176, "initial_val_acc": 0.42, "final_val_acc": 0.7, "best_val_acc": 0.7}, "improved_stage": {"initial_val_loss": 0.39841827750205994, "final_val_loss": 0.1694750189781189, "initial_val_acc": 0.88, "final_val_acc": 0.96, "best_val_acc": 0.96, "best_epoch": 7}, "improvement": 0.26, "first_improvement_epoch": 1}}
|
40
|
{"target_pattern": "ends_with", "degraded_accuracy": 0.58, "improved_accuracy": 0.78, "improvement": 0.20000000000000007, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 2084, "learning_rate": 0.07613021159903365, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "ends_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["ends_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.388896,
-0.076341,
0.102071,
-0.498515,
0.345834
],
[
0.464564,
-0.109712,
-0.195431,
-0.095248,
0.43856
],
[
-0.502554,
-0.244549,
0.164886,
0.220243,
1.050065
],
[
-0.296963,
0.174377,
0.066301,
0.168926,
-0.991912
],
[
-0.700244,
-0.53192,
-0.552613,
-0.054345,
-0.223
],
[
0.739396,
-0.112262,
-0.078898,
-0.384687,
0.031831
],
[
-0.440265,
0.451636,
0.252654,
0.228348,
-0.476928
],
[
-0.101593,
-0.237787,
0.383359,
0.095351,
-0.84944
]
],
"network.0.bias": [
-0.265049,
0.563351,
0.045538,
0.431804,
-0.596582,
0.14807,
-0.087178,
0.334263
],
"network.2.weight": [
[
0.000611,
0.27097,
0.958709,
-0.068399,
0.316792,
-0.270839,
0.3047,
-0.376894
],
[
0.146379,
-0.028965,
-0.337845,
0.072932,
-0.211553,
0.014745,
-0.460602,
-0.484833
],
[
0.097868,
0.320699,
0.769431,
-0.220845,
0.078783,
0.335818,
-0.241861,
-0.224117
],
[
-0.129189,
-0.095569,
-0.319079,
0.373867,
0.078455,
-0.396071,
0.582442,
0.612767
],
[
0.115895,
0.729566,
0.935262,
-0.678243,
0.353549,
0.670145,
-0.284442,
-0.162325
],
[
-0.17055,
0.492571,
-0.241193,
0.278416,
-0.343345,
0.238777,
-0.368831,
0.207115
],
[
-0.94921,
-0.188891,
0.084727,
0.082328,
-0.073629,
-0.362053,
0.178909,
-0.017969
],
[
0.634576,
-0.60835,
-0.643282,
-0.066351,
-0.1109,
0.023806,
0.356232,
0.176553
]
],
"network.2.bias": [
0.066929,
-0.327092,
0.509466,
0.19096,
0.475078,
0.004501,
-0.310928,
-0.240431
],
"network.4.weight": [
[
0.220232,
-0.088,
-0.043844,
0.349329,
0.184495,
0.381737,
0.143931,
0.406738
],
[
0.035783,
0.130498,
0.013354,
-0.686245,
0.767861,
0.108578,
-0.666086,
-0.026062
],
[
0.325571,
0.307256,
0.30382,
-0.254232,
0.576663,
-0.254364,
-0.178388,
-0.357497
],
[
-0.865198,
0.339242,
-0.531385,
-0.213226,
0.005232,
0.066059,
-0.493174,
-0.013059
],
[
0.492832,
-0.010955,
0.638096,
-0.592798,
0.78522,
0.107987,
0.122315,
-0.703075
],
[
0.361038,
0.09987,
0.152767,
-0.776954,
0.605123,
0.374637,
-0.031007,
-0.514455
],
[
0.179606,
-0.475072,
0.43376,
-0.42693,
0.519518,
0.190897,
0.031249,
-0.509276
],
[
-0.463677,
-0.213152,
-0.239207,
-0.42198,
-0.389708,
0.099451,
-0.323992,
0.126208
]
],
"network.4.bias": [
0.39288,
0.114244,
-0.194066,
-0.390658,
-0.043897,
0.105868,
0.102463,
-0.144831
],
"network.6.weight": [
[
-0.531674,
-0.527558,
-0.207455,
-0.389337,
-0.333734,
-0.080054,
-0.360297,
0.058745
],
[
-0.155713,
0.109399,
0.014098,
-0.285022,
-0.105249,
0.052354,
0.058978,
-0.105466
],
[
-0.547085,
-0.084852,
-0.167855,
-0.022197,
0.113126,
-0.083069,
-0.195918,
-0.053955
],
[
-0.140995,
0.672546,
0.29224,
0.14679,
0.745853,
0.592183,
0.205976,
0.421515
],
[
-0.080936,
0.389367,
0.73108,
-0.146069,
0.451584,
0.775655,
0.332484,
-0.026014
],
[
0.034227,
0.339812,
0.709226,
0.178475,
0.770902,
0.481875,
0.636269,
0.513392
],
[
0.392335,
0.109451,
-0.543195,
-0.025525,
-0.051505,
-0.457317,
-0.421596,
-0.533145
],
[
0.374183,
0.324197,
-0.220897,
-0.028952,
-0.247479,
-0.498071,
-0.186189,
0.069614
]
],
"network.6.bias": [
-0.099903,
-0.489081,
-0.303908,
-0.042962,
-0.279009,
-0.24011,
0.634162,
0.520339
],
"network.8.weight": [
[
0.029204,
-0.136619,
0.067914,
-0.467167,
-0.234478,
-0.414316,
0.643976,
0.46118
]
],
"network.8.bias": [
0.606936
]
}
## Activation Signature
### 0
mean: [-0.349496, 0.869958, 1.086323, -0.342604, -3.983193, -0.086849, 0.617154, -0.267840]
std: [1.547504, 1.261258, 2.252962, 1.998992, 2.673750, 1.694431, 1.558184, 1.604016]
### 2
mean: [1.616312, -1.380903, 1.723426, 0.386708, 2.271594, 0.141723, -0.805760, -1.108474]
std: [1.960400, 0.809442, 2.048407, 1.793177, 3.119458, 0.766242, 1.425582, 1.673951]
### 4
mean: [1.665282, 1.458605, 1.944999, -2.965130, 3.295928, 1.837434, 2.096758, -2.687882]
std: [0.705559, 2.966180, 2.968405, 2.504218, 4.829532, 3.379042, 3.053724, 2.114586]
### 6
mean: [-4.783035, -0.619019, -1.997565, 6.366692, 6.258611, 7.544556, -1.948124, -1.218312]
std: [4.841861, 0.113584, 1.263921, 7.913437, 8.046103, 9.365412, 3.730514, 2.681009]
### 8
mean: [-6.540194]
std: [9.808219]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
ends_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.388896,
-0.076341,
0.102071,
-0.498515,
0.345834
],
[
0.464564,
-0.109712,
-0.195431,
-0.095248,
0.43856
],
[
-0.502554,
-0.244549,
0.164886,
0.220243,
1.050065
],
[
-0.296963,
0.174377,
0.066301,
0.168926,
-0.991912
],
[
-0.700244,
-0.53192,
-0.552613,
-0.054345,
-0.223
],
[
0.739396,
-0.112262,
-0.078898,
-0.384687,
0.031831
],
[
-0.440265,
0.451636,
0.252654,
0.228348,
-0.476928
],
[
-0.101593,
-0.237787,
0.383359,
0.095351,
-0.84944
]
],
"network.0.bias": [
-0.265049,
0.563351,
0.045538,
0.431804,
-0.596582,
0.14807,
-0.087178,
0.334263
],
"network.2.weight": [
[
0.000611,
0.27097,
0.958709,
-0.068399,
0.316792,
-0.270839,
0.3047,
-0.376894
],
[
0.146379,
-0.028965,
-0.337845,
0.072932,
-0.211553,
0.014745,
-0.460602,
-0.484833
],
[
0.097868,
0.320699,
0.769431,
-0.220845,
0.078783,
0.335818,
-0.241861,
-0.224117
],
[
-0.129189,
-0.095569,
-0.319079,
0.373867,
0.078455,
-0.396071,
0.582442,
0.612767
],
[
0.115895,
0.729566,
0.935262,
-0.678243,
0.353549,
0.670145,
-0.284442,
-0.162325
],
[
-0.17055,
0.492571,
-0.241193,
0.278416,
-0.343345,
0.238777,
-0.368831,
0.207115
],
[
-0.94921,
-0.188891,
0.084727,
0.082328,
-0.073629,
-0.362053,
0.178909,
-0.017969
],
[
0.634576,
-0.60835,
-0.643282,
-0.066351,
-0.1109,
0.023806,
0.356232,
0.176553
]
],
"network.2.bias": [
0.066929,
-0.327092,
0.509466,
0.19096,
0.475078,
0.004501,
-0.310928,
-0.240431
],
"network.4.weight": [
[
0.220232,
-0.088,
-0.043844,
0.349329,
0.184495,
0.381737,
0.143931,
0.406738
],
[
0.035783,
0.130498,
0.013354,
-0.686245,
0.767861,
0.108578,
-0.666086,
-0.026062
],
[
0.325571,
0.307256,
0.30382,
-0.254232,
0.576663,
-0.254364,
-0.178388,
-0.357497
],
[
-0.865198,
0.339242,
-0.531385,
-0.213226,
0.005232,
0.066059,
-0.493174,
-0.013059
],
[
0.492832,
-0.010955,
0.638096,
-0.592798,
0.78522,
0.107987,
0.122315,
-0.703075
],
[
0.361038,
0.09987,
0.152767,
-0.776954,
0.605123,
0.374637,
-0.031007,
-0.514455
],
[
0.179606,
-0.475072,
0.43376,
-0.42693,
0.519518,
0.190897,
0.031249,
-0.509276
],
[
-0.463677,
-0.213152,
-0.239207,
-0.42198,
-0.389708,
0.099451,
-0.323992,
0.126208
]
],
"network.4.bias": [
0.39288,
0.114244,
-0.194066,
-0.390658,
-0.043897,
0.105868,
0.102463,
-0.144831
],
"network.6.weight": [
[
-0.531674,
-0.527558,
-0.207455,
-0.389337,
-0.333734,
-0.080054,
-0.360297,
0.058745
],
[
-0.155713,
0.109399,
0.014098,
-0.285022,
-0.105249,
0.052354,
0.058978,
-0.105466
],
[
-0.547085,
-0.084852,
-0.167855,
-0.022197,
0.113126,
-0.083069,
-0.195918,
-0.053955
],
[
-0.140995,
0.672546,
0.29224,
0.14679,
0.745853,
0.592183,
0.205976,
0.421515
],
[
-0.080936,
0.389367,
0.73108,
-0.146069,
0.451584,
0.775655,
0.332484,
-0.026014
],
[
0.034227,
0.339812,
0.709226,
0.178475,
0.770902,
0.481875,
0.636269,
0.513392
],
[
0.392335,
0.109451,
-0.543195,
-0.025525,
-0.051505,
-0.457317,
-0.421596,
-0.533145
],
[
0.374183,
0.324197,
-0.220897,
-0.028952,
-0.247479,
-0.498071,
-0.186189,
0.069614
]
],
"network.6.bias": [
-0.099903,
-0.489081,
-0.303908,
-0.042962,
-0.279009,
-0.24011,
0.634162,
0.520339
],
"network.8.weight": [
[
0.029204,
-0.136619,
0.067914,
-0.467167,
-0.234478,
-0.414316,
0.643976,
0.46118
]
],
"network.8.bias": [
0.606936
]
}
## Activation Signature
### 0
mean: [-0.349496, 0.869958, 1.086323, -0.342604, -3.983193, -0.086849, 0.617154, -0.267840]
std: [1.547504, 1.261258, 2.252962, 1.998992, 2.673750, 1.694431, 1.558184, 1.604016]
### 2
mean: [1.616312, -1.380903, 1.723426, 0.386708, 2.271594, 0.141723, -0.805760, -1.108474]
std: [1.960400, 0.809442, 2.048407, 1.793177, 3.119458, 0.766242, 1.425582, 1.673951]
### 4
mean: [1.665282, 1.458605, 1.944999, -2.965130, 3.295928, 1.837434, 2.096758, -2.687882]
std: [0.705559, 2.966180, 2.968405, 2.504218, 4.829532, 3.379042, 3.053724, 2.114586]
### 6
mean: [-4.783035, -0.619019, -1.997565, 6.366692, 6.258611, 7.544556, -1.948124, -1.218312]
std: [4.841861, 0.113584, 1.263921, 7.913437, 8.046103, 9.365412, 3.730514, 2.681009]
### 8
mean: [-6.540194]
std: [9.808219]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
ends_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -0.34949609637260437, "std": 1.547504186630249}, "1": {"mean": 0.8699579238891602, "std": 1.2612584829330444}, "2": {"mean": 1.0863226652145386, "std": 2.2529616355895996}, "3": {"mean": -0.34260427951812744, "std": 1.99899160861969}, "4": {"mean": -3.9831926822662354, "std": 2.6737496852874756}, "5": {"mean": -0.08684898912906647, "std": 1.694431185722351}, "6": {"mean": 0.6171543598175049, "std": 1.5581836700439453}, "7": {"mean": -0.2678397297859192, "std": 1.6040157079696655}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 1.61631178855896, "std": 1.9604002237319946}, "1": {"mean": -1.3809034824371338, "std": 0.8094421625137329}, "2": {"mean": 1.7234264612197876, "std": 2.0484070777893066}, "3": {"mean": 0.38670799136161804, "std": 1.7931770086288452}, "4": {"mean": 2.2715938091278076, "std": 3.1194581985473633}, "5": {"mean": 0.14172285795211792, "std": 0.7662420272827148}, "6": {"mean": -0.8057602047920227, "std": 1.4255824089050293}, "7": {"mean": -1.1084742546081543, "std": 1.6739510297775269}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 1.6652824878692627, "std": 0.7055588364601135}, "1": {"mean": 1.458605408668518, "std": 2.9661800861358643}, "2": {"mean": 1.9449986219406128, "std": 2.96840500831604}, "3": {"mean": -2.9651296138763428, "std": 2.5042178630828857}, "4": {"mean": 3.2959275245666504, "std": 4.829532146453857}, "5": {"mean": 1.8374344110488892, "std": 3.3790416717529297}, "6": {"mean": 2.096757650375366, "std": 3.0537238121032715}, "7": {"mean": -2.6878819465637207, "std": 2.1145858764648438}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -4.7830352783203125, "std": 4.841860771179199}, "1": {"mean": -0.6190194487571716, "std": 0.1135844811797142}, "2": {"mean": -1.9975651502609253, "std": 1.2639206647872925}, "3": {"mean": 6.366691589355469, "std": 7.913437366485596}, "4": {"mean": 6.25861120223999, "std": 8.046102523803711}, "5": {"mean": 7.544556140899658, "std": 9.365411758422852}, "6": {"mean": -1.9481244087219238, "std": 3.730513572692871}, "7": {"mean": -1.2183117866516113, "std": 2.6810085773468018}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -6.540193557739258, "std": 9.808218955993652}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.388896, -0.076341, 0.102071, -0.498515, 0.345834], [0.464564, -0.109712, -0.195431, -0.095248, 0.43856], [-0.502554, -0.244549, 0.164886, 0.220243, 1.050065], [-0.296963, 0.174377, 0.066301, 0.168926, -0.991912], [-0.700244, -0.53192, -0.552613, -0.054345, -0.223], [0.739396, -0.112262, -0.078898, -0.384687, 0.031831], [-0.440265, 0.451636, 0.252654, 0.228348, -0.476928], [-0.101593, -0.237787, 0.383359, 0.095351, -0.84944]], "network.0.bias": [-0.265049, 0.563351, 0.045538, 0.431804, -0.596582, 0.14807, -0.087178, 0.334263], "network.2.weight": [[0.000611, 0.27097, 0.958709, -0.068399, 0.316792, -0.270839, 0.3047, -0.376894], [0.146379, -0.028965, -0.337845, 0.072932, -0.211553, 0.014745, -0.460602, -0.484833], [0.097868, 0.320699, 0.769431, -0.220845, 0.078783, 0.335818, -0.241861, -0.224117], [-0.129189, -0.095569, -0.319079, 0.373867, 0.078455, -0.396071, 0.582442, 0.612767], [0.115895, 0.729566, 0.935262, -0.678243, 0.353549, 0.670145, -0.284442, -0.162325], [-0.17055, 0.492571, -0.241193, 0.278416, -0.343345, 0.238777, -0.368831, 0.207115], [-0.94921, -0.188891, 0.084727, 0.082328, -0.073629, -0.362053, 0.178909, -0.017969], [0.634576, -0.60835, -0.643282, -0.066351, -0.1109, 0.023806, 0.356232, 0.176553]], "network.2.bias": [0.066929, -0.327092, 0.509466, 0.19096, 0.475078, 0.004501, -0.310928, -0.240431], "network.4.weight": [[0.220232, -0.088, -0.043844, 0.349329, 0.184495, 0.381737, 0.143931, 0.406738], [0.035783, 0.130498, 0.013354, -0.686245, 0.767861, 0.108578, -0.666086, -0.026062], [0.325571, 0.307256, 0.30382, -0.254232, 0.576663, -0.254364, -0.178388, -0.357497], [-0.865198, 0.339242, -0.531385, -0.213226, 0.005232, 0.066059, -0.493174, -0.013059], [0.492832, -0.010955, 0.638096, -0.592798, 0.78522, 0.107987, 0.122315, -0.703075], [0.361038, 0.09987, 0.152767, -0.776954, 0.605123, 0.374637, -0.031007, -0.514455], [0.179606, -0.475072, 0.43376, -0.42693, 0.519518, 0.190897, 0.031249, -0.509276], [-0.463677, -0.213152, -0.239207, -0.42198, -0.389708, 0.099451, -0.323992, 0.126208]], "network.4.bias": [0.39288, 0.114244, -0.194066, -0.390658, -0.043897, 0.105868, 0.102463, -0.144831], "network.6.weight": [[-0.531674, -0.527558, -0.207455, -0.389337, -0.333734, -0.080054, -0.360297, 0.058745], [-0.155713, 0.109399, 0.014098, -0.285022, -0.105249, 0.052354, 0.058978, -0.105466], [-0.547085, -0.084852, -0.167855, -0.022197, 0.113126, -0.083069, -0.195918, -0.053955], [-0.140995, 0.672546, 0.29224, 0.14679, 0.745853, 0.592183, 0.205976, 0.421515], [-0.080936, 0.389367, 0.73108, -0.146069, 0.451584, 0.775655, 0.332484, -0.026014], [0.034227, 0.339812, 0.709226, 0.178475, 0.770902, 0.481875, 0.636269, 0.513392], [0.392335, 0.109451, -0.543195, -0.025525, -0.051505, -0.457317, -0.421596, -0.533145], [0.374183, 0.324197, -0.220897, -0.028952, -0.247479, -0.498071, -0.186189, 0.069614]], "network.6.bias": [-0.099903, -0.489081, -0.303908, -0.042962, -0.279009, -0.24011, 0.634162, 0.520339], "network.8.weight": [[0.029204, -0.136619, 0.067914, -0.467167, -0.234478, -0.414316, 0.643976, 0.46118]], "network.8.bias": [0.606936]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7116595208644867, "train_acc": 0.43, "val_loss": 0.6481384038925171, "val_acc": 0.58}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6410282850265503, "train_acc": 0.56, "val_loss": 0.5449063181877136, "val_acc": 0.58}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.598608523607254, "train_acc": 0.585, "val_loss": 0.4776480197906494, "val_acc": 0.74}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.47113656997680664, "train_acc": 0.835, "val_loss": 0.4435732960700989, "val_acc": 0.76}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.32947850227355957, "train_acc": 0.89, "val_loss": 0.5477588772773743, "val_acc": 0.76}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.31863756477832794, "train_acc": 0.865, "val_loss": 0.5908229351043701, "val_acc": 0.78}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.27146774530410767, "train_acc": 0.91, "val_loss": 0.6722450256347656, "val_acc": 0.78}], "summary": {"total_epochs": 7, "degraded_epochs": 2, "improved_epochs": 5, "patterns": ["ends_with"], "degraded_stage": {"initial_val_loss": 0.6481384038925171, "final_val_loss": 0.5449063181877136, "initial_val_acc": 0.58, "final_val_acc": 0.58, "best_val_acc": 0.58}, "improved_stage": {"initial_val_loss": 0.4776480197906494, "final_val_loss": 0.6722450256347656, "initial_val_acc": 0.74, "final_val_acc": 0.78, "best_val_acc": 0.78, "best_epoch": 5}, "improvement": 0.20000000000000007, "first_improvement_epoch": 1}}
|
41
|
{"target_pattern": "ends_with", "degraded_accuracy": 0.58, "improved_accuracy": 0.96, "improvement": 0.38, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 2336, "learning_rate": 0.027281424600688553, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "ends_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["ends_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.161891,
0.084214,
0.404687,
0.527906,
-0.396693
],
[
-0.20482,
0.278678,
-0.16929,
-0.349265,
-0.027681
],
[
-0.807691,
0.034478,
0.106431,
-0.043396,
0.342171
],
[
-0.870324,
0.012155,
-0.160704,
-0.109193,
0.529424
],
[
-0.382691,
0.185787,
-0.141041,
0.115818,
0.727171
],
[
-0.838261,
-0.069857,
0.225162,
0.292673,
0.897909
]
],
"network.0.bias": [
0.400435,
-0.056291,
0.486324,
-0.204332,
-0.167437,
-0.098703
],
"network.2.weight": [
[
0.606441,
0.144371,
-0.852941,
0.675367,
-0.280519,
-0.381997
],
[
-0.027824,
-0.137746,
0.56655,
-0.51932,
-0.392655,
-0.384969
],
[
0.325119,
0.573242,
-0.793979,
0.015466,
-0.314606,
-0.051963
],
[
-0.168885,
-0.136381,
0.092945,
0.121444,
0.106198,
-0.229145
],
[
0.30667,
-0.736034,
-0.20524,
0.197811,
0.254617,
0.33279
],
[
0.317989,
-0.605427,
0.365937,
-0.221988,
0.520061,
0.889419
]
],
"network.2.bias": [
-0.026529,
0.111767,
-0.037425,
-0.485308,
0.576923,
0.636859
],
"network.4.weight": [
[
0.537771,
-0.178742,
0.328123,
-0.028806,
-0.23423,
-0.291971
],
[
0.428908,
0.059817,
0.201126,
-0.121302,
-0.098702,
-0.111064
],
[
-0.685329,
-0.159625,
-0.562791,
-0.540772,
0.390455,
0.66843
],
[
0.758317,
-0.194615,
0.431847,
-0.113471,
0.349346,
-0.047374
],
[
0.286021,
0.021273,
0.275088,
0.387856,
0.235597,
-0.157561
],
[
-0.459628,
0.640747,
0.055512,
0.129653,
-0.103073,
-0.128919
]
],
"network.4.bias": [
0.299405,
0.062536,
0.483655,
-0.420226,
0.141606,
-0.188853
],
"network.6.weight": [
[
-0.050606,
0.033346,
0.489739,
-0.445478,
-0.361718,
-0.200463
],
[
0.104111,
-0.126951,
-0.319597,
0.559602,
-0.178485,
0.0669
],
[
-0.608401,
-0.232899,
0.963613,
-0.458384,
-0.044295,
-0.446864
],
[
0.445948,
0.522425,
-0.214875,
0.020132,
0.021515,
-0.630876
],
[
-0.070938,
0.506386,
-0.108415,
0.465619,
0.559359,
0.103147
],
[
-0.381692,
-0.434544,
0.734732,
-0.388582,
0.04621,
-0.451192
]
],
"network.6.bias": [
0.375404,
0.260306,
0.238221,
-0.180953,
0.213652,
0.17467
],
"network.8.weight": [
[
-0.275484,
0.330605,
-0.73679,
0.031502,
0.328874,
-0.80528
]
],
"network.8.bias": [
0.301866
]
}
## Activation Signature
### 0
mean: [2.243514, -0.940111, 0.095296, -1.182149, 0.545463, 0.936212]
std: [1.508209, 0.840105, 1.583901, 1.918419, 1.484101, 2.348956]
### 2
mean: [0.199019, -0.495835, -0.091353, -1.014150, 1.847717, 3.125070]
std: [1.580635, 1.085718, 1.189757, 0.356342, 0.960116, 2.270084]
### 4
mean: [-0.595031, -0.120404, 2.738533, 0.643531, 0.265491, -1.046784]
std: [1.108646, 0.584857, 2.162040, 0.907767, 0.462737, 0.554707]
### 6
mean: [1.355697, -0.303319, 2.542747, -0.613313, 0.353536, 1.927644]
std: [1.348001, 0.883982, 2.436281, 0.688919, 0.842116, 1.893321]
### 8
mean: [-3.493510]
std: [3.635813]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
ends_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.161891,
0.084214,
0.404687,
0.527906,
-0.396693
],
[
-0.20482,
0.278678,
-0.16929,
-0.349265,
-0.027681
],
[
-0.807691,
0.034478,
0.106431,
-0.043396,
0.342171
],
[
-0.870324,
0.012155,
-0.160704,
-0.109193,
0.529424
],
[
-0.382691,
0.185787,
-0.141041,
0.115818,
0.727171
],
[
-0.838261,
-0.069857,
0.225162,
0.292673,
0.897909
]
],
"network.0.bias": [
0.400435,
-0.056291,
0.486324,
-0.204332,
-0.167437,
-0.098703
],
"network.2.weight": [
[
0.606441,
0.144371,
-0.852941,
0.675367,
-0.280519,
-0.381997
],
[
-0.027824,
-0.137746,
0.56655,
-0.51932,
-0.392655,
-0.384969
],
[
0.325119,
0.573242,
-0.793979,
0.015466,
-0.314606,
-0.051963
],
[
-0.168885,
-0.136381,
0.092945,
0.121444,
0.106198,
-0.229145
],
[
0.30667,
-0.736034,
-0.20524,
0.197811,
0.254617,
0.33279
],
[
0.317989,
-0.605427,
0.365937,
-0.221988,
0.520061,
0.889419
]
],
"network.2.bias": [
-0.026529,
0.111767,
-0.037425,
-0.485308,
0.576923,
0.636859
],
"network.4.weight": [
[
0.537771,
-0.178742,
0.328123,
-0.028806,
-0.23423,
-0.291971
],
[
0.428908,
0.059817,
0.201126,
-0.121302,
-0.098702,
-0.111064
],
[
-0.685329,
-0.159625,
-0.562791,
-0.540772,
0.390455,
0.66843
],
[
0.758317,
-0.194615,
0.431847,
-0.113471,
0.349346,
-0.047374
],
[
0.286021,
0.021273,
0.275088,
0.387856,
0.235597,
-0.157561
],
[
-0.459628,
0.640747,
0.055512,
0.129653,
-0.103073,
-0.128919
]
],
"network.4.bias": [
0.299405,
0.062536,
0.483655,
-0.420226,
0.141606,
-0.188853
],
"network.6.weight": [
[
-0.050606,
0.033346,
0.489739,
-0.445478,
-0.361718,
-0.200463
],
[
0.104111,
-0.126951,
-0.319597,
0.559602,
-0.178485,
0.0669
],
[
-0.608401,
-0.232899,
0.963613,
-0.458384,
-0.044295,
-0.446864
],
[
0.445948,
0.522425,
-0.214875,
0.020132,
0.021515,
-0.630876
],
[
-0.070938,
0.506386,
-0.108415,
0.465619,
0.559359,
0.103147
],
[
-0.381692,
-0.434544,
0.734732,
-0.388582,
0.04621,
-0.451192
]
],
"network.6.bias": [
0.375404,
0.260306,
0.238221,
-0.180953,
0.213652,
0.17467
],
"network.8.weight": [
[
-0.275484,
0.330605,
-0.73679,
0.031502,
0.328874,
-0.80528
]
],
"network.8.bias": [
0.301866
]
}
## Activation Signature
### 0
mean: [2.243514, -0.940111, 0.095296, -1.182149, 0.545463, 0.936212]
std: [1.508209, 0.840105, 1.583901, 1.918419, 1.484101, 2.348956]
### 2
mean: [0.199019, -0.495835, -0.091353, -1.014150, 1.847717, 3.125070]
std: [1.580635, 1.085718, 1.189757, 0.356342, 0.960116, 2.270084]
### 4
mean: [-0.595031, -0.120404, 2.738533, 0.643531, 0.265491, -1.046784]
std: [1.108646, 0.584857, 2.162040, 0.907767, 0.462737, 0.554707]
### 6
mean: [1.355697, -0.303319, 2.542747, -0.613313, 0.353536, 1.927644]
std: [1.348001, 0.883982, 2.436281, 0.688919, 0.842116, 1.893321]
### 8
mean: [-3.493510]
std: [3.635813]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
ends_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 2.243513584136963, "std": 1.508209228515625}, "1": {"mean": -0.9401110410690308, "std": 0.8401045799255371}, "2": {"mean": 0.09529591351747513, "std": 1.583900809288025}, "3": {"mean": -1.1821486949920654, "std": 1.9184192419052124}, "4": {"mean": 0.5454629063606262, "std": 1.4841011762619019}, "5": {"mean": 0.936212420463562, "std": 2.3489556312561035}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.19901923835277557, "std": 1.5806349515914917}, "1": {"mean": -0.4958352744579315, "std": 1.0857176780700684}, "2": {"mean": -0.09135258197784424, "std": 1.1897571086883545}, "3": {"mean": -1.0141499042510986, "std": 0.35634177923202515}, "4": {"mean": 1.8477174043655396, "std": 0.9601161479949951}, "5": {"mean": 3.1250696182250977, "std": 2.2700841426849365}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.5950312614440918, "std": 1.1086456775665283}, "1": {"mean": -0.12040448933839798, "std": 0.5848572254180908}, "2": {"mean": 2.7385334968566895, "std": 2.1620395183563232}, "3": {"mean": 0.6435312628746033, "std": 0.9077667593955994}, "4": {"mean": 0.2654910087585449, "std": 0.46273717284202576}, "5": {"mean": -1.0467841625213623, "std": 0.5547065138816833}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 1.3556969165802002, "std": 1.3480005264282227}, "1": {"mean": -0.3033188581466675, "std": 0.8839824199676514}, "2": {"mean": 2.5427465438842773, "std": 2.4362809658050537}, "3": {"mean": -0.6133133172988892, "std": 0.688919186592102}, "4": {"mean": 0.35353633761405945, "std": 0.8421156406402588}, "5": {"mean": 1.9276443719863892, "std": 1.8933212757110596}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -3.4935100078582764, "std": 3.635812520980835}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.161891, 0.084214, 0.404687, 0.527906, -0.396693], [-0.20482, 0.278678, -0.16929, -0.349265, -0.027681], [-0.807691, 0.034478, 0.106431, -0.043396, 0.342171], [-0.870324, 0.012155, -0.160704, -0.109193, 0.529424], [-0.382691, 0.185787, -0.141041, 0.115818, 0.727171], [-0.838261, -0.069857, 0.225162, 0.292673, 0.897909]], "network.0.bias": [0.400435, -0.056291, 0.486324, -0.204332, -0.167437, -0.098703], "network.2.weight": [[0.606441, 0.144371, -0.852941, 0.675367, -0.280519, -0.381997], [-0.027824, -0.137746, 0.56655, -0.51932, -0.392655, -0.384969], [0.325119, 0.573242, -0.793979, 0.015466, -0.314606, -0.051963], [-0.168885, -0.136381, 0.092945, 0.121444, 0.106198, -0.229145], [0.30667, -0.736034, -0.20524, 0.197811, 0.254617, 0.33279], [0.317989, -0.605427, 0.365937, -0.221988, 0.520061, 0.889419]], "network.2.bias": [-0.026529, 0.111767, -0.037425, -0.485308, 0.576923, 0.636859], "network.4.weight": [[0.537771, -0.178742, 0.328123, -0.028806, -0.23423, -0.291971], [0.428908, 0.059817, 0.201126, -0.121302, -0.098702, -0.111064], [-0.685329, -0.159625, -0.562791, -0.540772, 0.390455, 0.66843], [0.758317, -0.194615, 0.431847, -0.113471, 0.349346, -0.047374], [0.286021, 0.021273, 0.275088, 0.387856, 0.235597, -0.157561], [-0.459628, 0.640747, 0.055512, 0.129653, -0.103073, -0.128919]], "network.4.bias": [0.299405, 0.062536, 0.483655, -0.420226, 0.141606, -0.188853], "network.6.weight": [[-0.050606, 0.033346, 0.489739, -0.445478, -0.361718, -0.200463], [0.104111, -0.126951, -0.319597, 0.559602, -0.178485, 0.0669], [-0.608401, -0.232899, 0.963613, -0.458384, -0.044295, -0.446864], [0.445948, 0.522425, -0.214875, 0.020132, 0.021515, -0.630876], [-0.070938, 0.506386, -0.108415, 0.465619, 0.559359, 0.103147], [-0.381692, -0.434544, 0.734732, -0.388582, 0.04621, -0.451192]], "network.6.bias": [0.375404, 0.260306, 0.238221, -0.180953, 0.213652, 0.17467], "network.8.weight": [[-0.275484, 0.330605, -0.73679, 0.031502, 0.328874, -0.80528]], "network.8.bias": [0.301866]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7025384902954102, "train_acc": 0.435, "val_loss": 0.6880980730056763, "val_acc": 0.58}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6831613183021545, "train_acc": 0.565, "val_loss": 0.6696557402610779, "val_acc": 0.58}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6688477694988251, "train_acc": 0.565, "val_loss": 0.6505634188652039, "val_acc": 0.58}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.684164822101593, "train_acc": 0.48, "val_loss": 0.6507313251495361, "val_acc": 0.58}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.6509829163551331, "train_acc": 0.655, "val_loss": 0.6286749839782715, "val_acc": 0.72}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.600348711013794, "train_acc": 0.735, "val_loss": 0.5636680722236633, "val_acc": 0.72}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.5480456352233887, "train_acc": 0.775, "val_loss": 0.4918942153453827, "val_acc": 0.74}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.4856690317392349, "train_acc": 0.805, "val_loss": 0.4507330358028412, "val_acc": 0.76}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.42754825949668884, "train_acc": 0.8, "val_loss": 0.4267332851886749, "val_acc": 0.78}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.4008858650922775, "train_acc": 0.82, "val_loss": 0.2940838038921356, "val_acc": 0.84}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.36625224351882935, "train_acc": 0.875, "val_loss": 0.3011467456817627, "val_acc": 0.84}, {"stage": "improved", "epoch": 8, "global_epoch": 11, "train_loss": 0.3597070723772049, "train_acc": 0.84, "val_loss": 0.21949610114097595, "val_acc": 0.94}, {"stage": "improved", "epoch": 9, "global_epoch": 12, "train_loss": 0.301668256521225, "train_acc": 0.89, "val_loss": 0.17210653424263, "val_acc": 0.96}], "summary": {"total_epochs": 13, "degraded_epochs": 3, "improved_epochs": 10, "patterns": ["ends_with"], "degraded_stage": {"initial_val_loss": 0.6880980730056763, "final_val_loss": 0.6505634188652039, "initial_val_acc": 0.58, "final_val_acc": 0.58, "best_val_acc": 0.58}, "improved_stage": {"initial_val_loss": 0.6507313251495361, "final_val_loss": 0.17210653424263, "initial_val_acc": 0.58, "final_val_acc": 0.96, "best_val_acc": 0.96, "best_epoch": 12}, "improvement": 0.38, "first_improvement_epoch": 2}}
|
42
|
{"target_pattern": "palindrome", "degraded_accuracy": 0.56, "improved_accuracy": 0.92, "improvement": 0.36, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 7385, "learning_rate": 0.03768603114736421, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "palindrome", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["palindrome"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.447508,
-0.104747,
0.018937,
0.231872,
0.704269
],
[
-0.173712,
-0.26278,
-0.070731,
0.436484,
0.631075
],
[
-0.338535,
0.475308,
0.137348,
0.371907,
-0.5224
],
[
0.51975,
-0.165959,
-0.206193,
0.188862,
-0.252589
],
[
-0.11849,
-0.239171,
-0.315832,
-0.046167,
-0.097884
]
],
"network.0.bias": [
-0.043419,
0.075723,
-0.208389,
0.262787,
0.186463
],
"network.2.weight": [
[
-0.256541,
0.092051,
-0.25534,
-0.343807,
0.020394
],
[
0.584801,
0.571732,
-0.095973,
0.1035,
-0.336275
],
[
-0.582634,
-0.15031,
-0.368369,
-0.275322,
-0.558524
],
[
-0.351899,
-0.504534,
0.352491,
-0.298861,
-0.146093
],
[
-0.429075,
0.047115,
-0.332672,
-0.414905,
-0.291928
]
],
"network.2.bias": [
-0.325727,
0.325202,
0.106962,
0.236476,
-0.32606
],
"network.4.weight": [
[
-0.361781,
0.524852,
-0.160067,
0.476246,
-0.209499
],
[
-0.179615,
0.463771,
0.01837,
-0.188869,
0.119766
],
[
-0.180594,
0.458482,
-0.019179,
-0.447905,
0.207805
],
[
0.085771,
-0.314725,
-0.065762,
0.00865,
0.161807
],
[
0.00417,
-0.318865,
0.444376,
-0.282966,
0.00196
]
],
"network.4.bias": [
-0.269668,
-0.149247,
0.165218,
-0.117863,
-0.312257
],
"network.6.weight": [
[
0.171528,
-0.35608,
0.145199,
-0.345107,
-0.096482
],
[
0.596662,
0.173187,
0.382954,
0.296076,
0.173969
],
[
-0.131153,
0.415486,
0.704282,
-0.125659,
0.225347
],
[
0.055814,
0.319813,
0.157903,
0.311243,
-0.037803
],
[
0.4315,
-0.309584,
-0.217599,
0.450864,
0.21308
]
],
"network.6.bias": [
-0.39091,
0.23747,
0.267817,
0.092252,
0.247411
],
"network.8.weight": [
[
0.02361,
-0.443902,
-0.332722,
-0.08013,
0.493283
]
],
"network.8.bias": [
0.240494
]
}
## Activation Signature
### 0
mean: [1.728001, 0.947295, 0.676586, 0.269937, -1.276279]
std: [1.794214, 1.605603, 1.686549, 1.018323, 1.024635]
### 2
mean: [-1.110998, 1.951174, -1.613205, -0.763390, -1.586412]
std: [0.537016, 1.732391, 1.186077, 1.408365, 0.832483]
### 4
mean: [0.841602, 0.721282, 0.978087, -0.730463, -0.986227]
std: [0.844266, 0.837164, 0.881479, 0.546568, 0.513578]
### 6
mean: [-0.368834, 1.257236, 1.176308, 0.539136, 0.161006]
std: [0.026896, 0.956338, 0.823352, 0.435823, 0.113015]
### 8
mean: [-0.672759]
std: [0.769901]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
palindrome
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.447508,
-0.104747,
0.018937,
0.231872,
0.704269
],
[
-0.173712,
-0.26278,
-0.070731,
0.436484,
0.631075
],
[
-0.338535,
0.475308,
0.137348,
0.371907,
-0.5224
],
[
0.51975,
-0.165959,
-0.206193,
0.188862,
-0.252589
],
[
-0.11849,
-0.239171,
-0.315832,
-0.046167,
-0.097884
]
],
"network.0.bias": [
-0.043419,
0.075723,
-0.208389,
0.262787,
0.186463
],
"network.2.weight": [
[
-0.256541,
0.092051,
-0.25534,
-0.343807,
0.020394
],
[
0.584801,
0.571732,
-0.095973,
0.1035,
-0.336275
],
[
-0.582634,
-0.15031,
-0.368369,
-0.275322,
-0.558524
],
[
-0.351899,
-0.504534,
0.352491,
-0.298861,
-0.146093
],
[
-0.429075,
0.047115,
-0.332672,
-0.414905,
-0.291928
]
],
"network.2.bias": [
-0.325727,
0.325202,
0.106962,
0.236476,
-0.32606
],
"network.4.weight": [
[
-0.361781,
0.524852,
-0.160067,
0.476246,
-0.209499
],
[
-0.179615,
0.463771,
0.01837,
-0.188869,
0.119766
],
[
-0.180594,
0.458482,
-0.019179,
-0.447905,
0.207805
],
[
0.085771,
-0.314725,
-0.065762,
0.00865,
0.161807
],
[
0.00417,
-0.318865,
0.444376,
-0.282966,
0.00196
]
],
"network.4.bias": [
-0.269668,
-0.149247,
0.165218,
-0.117863,
-0.312257
],
"network.6.weight": [
[
0.171528,
-0.35608,
0.145199,
-0.345107,
-0.096482
],
[
0.596662,
0.173187,
0.382954,
0.296076,
0.173969
],
[
-0.131153,
0.415486,
0.704282,
-0.125659,
0.225347
],
[
0.055814,
0.319813,
0.157903,
0.311243,
-0.037803
],
[
0.4315,
-0.309584,
-0.217599,
0.450864,
0.21308
]
],
"network.6.bias": [
-0.39091,
0.23747,
0.267817,
0.092252,
0.247411
],
"network.8.weight": [
[
0.02361,
-0.443902,
-0.332722,
-0.08013,
0.493283
]
],
"network.8.bias": [
0.240494
]
}
## Activation Signature
### 0
mean: [1.728001, 0.947295, 0.676586, 0.269937, -1.276279]
std: [1.794214, 1.605603, 1.686549, 1.018323, 1.024635]
### 2
mean: [-1.110998, 1.951174, -1.613205, -0.763390, -1.586412]
std: [0.537016, 1.732391, 1.186077, 1.408365, 0.832483]
### 4
mean: [0.841602, 0.721282, 0.978087, -0.730463, -0.986227]
std: [0.844266, 0.837164, 0.881479, 0.546568, 0.513578]
### 6
mean: [-0.368834, 1.257236, 1.176308, 0.539136, 0.161006]
std: [0.026896, 0.956338, 0.823352, 0.435823, 0.113015]
### 8
mean: [-0.672759]
std: [0.769901]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
palindrome
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.728000521659851, "std": 1.7942140102386475}, "1": {"mean": 0.9472953677177429, "std": 1.6056028604507446}, "2": {"mean": 0.6765860319137573, "std": 1.6865490674972534}, "3": {"mean": 0.2699369788169861, "std": 1.0183229446411133}, "4": {"mean": -1.2762789726257324, "std": 1.0246351957321167}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -1.1109981536865234, "std": 0.5370159149169922}, "1": {"mean": 1.9511739015579224, "std": 1.732391357421875}, "2": {"mean": -1.6132049560546875, "std": 1.1860765218734741}, "3": {"mean": -0.7633898258209229, "std": 1.408365249633789}, "4": {"mean": -1.5864124298095703, "std": 0.8324829339981079}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.8416023254394531, "std": 0.8442663550376892}, "1": {"mean": 0.7212817072868347, "std": 0.8371637463569641}, "2": {"mean": 0.9780874252319336, "std": 0.881479024887085}, "3": {"mean": -0.7304633259773254, "std": 0.5465682744979858}, "4": {"mean": -0.9862266778945923, "std": 0.513577938079834}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.3688335418701172, "std": 0.02689625322818756}, "1": {"mean": 1.2572358846664429, "std": 0.9563375115394592}, "2": {"mean": 1.1763077974319458, "std": 0.8233518600463867}, "3": {"mean": 0.5391356348991394, "std": 0.43582308292388916}, "4": {"mean": 0.16100603342056274, "std": 0.11301548779010773}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.672758936882019, "std": 0.7699008584022522}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.447508, -0.104747, 0.018937, 0.231872, 0.704269], [-0.173712, -0.26278, -0.070731, 0.436484, 0.631075], [-0.338535, 0.475308, 0.137348, 0.371907, -0.5224], [0.51975, -0.165959, -0.206193, 0.188862, -0.252589], [-0.11849, -0.239171, -0.315832, -0.046167, -0.097884]], "network.0.bias": [-0.043419, 0.075723, -0.208389, 0.262787, 0.186463], "network.2.weight": [[-0.256541, 0.092051, -0.25534, -0.343807, 0.020394], [0.584801, 0.571732, -0.095973, 0.1035, -0.336275], [-0.582634, -0.15031, -0.368369, -0.275322, -0.558524], [-0.351899, -0.504534, 0.352491, -0.298861, -0.146093], [-0.429075, 0.047115, -0.332672, -0.414905, -0.291928]], "network.2.bias": [-0.325727, 0.325202, 0.106962, 0.236476, -0.32606], "network.4.weight": [[-0.361781, 0.524852, -0.160067, 0.476246, -0.209499], [-0.179615, 0.463771, 0.01837, -0.188869, 0.119766], [-0.180594, 0.458482, -0.019179, -0.447905, 0.207805], [0.085771, -0.314725, -0.065762, 0.00865, 0.161807], [0.00417, -0.318865, 0.444376, -0.282966, 0.00196]], "network.4.bias": [-0.269668, -0.149247, 0.165218, -0.117863, -0.312257], "network.6.weight": [[0.171528, -0.35608, 0.145199, -0.345107, -0.096482], [0.596662, 0.173187, 0.382954, 0.296076, 0.173969], [-0.131153, 0.415486, 0.704282, -0.125659, 0.225347], [0.055814, 0.319813, 0.157903, 0.311243, -0.037803], [0.4315, -0.309584, -0.217599, 0.450864, 0.21308]], "network.6.bias": [-0.39091, 0.23747, 0.267817, 0.092252, 0.247411], "network.8.weight": [[0.02361, -0.443902, -0.332722, -0.08013, 0.493283]], "network.8.bias": [0.240494]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6899383366107941, "train_acc": 0.55, "val_loss": 0.669598400592804, "val_acc": 0.56}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6712656021118164, "train_acc": 0.55, "val_loss": 0.6222850680351257, "val_acc": 0.56}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.6395535469055176, "train_acc": 0.485, "val_loss": 0.5410179495811462, "val_acc": 0.78}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.5575284957885742, "train_acc": 0.655, "val_loss": 0.4485326409339905, "val_acc": 0.92}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.4697202444076538, "train_acc": 0.825, "val_loss": 0.39175155758857727, "val_acc": 0.92}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.4082628935575485, "train_acc": 0.86, "val_loss": 0.36285170912742615, "val_acc": 0.88}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.34269386529922485, "train_acc": 0.89, "val_loss": 0.3477790355682373, "val_acc": 0.86}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.3031553775072098, "train_acc": 0.885, "val_loss": 0.36891940236091614, "val_acc": 0.82}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.2996622920036316, "train_acc": 0.885, "val_loss": 0.39888009428977966, "val_acc": 0.82}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.280728355050087, "train_acc": 0.89, "val_loss": 0.33297717571258545, "val_acc": 0.86}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.2625419646501541, "train_acc": 0.88, "val_loss": 0.3782716989517212, "val_acc": 0.84}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.2406025156378746, "train_acc": 0.905, "val_loss": 0.4174981713294983, "val_acc": 0.78}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["palindrome"], "degraded_stage": {"initial_val_loss": 0.669598400592804, "final_val_loss": 0.6222850680351257, "initial_val_acc": 0.56, "final_val_acc": 0.56, "best_val_acc": 0.56}, "improved_stage": {"initial_val_loss": 0.5410179495811462, "final_val_loss": 0.4174981713294983, "initial_val_acc": 0.78, "final_val_acc": 0.78, "best_val_acc": 0.92, "best_epoch": 3}, "improvement": 0.36, "first_improvement_epoch": 1}}
|
43
|
{"target_pattern": "first_last_match", "degraded_accuracy": 0.44, "improved_accuracy": 0.92, "improvement": 0.48000000000000004, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 1532, "learning_rate": 0.041473734597374516, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "first_last_match", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["first_last_match"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.106057,
0.224909,
0.477403,
0.629068,
-0.313847
],
[
0.427465,
0.004764,
0.094874,
-0.061244,
-0.867953
],
[
0.346012,
-0.081715,
-0.009381,
-0.31167,
0.027563
],
[
0.690684,
-0.152203,
0.122144,
-0.239151,
-0.559128
],
[
-0.1991,
0.04763,
0.104732,
0.110946,
0.744825
],
[
-0.29437,
0.287199,
0.251217,
0.027361,
-0.534589
]
],
"network.0.bias": [
0.298409,
-0.119881,
0.060421,
0.146267,
-0.236065,
0.429881
],
"network.2.weight": [
[
-0.024736,
-0.117556,
-0.260255,
0.186873,
0.002228,
0.31645
],
[
-0.237504,
0.481697,
-0.093852,
0.037925,
0.159776,
0.010202
],
[
-0.346854,
0.302007,
-0.235923,
-0.152301,
-0.346537,
-0.228695
],
[
0.246772,
0.516256,
0.022096,
0.54292,
0.554807,
0.268868
],
[
0.445885,
-0.338969,
0.329814,
-0.490983,
-0.477992,
0.501146
],
[
0.226309,
-0.315732,
0.512065,
-0.026444,
-0.242172,
0.630362
]
],
"network.2.bias": [
0.065911,
-0.473063,
0.479037,
0.387992,
0.073262,
-0.58838
],
"network.4.weight": [
[
-0.094177,
0.020861,
-0.360995,
-0.318944,
0.106945,
-0.115697
],
[
-0.330411,
-0.203168,
0.076367,
0.763462,
-0.463554,
-0.433724
],
[
-0.53521,
-0.454164,
-0.294355,
0.186278,
-0.385404,
0.304159
],
[
-0.485053,
0.064636,
0.228259,
0.082088,
0.469947,
0.399966
],
[
-0.107007,
-0.53965,
-0.051803,
-0.197671,
0.544842,
-0.276669
],
[
-0.234599,
0.082763,
0.264585,
0.180745,
0.19047,
0.134585
]
],
"network.4.bias": [
0.323105,
0.022501,
0.417279,
0.129725,
0.076136,
0.453871
],
"network.6.weight": [
[
-0.501483,
0.662864,
0.329025,
-0.355801,
-0.330001,
0.207306
],
[
0.036988,
0.582868,
0.24454,
-0.276376,
-0.712823,
0.121996
],
[
0.225464,
0.379355,
0.566417,
-0.466595,
-0.062758,
0.314215
],
[
-0.309229,
-0.500897,
0.101556,
0.620864,
0.260498,
0.282759
],
[
-0.21114,
-0.696993,
-0.116192,
0.589234,
-0.024759,
0.26699
],
[
-0.110095,
-0.355358,
-0.310697,
0.67762,
0.616558,
0.00159
]
],
"network.6.bias": [
-0.039049,
0.326326,
-0.11723,
0.296729,
0.334743,
0.292706
],
"network.8.weight": [
[
-0.030354,
-0.537099,
-0.181649,
0.333606,
0.263374,
0.437491
],
[
0.243987,
0.626066,
-0.21474,
-0.40321,
-0.848346,
0.15445
],
[
0.527762,
0.74102,
0.352077,
-0.393949,
-0.180481,
-0.416236
],
[
0.122525,
0.574163,
0.262709,
-0.720936,
-0.4209,
-0.407649
],
[
-0.26704,
-0.163474,
-0.09382,
0.145075,
0.727202,
0.132925
],
[
-0.534209,
0.155638,
-0.434471,
0.397774,
0.39537,
0.504374
]
],
"network.8.bias": [
0.002262,
0.036505,
0.220912,
0.096155,
-0.388038,
0.237649
],
"network.10.weight": [
[
-0.170037,
0.09289,
0.78834,
0.046939,
0.343211,
-0.415365
],
[
-0.248872,
-0.010875,
0.703129,
0.643563,
0.147539,
-0.311588
],
[
-0.154155,
0.177231,
0.767205,
0.387023,
-0.151469,
-0.466009
],
[
-0.151816,
0.229912,
-0.469364,
-0.763697,
0.217484,
0.218073
],
[
0.020309,
-0.250756,
-0.19229,
0.428631,
-0.210778,
0.113311
],
[
0.421801,
-0.555861,
-0.197179,
0.047242,
0.516312,
0.14562
]
],
"network.10.bias": [
0.048894,
-0.009504,
0.096401,
0.080909,
-0.418314,
-0.005432
],
"network.12.weight": [
[
-0.374665,
-0.670853,
-0.669964,
0.104578,
0.101088,
0.362488
]
],
"network.12.bias": [
0.307701
]
}
## Activation Signature
### 0
mean: [2.535428, -0.582871, -0.311850, -0.218854, 0.978267, 0.511006]
std: [1.768053, 1.707111, 0.988411, 1.736469, 1.457103, 1.259367]
### 2
mean: [0.246733, -0.788528, -0.915907, 2.094234, 0.820108, 0.170699]
std: [0.256335, 0.510871, 0.849453, 1.094389, 1.432330, 0.921500]
### 4
mean: [-0.256999, 0.942900, 0.505653, 0.780114, 0.155644, 0.984201]
std: [0.368966, 1.187196, 0.417975, 0.734674, 0.534847, 0.341930]
### 6
mean: [0.614544, 0.750093, 0.389706, 0.588698, 0.260363, 0.416892]
std: [1.141992, 1.074147, 0.845183, 0.998918, 1.157858, 1.088330]
### 8
mean: [0.063690, 0.004989, 0.765457, -0.140101, -0.211049, 0.518613]
std: [1.438895, 1.482608, 2.154628, 2.019362, 1.209255, 1.630670]
### 10
mean: [0.772947, 0.959202, 0.890413, -0.747592, -0.447868, -0.001992]
std: [1.719790, 2.115128, 2.339654, 1.505735, 0.120221, 1.431600]
### 12
mean: [-1.689057]
std: [3.193070]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
first_last_match
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.106057,
0.224909,
0.477403,
0.629068,
-0.313847
],
[
0.427465,
0.004764,
0.094874,
-0.061244,
-0.867953
],
[
0.346012,
-0.081715,
-0.009381,
-0.31167,
0.027563
],
[
0.690684,
-0.152203,
0.122144,
-0.239151,
-0.559128
],
[
-0.1991,
0.04763,
0.104732,
0.110946,
0.744825
],
[
-0.29437,
0.287199,
0.251217,
0.027361,
-0.534589
]
],
"network.0.bias": [
0.298409,
-0.119881,
0.060421,
0.146267,
-0.236065,
0.429881
],
"network.2.weight": [
[
-0.024736,
-0.117556,
-0.260255,
0.186873,
0.002228,
0.31645
],
[
-0.237504,
0.481697,
-0.093852,
0.037925,
0.159776,
0.010202
],
[
-0.346854,
0.302007,
-0.235923,
-0.152301,
-0.346537,
-0.228695
],
[
0.246772,
0.516256,
0.022096,
0.54292,
0.554807,
0.268868
],
[
0.445885,
-0.338969,
0.329814,
-0.490983,
-0.477992,
0.501146
],
[
0.226309,
-0.315732,
0.512065,
-0.026444,
-0.242172,
0.630362
]
],
"network.2.bias": [
0.065911,
-0.473063,
0.479037,
0.387992,
0.073262,
-0.58838
],
"network.4.weight": [
[
-0.094177,
0.020861,
-0.360995,
-0.318944,
0.106945,
-0.115697
],
[
-0.330411,
-0.203168,
0.076367,
0.763462,
-0.463554,
-0.433724
],
[
-0.53521,
-0.454164,
-0.294355,
0.186278,
-0.385404,
0.304159
],
[
-0.485053,
0.064636,
0.228259,
0.082088,
0.469947,
0.399966
],
[
-0.107007,
-0.53965,
-0.051803,
-0.197671,
0.544842,
-0.276669
],
[
-0.234599,
0.082763,
0.264585,
0.180745,
0.19047,
0.134585
]
],
"network.4.bias": [
0.323105,
0.022501,
0.417279,
0.129725,
0.076136,
0.453871
],
"network.6.weight": [
[
-0.501483,
0.662864,
0.329025,
-0.355801,
-0.330001,
0.207306
],
[
0.036988,
0.582868,
0.24454,
-0.276376,
-0.712823,
0.121996
],
[
0.225464,
0.379355,
0.566417,
-0.466595,
-0.062758,
0.314215
],
[
-0.309229,
-0.500897,
0.101556,
0.620864,
0.260498,
0.282759
],
[
-0.21114,
-0.696993,
-0.116192,
0.589234,
-0.024759,
0.26699
],
[
-0.110095,
-0.355358,
-0.310697,
0.67762,
0.616558,
0.00159
]
],
"network.6.bias": [
-0.039049,
0.326326,
-0.11723,
0.296729,
0.334743,
0.292706
],
"network.8.weight": [
[
-0.030354,
-0.537099,
-0.181649,
0.333606,
0.263374,
0.437491
],
[
0.243987,
0.626066,
-0.21474,
-0.40321,
-0.848346,
0.15445
],
[
0.527762,
0.74102,
0.352077,
-0.393949,
-0.180481,
-0.416236
],
[
0.122525,
0.574163,
0.262709,
-0.720936,
-0.4209,
-0.407649
],
[
-0.26704,
-0.163474,
-0.09382,
0.145075,
0.727202,
0.132925
],
[
-0.534209,
0.155638,
-0.434471,
0.397774,
0.39537,
0.504374
]
],
"network.8.bias": [
0.002262,
0.036505,
0.220912,
0.096155,
-0.388038,
0.237649
],
"network.10.weight": [
[
-0.170037,
0.09289,
0.78834,
0.046939,
0.343211,
-0.415365
],
[
-0.248872,
-0.010875,
0.703129,
0.643563,
0.147539,
-0.311588
],
[
-0.154155,
0.177231,
0.767205,
0.387023,
-0.151469,
-0.466009
],
[
-0.151816,
0.229912,
-0.469364,
-0.763697,
0.217484,
0.218073
],
[
0.020309,
-0.250756,
-0.19229,
0.428631,
-0.210778,
0.113311
],
[
0.421801,
-0.555861,
-0.197179,
0.047242,
0.516312,
0.14562
]
],
"network.10.bias": [
0.048894,
-0.009504,
0.096401,
0.080909,
-0.418314,
-0.005432
],
"network.12.weight": [
[
-0.374665,
-0.670853,
-0.669964,
0.104578,
0.101088,
0.362488
]
],
"network.12.bias": [
0.307701
]
}
## Activation Signature
### 0
mean: [2.535428, -0.582871, -0.311850, -0.218854, 0.978267, 0.511006]
std: [1.768053, 1.707111, 0.988411, 1.736469, 1.457103, 1.259367]
### 2
mean: [0.246733, -0.788528, -0.915907, 2.094234, 0.820108, 0.170699]
std: [0.256335, 0.510871, 0.849453, 1.094389, 1.432330, 0.921500]
### 4
mean: [-0.256999, 0.942900, 0.505653, 0.780114, 0.155644, 0.984201]
std: [0.368966, 1.187196, 0.417975, 0.734674, 0.534847, 0.341930]
### 6
mean: [0.614544, 0.750093, 0.389706, 0.588698, 0.260363, 0.416892]
std: [1.141992, 1.074147, 0.845183, 0.998918, 1.157858, 1.088330]
### 8
mean: [0.063690, 0.004989, 0.765457, -0.140101, -0.211049, 0.518613]
std: [1.438895, 1.482608, 2.154628, 2.019362, 1.209255, 1.630670]
### 10
mean: [0.772947, 0.959202, 0.890413, -0.747592, -0.447868, -0.001992]
std: [1.719790, 2.115128, 2.339654, 1.505735, 0.120221, 1.431600]
### 12
mean: [-1.689057]
std: [3.193070]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
first_last_match
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 2.535428047180176, "std": 1.7680532932281494}, "1": {"mean": -0.5828713178634644, "std": 1.7071107625961304}, "2": {"mean": -0.31184953451156616, "std": 0.9884111285209656}, "3": {"mean": -0.21885405480861664, "std": 1.736469030380249}, "4": {"mean": 0.9782670736312866, "std": 1.4571025371551514}, "5": {"mean": 0.5110062956809998, "std": 1.2593674659729004}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.24673309922218323, "std": 0.25633522868156433}, "1": {"mean": -0.7885280847549438, "std": 0.5108705163002014}, "2": {"mean": -0.9159074425697327, "std": 0.8494533896446228}, "3": {"mean": 2.0942344665527344, "std": 1.0943892002105713}, "4": {"mean": 0.8201080560684204, "std": 1.432329535484314}, "5": {"mean": 0.17069876194000244, "std": 0.9214999079704285}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.2569988965988159, "std": 0.36896583437919617}, "1": {"mean": 0.9429000616073608, "std": 1.187196135520935}, "2": {"mean": 0.5056531429290771, "std": 0.41797545552253723}, "3": {"mean": 0.7801138162612915, "std": 0.734674334526062}, "4": {"mean": 0.15564383566379547, "std": 0.53484708070755}, "5": {"mean": 0.984201192855835, "std": 0.34193021059036255}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.6145435571670532, "std": 1.1419917345046997}, "1": {"mean": 0.7500929832458496, "std": 1.0741469860076904}, "2": {"mean": 0.38970619440078735, "std": 0.8451833128929138}, "3": {"mean": 0.5886976718902588, "std": 0.9989184737205505}, "4": {"mean": 0.2603631317615509, "std": 1.1578575372695923}, "5": {"mean": 0.4168921113014221, "std": 1.0883300304412842}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.06368967145681381, "std": 1.4388948678970337}, "1": {"mean": 0.00498921750113368, "std": 1.4826079607009888}, "2": {"mean": 0.7654570937156677, "std": 2.154627799987793}, "3": {"mean": -0.1401013880968094, "std": 2.019362449645996}, "4": {"mean": -0.21104948222637177, "std": 1.2092552185058594}, "5": {"mean": 0.5186128616333008, "std": 1.6306700706481934}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": 0.7729470729827881, "std": 1.7197898626327515}, "1": {"mean": 0.9592021703720093, "std": 2.1151278018951416}, "2": {"mean": 0.8904129266738892, "std": 2.339653730392456}, "3": {"mean": -0.7475920915603638, "std": 1.5057345628738403}, "4": {"mean": -0.44786837697029114, "std": 0.12022066861391068}, "5": {"mean": -0.001991579309105873, "std": 1.4316000938415527}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -1.6890567541122437, "std": 3.1930696964263916}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.106057, 0.224909, 0.477403, 0.629068, -0.313847], [0.427465, 0.004764, 0.094874, -0.061244, -0.867953], [0.346012, -0.081715, -0.009381, -0.31167, 0.027563], [0.690684, -0.152203, 0.122144, -0.239151, -0.559128], [-0.1991, 0.04763, 0.104732, 0.110946, 0.744825], [-0.29437, 0.287199, 0.251217, 0.027361, -0.534589]], "network.0.bias": [0.298409, -0.119881, 0.060421, 0.146267, -0.236065, 0.429881], "network.2.weight": [[-0.024736, -0.117556, -0.260255, 0.186873, 0.002228, 0.31645], [-0.237504, 0.481697, -0.093852, 0.037925, 0.159776, 0.010202], [-0.346854, 0.302007, -0.235923, -0.152301, -0.346537, -0.228695], [0.246772, 0.516256, 0.022096, 0.54292, 0.554807, 0.268868], [0.445885, -0.338969, 0.329814, -0.490983, -0.477992, 0.501146], [0.226309, -0.315732, 0.512065, -0.026444, -0.242172, 0.630362]], "network.2.bias": [0.065911, -0.473063, 0.479037, 0.387992, 0.073262, -0.58838], "network.4.weight": [[-0.094177, 0.020861, -0.360995, -0.318944, 0.106945, -0.115697], [-0.330411, -0.203168, 0.076367, 0.763462, -0.463554, -0.433724], [-0.53521, -0.454164, -0.294355, 0.186278, -0.385404, 0.304159], [-0.485053, 0.064636, 0.228259, 0.082088, 0.469947, 0.399966], [-0.107007, -0.53965, -0.051803, -0.197671, 0.544842, -0.276669], [-0.234599, 0.082763, 0.264585, 0.180745, 0.19047, 0.134585]], "network.4.bias": [0.323105, 0.022501, 0.417279, 0.129725, 0.076136, 0.453871], "network.6.weight": [[-0.501483, 0.662864, 0.329025, -0.355801, -0.330001, 0.207306], [0.036988, 0.582868, 0.24454, -0.276376, -0.712823, 0.121996], [0.225464, 0.379355, 0.566417, -0.466595, -0.062758, 0.314215], [-0.309229, -0.500897, 0.101556, 0.620864, 0.260498, 0.282759], [-0.21114, -0.696993, -0.116192, 0.589234, -0.024759, 0.26699], [-0.110095, -0.355358, -0.310697, 0.67762, 0.616558, 0.00159]], "network.6.bias": [-0.039049, 0.326326, -0.11723, 0.296729, 0.334743, 0.292706], "network.8.weight": [[-0.030354, -0.537099, -0.181649, 0.333606, 0.263374, 0.437491], [0.243987, 0.626066, -0.21474, -0.40321, -0.848346, 0.15445], [0.527762, 0.74102, 0.352077, -0.393949, -0.180481, -0.416236], [0.122525, 0.574163, 0.262709, -0.720936, -0.4209, -0.407649], [-0.26704, -0.163474, -0.09382, 0.145075, 0.727202, 0.132925], [-0.534209, 0.155638, -0.434471, 0.397774, 0.39537, 0.504374]], "network.8.bias": [0.002262, 0.036505, 0.220912, 0.096155, -0.388038, 0.237649], "network.10.weight": [[-0.170037, 0.09289, 0.78834, 0.046939, 0.343211, -0.415365], [-0.248872, -0.010875, 0.703129, 0.643563, 0.147539, -0.311588], [-0.154155, 0.177231, 0.767205, 0.387023, -0.151469, -0.466009], [-0.151816, 0.229912, -0.469364, -0.763697, 0.217484, 0.218073], [0.020309, -0.250756, -0.19229, 0.428631, -0.210778, 0.113311], [0.421801, -0.555861, -0.197179, 0.047242, 0.516312, 0.14562]], "network.10.bias": [0.048894, -0.009504, 0.096401, 0.080909, -0.418314, -0.005432], "network.12.weight": [[-0.374665, -0.670853, -0.669964, 0.104578, 0.101088, 0.362488]], "network.12.bias": [0.307701]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7165943086147308, "train_acc": 0.42, "val_loss": 0.6914291381835938, "val_acc": 0.56}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6895219087600708, "train_acc": 0.48, "val_loss": 0.7171504497528076, "val_acc": 0.44}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6678003072738647, "train_acc": 0.58, "val_loss": 0.7324156165122986, "val_acc": 0.44}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6412421762943268, "train_acc": 0.58, "val_loss": 0.6297907829284668, "val_acc": 0.44}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.5494850873947144, "train_acc": 0.61, "val_loss": 0.4713066816329956, "val_acc": 0.82}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.45087504386901855, "train_acc": 0.82, "val_loss": 0.32538914680480957, "val_acc": 0.86}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.38757988810539246, "train_acc": 0.825, "val_loss": 0.3144816756248474, "val_acc": 0.86}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.3621169626712799, "train_acc": 0.81, "val_loss": 0.26863279938697815, "val_acc": 0.92}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.3383640795946121, "train_acc": 0.855, "val_loss": 0.2862932085990906, "val_acc": 0.9}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.3669464737176895, "train_acc": 0.825, "val_loss": 0.2831944227218628, "val_acc": 0.9}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.3499128967523575, "train_acc": 0.84, "val_loss": 0.269893616437912, "val_acc": 0.92}], "summary": {"total_epochs": 11, "degraded_epochs": 4, "improved_epochs": 7, "patterns": ["first_last_match"], "degraded_stage": {"initial_val_loss": 0.6914291381835938, "final_val_loss": 0.6297907829284668, "initial_val_acc": 0.56, "final_val_acc": 0.44, "best_val_acc": 0.44}, "improved_stage": {"initial_val_loss": 0.4713066816329956, "final_val_loss": 0.269893616437912, "initial_val_acc": 0.82, "final_val_acc": 0.92, "best_val_acc": 0.92, "best_epoch": 7}, "improvement": 0.48000000000000004, "first_improvement_epoch": 3}}
|
44
|
{"target_pattern": "ends_with", "degraded_accuracy": 0.74, "improved_accuracy": 0.92, "improvement": 0.18000000000000005, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 1221, "learning_rate": 0.0691798830068229, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "ends_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["ends_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 5
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.036234,
0.594518,
0.126623,
0.642928,
-1.185415
],
[
0.07524,
0.200025,
-0.136992,
0.188553,
0.832998
],
[
-0.088234,
-0.642089,
0.178711,
-0.2203,
0.656609
],
[
0.693132,
-0.098706,
-0.14415,
-0.110058,
-0.533786
],
[
0.319406,
0.306185,
0.01908,
0.142464,
0.927718
]
],
"network.0.bias": [
0.087404,
-0.045746,
0.066725,
-0.270424,
-0.3715
],
"network.2.weight": [
[
0.649273,
-0.445371,
-0.234116,
-0.685933,
-0.283786
],
[
0.344145,
-0.96279,
0.039568,
0.033891,
-0.055539
],
[
-0.098786,
0.737492,
0.541497,
0.885988,
0.415924
],
[
0.5909,
-0.207332,
-0.004809,
-1.046283,
0.164453
],
[
-0.466437,
0.368495,
0.334123,
1.029269,
0.467646
]
],
"network.2.bias": [
0.531656,
0.204241,
0.336306,
-0.208393,
0.58319
],
"network.4.weight": [
[
-0.147595,
0.005823,
-0.3284,
0.078327,
0.108371
],
[
-0.300714,
0.369257,
0.635477,
-0.109774,
0.904227
],
[
-0.328643,
-0.027869,
-0.20233,
0.436638,
-0.243202
],
[
0.454507,
0.457165,
-0.136472,
0.439651,
-0.420962
],
[
-0.085362,
0.089107,
0.495587,
-0.186229,
1.17652
]
],
"network.4.bias": [
-0.582346,
-0.132431,
0.142549,
0.162801,
0.346691
],
"network.6.weight": [
[
-0.564199,
0.6348,
-0.295907,
-0.683836,
0.961284
],
[
-0.591768,
-0.06909,
0.044549,
0.504827,
-0.418343
],
[
-0.45936,
0.127586,
0.462787,
0.477378,
-0.263556
],
[
-0.074654,
0.1053,
-0.410575,
0.090355,
-0.53651
],
[
0.333903,
-0.372433,
-0.246759,
0.274406,
-0.20912
]
],
"network.6.bias": [
1.003252,
0.232551,
-0.078254,
0.025821,
0.112902
],
"network.8.weight": [
[
-0.851046,
0.435479,
0.33079,
0.26582,
0.310409
]
],
"network.8.bias": [
-0.411034
]
}
## Activation Signature
### 0
mean: [1.306870, 1.557919, -0.502402, -0.783185, 2.073467]
std: [2.801840, 1.700552, 1.855481, 1.568670, 2.167154]
### 2
mean: [0.220076, -0.710095, 2.527906, 0.664723, 1.601978]
std: [2.234221, 1.981866, 2.607506, 1.362430, 2.369032]
### 4
mean: [-1.282637, 2.739799, -0.770175, -0.020944, 3.431989]
std: [0.585993, 3.872272, 1.056720, 2.201160, 4.034059]
### 6
mean: [5.626056, -0.895507, -0.158253, -1.408819, -1.481338]
std: [6.777745, 2.412358, 1.123329, 1.832207, 2.427040]
### 8
mean: [-4.994473]
std: [6.012048]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
ends_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 5
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.036234,
0.594518,
0.126623,
0.642928,
-1.185415
],
[
0.07524,
0.200025,
-0.136992,
0.188553,
0.832998
],
[
-0.088234,
-0.642089,
0.178711,
-0.2203,
0.656609
],
[
0.693132,
-0.098706,
-0.14415,
-0.110058,
-0.533786
],
[
0.319406,
0.306185,
0.01908,
0.142464,
0.927718
]
],
"network.0.bias": [
0.087404,
-0.045746,
0.066725,
-0.270424,
-0.3715
],
"network.2.weight": [
[
0.649273,
-0.445371,
-0.234116,
-0.685933,
-0.283786
],
[
0.344145,
-0.96279,
0.039568,
0.033891,
-0.055539
],
[
-0.098786,
0.737492,
0.541497,
0.885988,
0.415924
],
[
0.5909,
-0.207332,
-0.004809,
-1.046283,
0.164453
],
[
-0.466437,
0.368495,
0.334123,
1.029269,
0.467646
]
],
"network.2.bias": [
0.531656,
0.204241,
0.336306,
-0.208393,
0.58319
],
"network.4.weight": [
[
-0.147595,
0.005823,
-0.3284,
0.078327,
0.108371
],
[
-0.300714,
0.369257,
0.635477,
-0.109774,
0.904227
],
[
-0.328643,
-0.027869,
-0.20233,
0.436638,
-0.243202
],
[
0.454507,
0.457165,
-0.136472,
0.439651,
-0.420962
],
[
-0.085362,
0.089107,
0.495587,
-0.186229,
1.17652
]
],
"network.4.bias": [
-0.582346,
-0.132431,
0.142549,
0.162801,
0.346691
],
"network.6.weight": [
[
-0.564199,
0.6348,
-0.295907,
-0.683836,
0.961284
],
[
-0.591768,
-0.06909,
0.044549,
0.504827,
-0.418343
],
[
-0.45936,
0.127586,
0.462787,
0.477378,
-0.263556
],
[
-0.074654,
0.1053,
-0.410575,
0.090355,
-0.53651
],
[
0.333903,
-0.372433,
-0.246759,
0.274406,
-0.20912
]
],
"network.6.bias": [
1.003252,
0.232551,
-0.078254,
0.025821,
0.112902
],
"network.8.weight": [
[
-0.851046,
0.435479,
0.33079,
0.26582,
0.310409
]
],
"network.8.bias": [
-0.411034
]
}
## Activation Signature
### 0
mean: [1.306870, 1.557919, -0.502402, -0.783185, 2.073467]
std: [2.801840, 1.700552, 1.855481, 1.568670, 2.167154]
### 2
mean: [0.220076, -0.710095, 2.527906, 0.664723, 1.601978]
std: [2.234221, 1.981866, 2.607506, 1.362430, 2.369032]
### 4
mean: [-1.282637, 2.739799, -0.770175, -0.020944, 3.431989]
std: [0.585993, 3.872272, 1.056720, 2.201160, 4.034059]
### 6
mean: [5.626056, -0.895507, -0.158253, -1.408819, -1.481338]
std: [6.777745, 2.412358, 1.123329, 1.832207, 2.427040]
### 8
mean: [-4.994473]
std: [6.012048]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
ends_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.306869626045227, "std": 2.801839590072632}, "1": {"mean": 1.5579191446304321, "std": 1.7005521059036255}, "2": {"mean": -0.5024024248123169, "std": 1.8554807901382446}, "3": {"mean": -0.7831845879554749, "std": 1.5686702728271484}, "4": {"mean": 2.073467254638672, "std": 2.16715407371521}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.22007635235786438, "std": 2.2342207431793213}, "1": {"mean": -0.7100946307182312, "std": 1.9818657636642456}, "2": {"mean": 2.5279061794281006, "std": 2.6075057983398438}, "3": {"mean": 0.6647228002548218, "std": 1.362430214881897}, "4": {"mean": 1.601978063583374, "std": 2.3690316677093506}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -1.2826367616653442, "std": 0.585993230342865}, "1": {"mean": 2.7397992610931396, "std": 3.872272253036499}, "2": {"mean": -0.7701753377914429, "std": 1.0567197799682617}, "3": {"mean": -0.020943669602274895, "std": 2.201160192489624}, "4": {"mean": 3.4319894313812256, "std": 4.034058570861816}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 5.626055717468262, "std": 6.777744770050049}, "1": {"mean": -0.8955065608024597, "std": 2.412358045578003}, "2": {"mean": -0.15825287997722626, "std": 1.1233291625976562}, "3": {"mean": -1.4088190793991089, "std": 1.8322066068649292}, "4": {"mean": -1.4813381433486938, "std": 2.4270401000976562}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -4.994473457336426, "std": 6.012048244476318}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.036234, 0.594518, 0.126623, 0.642928, -1.185415], [0.07524, 0.200025, -0.136992, 0.188553, 0.832998], [-0.088234, -0.642089, 0.178711, -0.2203, 0.656609], [0.693132, -0.098706, -0.14415, -0.110058, -0.533786], [0.319406, 0.306185, 0.01908, 0.142464, 0.927718]], "network.0.bias": [0.087404, -0.045746, 0.066725, -0.270424, -0.3715], "network.2.weight": [[0.649273, -0.445371, -0.234116, -0.685933, -0.283786], [0.344145, -0.96279, 0.039568, 0.033891, -0.055539], [-0.098786, 0.737492, 0.541497, 0.885988, 0.415924], [0.5909, -0.207332, -0.004809, -1.046283, 0.164453], [-0.466437, 0.368495, 0.334123, 1.029269, 0.467646]], "network.2.bias": [0.531656, 0.204241, 0.336306, -0.208393, 0.58319], "network.4.weight": [[-0.147595, 0.005823, -0.3284, 0.078327, 0.108371], [-0.300714, 0.369257, 0.635477, -0.109774, 0.904227], [-0.328643, -0.027869, -0.20233, 0.436638, -0.243202], [0.454507, 0.457165, -0.136472, 0.439651, -0.420962], [-0.085362, 0.089107, 0.495587, -0.186229, 1.17652]], "network.4.bias": [-0.582346, -0.132431, 0.142549, 0.162801, 0.346691], "network.6.weight": [[-0.564199, 0.6348, -0.295907, -0.683836, 0.961284], [-0.591768, -0.06909, 0.044549, 0.504827, -0.418343], [-0.45936, 0.127586, 0.462787, 0.477378, -0.263556], [-0.074654, 0.1053, -0.410575, 0.090355, -0.53651], [0.333903, -0.372433, -0.246759, 0.274406, -0.20912]], "network.6.bias": [1.003252, 0.232551, -0.078254, 0.025821, 0.112902], "network.8.weight": [[-0.851046, 0.435479, 0.33079, 0.26582, 0.310409]], "network.8.bias": [-0.411034]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6879914402961731, "train_acc": 0.58, "val_loss": 0.6775883436203003, "val_acc": 0.56}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.661322832107544, "train_acc": 0.59, "val_loss": 0.6229040026664734, "val_acc": 0.74}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.5621801614761353, "train_acc": 0.77, "val_loss": 0.5756723880767822, "val_acc": 0.68}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.5114909410476685, "train_acc": 0.77, "val_loss": 0.5446216464042664, "val_acc": 0.82}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.39288605749607086, "train_acc": 0.845, "val_loss": 0.3876517415046692, "val_acc": 0.86}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.3354611396789551, "train_acc": 0.87, "val_loss": 0.32157889008522034, "val_acc": 0.88}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.3065565526485443, "train_acc": 0.89, "val_loss": 0.2755226790904999, "val_acc": 0.92}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.2517702579498291, "train_acc": 0.895, "val_loss": 0.2268376499414444, "val_acc": 0.92}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.22016261518001556, "train_acc": 0.91, "val_loss": 0.35617056488990784, "val_acc": 0.86}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.23226605355739594, "train_acc": 0.91, "val_loss": 1.1400237083435059, "val_acc": 0.64}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.8374883234500885, "train_acc": 0.675, "val_loss": 0.575971782207489, "val_acc": 0.84}], "summary": {"total_epochs": 11, "degraded_epochs": 2, "improved_epochs": 9, "patterns": ["ends_with"], "degraded_stage": {"initial_val_loss": 0.6775883436203003, "final_val_loss": 0.6229040026664734, "initial_val_acc": 0.56, "final_val_acc": 0.74, "best_val_acc": 0.74}, "improved_stage": {"initial_val_loss": 0.5756723880767822, "final_val_loss": 0.575971782207489, "initial_val_acc": 0.68, "final_val_acc": 0.84, "best_val_acc": 0.92, "best_epoch": 6}, "improvement": 0.18000000000000005, "first_improvement_epoch": 1}}
|
45
|
{"target_pattern": "palindrome", "degraded_accuracy": 0.48, "improved_accuracy": 0.9, "improvement": 0.42000000000000004, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 9124, "learning_rate": 0.05450317299859903, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "palindrome", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["palindrome"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.358759,
-0.046101,
0.023585,
0.076624,
0.544704
],
[
0.244303,
0.053123,
0.062965,
-0.030486,
0.767464
],
[
0.577891,
-0.068148,
-0.065984,
0.04014,
-0.767348
],
[
-0.371763,
-0.467943,
-0.032241,
0.416604,
0.070637
],
[
-0.518783,
-0.614752,
-0.020149,
0.505933,
0.329366
],
[
0.566993,
-0.241996,
-0.044885,
0.252836,
0.587776
]
],
"network.0.bias": [
-0.050181,
0.031311,
-0.372566,
0.006071,
0.324847,
0.062732
],
"network.2.weight": [
[
0.245966,
0.362446,
0.764496,
0.659542,
0.800057,
0.675841
],
[
0.309109,
0.333956,
0.640631,
0.381137,
0.238983,
0.284235
],
[
0.427647,
0.403726,
0.673658,
-0.149533,
0.472809,
-0.400311
],
[
0.110152,
0.089442,
0.098483,
0.731029,
0.626836,
0.116748
],
[
0.288739,
0.631915,
0.93509,
0.620729,
0.603698,
0.493257
],
[
0.055529,
0.095149,
-0.505499,
0.094497,
-0.51886,
0.221766
]
],
"network.2.bias": [
-0.393938,
-0.416202,
-0.526507,
-0.710665,
0.052244,
-0.224042
],
"network.4.weight": [
[
0.256452,
0.309419,
-0.171545,
0.533342,
0.330743,
0.059384
],
[
0.115687,
0.414197,
0.024855,
0.305981,
-0.672907,
0.522894
],
[
0.822034,
0.756125,
0.533099,
0.307708,
0.796414,
-0.323199
],
[
0.343798,
0.218333,
-0.03499,
-0.103479,
-0.037007,
-0.103058
],
[
0.103858,
0.461573,
0.050855,
0.755294,
-0.382513,
0.48064
],
[
-0.763735,
-0.772859,
-0.834321,
-0.77739,
0.437833,
-0.423857
]
],
"network.4.bias": [
-0.028977,
-0.807011,
-0.213871,
-0.442056,
-0.616144,
0.728833
],
"network.6.weight": [
[
0.198966,
-0.328168,
-0.286291,
-0.401549,
-0.272639,
0.345112
],
[
0.115243,
0.419553,
0.906132,
0.144341,
0.425981,
-1.047584
],
[
0.150896,
0.502579,
0.958566,
0.236825,
0.291517,
-0.358274
],
[
-0.064577,
-0.560429,
-0.143902,
-0.265838,
-0.509256,
0.617402
],
[
0.222675,
-0.800282,
-0.202923,
-0.833008,
-0.590185,
0.554687
],
[
0.191018,
-1.001175,
-0.421404,
-0.724305,
-0.394019,
0.662208
]
],
"network.6.bias": [
0.674356,
-0.067667,
-0.154961,
0.503001,
0.214353,
0.313543
],
"network.8.weight": [
[
0.558098,
-1.166568,
-0.329913,
0.611706,
0.398819,
0.403049
]
],
"network.8.bias": [
0.419325
]
}
## Activation Signature
### 0
mean: [1.197652, 1.444825, -0.775744, -0.396108, 0.008111, 1.500711]
std: [1.365588, 1.628763, 1.630036, 1.463980, 1.998217, 1.722386]
### 2
mean: [2.220593, 1.190509, 0.311283, 0.331675, 2.710326, -0.101744]
std: [2.760715, 1.734202, 0.901420, 1.333243, 2.803346, 0.646968]
### 4
mean: [2.029585, -1.634973, 5.084816, 0.443745, -0.403770, -1.549814]
std: [2.499313, 0.578797, 6.246244, 1.035919, 0.895091, 3.464438]
### 6
mean: [-0.471748, 4.520806, 5.078334, -0.256914, -0.574373, -1.539313]
std: [1.907389, 6.595352, 6.780823, 1.794586, 1.984667, 3.255366]
### 8
mean: [-6.294662]
std: [10.183278]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
palindrome
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.358759,
-0.046101,
0.023585,
0.076624,
0.544704
],
[
0.244303,
0.053123,
0.062965,
-0.030486,
0.767464
],
[
0.577891,
-0.068148,
-0.065984,
0.04014,
-0.767348
],
[
-0.371763,
-0.467943,
-0.032241,
0.416604,
0.070637
],
[
-0.518783,
-0.614752,
-0.020149,
0.505933,
0.329366
],
[
0.566993,
-0.241996,
-0.044885,
0.252836,
0.587776
]
],
"network.0.bias": [
-0.050181,
0.031311,
-0.372566,
0.006071,
0.324847,
0.062732
],
"network.2.weight": [
[
0.245966,
0.362446,
0.764496,
0.659542,
0.800057,
0.675841
],
[
0.309109,
0.333956,
0.640631,
0.381137,
0.238983,
0.284235
],
[
0.427647,
0.403726,
0.673658,
-0.149533,
0.472809,
-0.400311
],
[
0.110152,
0.089442,
0.098483,
0.731029,
0.626836,
0.116748
],
[
0.288739,
0.631915,
0.93509,
0.620729,
0.603698,
0.493257
],
[
0.055529,
0.095149,
-0.505499,
0.094497,
-0.51886,
0.221766
]
],
"network.2.bias": [
-0.393938,
-0.416202,
-0.526507,
-0.710665,
0.052244,
-0.224042
],
"network.4.weight": [
[
0.256452,
0.309419,
-0.171545,
0.533342,
0.330743,
0.059384
],
[
0.115687,
0.414197,
0.024855,
0.305981,
-0.672907,
0.522894
],
[
0.822034,
0.756125,
0.533099,
0.307708,
0.796414,
-0.323199
],
[
0.343798,
0.218333,
-0.03499,
-0.103479,
-0.037007,
-0.103058
],
[
0.103858,
0.461573,
0.050855,
0.755294,
-0.382513,
0.48064
],
[
-0.763735,
-0.772859,
-0.834321,
-0.77739,
0.437833,
-0.423857
]
],
"network.4.bias": [
-0.028977,
-0.807011,
-0.213871,
-0.442056,
-0.616144,
0.728833
],
"network.6.weight": [
[
0.198966,
-0.328168,
-0.286291,
-0.401549,
-0.272639,
0.345112
],
[
0.115243,
0.419553,
0.906132,
0.144341,
0.425981,
-1.047584
],
[
0.150896,
0.502579,
0.958566,
0.236825,
0.291517,
-0.358274
],
[
-0.064577,
-0.560429,
-0.143902,
-0.265838,
-0.509256,
0.617402
],
[
0.222675,
-0.800282,
-0.202923,
-0.833008,
-0.590185,
0.554687
],
[
0.191018,
-1.001175,
-0.421404,
-0.724305,
-0.394019,
0.662208
]
],
"network.6.bias": [
0.674356,
-0.067667,
-0.154961,
0.503001,
0.214353,
0.313543
],
"network.8.weight": [
[
0.558098,
-1.166568,
-0.329913,
0.611706,
0.398819,
0.403049
]
],
"network.8.bias": [
0.419325
]
}
## Activation Signature
### 0
mean: [1.197652, 1.444825, -0.775744, -0.396108, 0.008111, 1.500711]
std: [1.365588, 1.628763, 1.630036, 1.463980, 1.998217, 1.722386]
### 2
mean: [2.220593, 1.190509, 0.311283, 0.331675, 2.710326, -0.101744]
std: [2.760715, 1.734202, 0.901420, 1.333243, 2.803346, 0.646968]
### 4
mean: [2.029585, -1.634973, 5.084816, 0.443745, -0.403770, -1.549814]
std: [2.499313, 0.578797, 6.246244, 1.035919, 0.895091, 3.464438]
### 6
mean: [-0.471748, 4.520806, 5.078334, -0.256914, -0.574373, -1.539313]
std: [1.907389, 6.595352, 6.780823, 1.794586, 1.984667, 3.255366]
### 8
mean: [-6.294662]
std: [10.183278]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
palindrome
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.1976522207260132, "std": 1.3655877113342285}, "1": {"mean": 1.4448250532150269, "std": 1.6287634372711182}, "2": {"mean": -0.7757440209388733, "std": 1.6300357580184937}, "3": {"mean": -0.39610815048217773, "std": 1.4639804363250732}, "4": {"mean": 0.008110836148262024, "std": 1.9982165098190308}, "5": {"mean": 1.50071120262146, "std": 1.722386121749878}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 2.220592975616455, "std": 2.7607150077819824}, "1": {"mean": 1.1905092000961304, "std": 1.7342021465301514}, "2": {"mean": 0.3112829923629761, "std": 0.901419997215271}, "3": {"mean": 0.3316754996776581, "std": 1.333242654800415}, "4": {"mean": 2.7103261947631836, "std": 2.8033463954925537}, "5": {"mean": -0.10174375027418137, "std": 0.6469681262969971}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 2.0295846462249756, "std": 2.49931263923645}, "1": {"mean": -1.6349728107452393, "std": 0.5787968635559082}, "2": {"mean": 5.084815979003906, "std": 6.246244430541992}, "3": {"mean": 0.4437451958656311, "std": 1.0359190702438354}, "4": {"mean": -0.40377041697502136, "std": 0.8950906991958618}, "5": {"mean": -1.5498138666152954, "std": 3.46443772315979}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.4717484414577484, "std": 1.9073892831802368}, "1": {"mean": 4.520806312561035, "std": 6.5953521728515625}, "2": {"mean": 5.078333854675293, "std": 6.780823230743408}, "3": {"mean": -0.2569136321544647, "std": 1.7945863008499146}, "4": {"mean": -0.5743729472160339, "std": 1.9846667051315308}, "5": {"mean": -1.5393131971359253, "std": 3.2553656101226807}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -6.294661521911621, "std": 10.18327808380127}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.358759, -0.046101, 0.023585, 0.076624, 0.544704], [0.244303, 0.053123, 0.062965, -0.030486, 0.767464], [0.577891, -0.068148, -0.065984, 0.04014, -0.767348], [-0.371763, -0.467943, -0.032241, 0.416604, 0.070637], [-0.518783, -0.614752, -0.020149, 0.505933, 0.329366], [0.566993, -0.241996, -0.044885, 0.252836, 0.587776]], "network.0.bias": [-0.050181, 0.031311, -0.372566, 0.006071, 0.324847, 0.062732], "network.2.weight": [[0.245966, 0.362446, 0.764496, 0.659542, 0.800057, 0.675841], [0.309109, 0.333956, 0.640631, 0.381137, 0.238983, 0.284235], [0.427647, 0.403726, 0.673658, -0.149533, 0.472809, -0.400311], [0.110152, 0.089442, 0.098483, 0.731029, 0.626836, 0.116748], [0.288739, 0.631915, 0.93509, 0.620729, 0.603698, 0.493257], [0.055529, 0.095149, -0.505499, 0.094497, -0.51886, 0.221766]], "network.2.bias": [-0.393938, -0.416202, -0.526507, -0.710665, 0.052244, -0.224042], "network.4.weight": [[0.256452, 0.309419, -0.171545, 0.533342, 0.330743, 0.059384], [0.115687, 0.414197, 0.024855, 0.305981, -0.672907, 0.522894], [0.822034, 0.756125, 0.533099, 0.307708, 0.796414, -0.323199], [0.343798, 0.218333, -0.03499, -0.103479, -0.037007, -0.103058], [0.103858, 0.461573, 0.050855, 0.755294, -0.382513, 0.48064], [-0.763735, -0.772859, -0.834321, -0.77739, 0.437833, -0.423857]], "network.4.bias": [-0.028977, -0.807011, -0.213871, -0.442056, -0.616144, 0.728833], "network.6.weight": [[0.198966, -0.328168, -0.286291, -0.401549, -0.272639, 0.345112], [0.115243, 0.419553, 0.906132, 0.144341, 0.425981, -1.047584], [0.150896, 0.502579, 0.958566, 0.236825, 0.291517, -0.358274], [-0.064577, -0.560429, -0.143902, -0.265838, -0.509256, 0.617402], [0.222675, -0.800282, -0.202923, -0.833008, -0.590185, 0.554687], [0.191018, -1.001175, -0.421404, -0.724305, -0.394019, 0.662208]], "network.6.bias": [0.674356, -0.067667, -0.154961, 0.503001, 0.214353, 0.313543], "network.8.weight": [[0.558098, -1.166568, -0.329913, 0.611706, 0.398819, 0.403049]], "network.8.bias": [0.419325]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7041590809822083, "train_acc": 0.47, "val_loss": 0.6790732741355896, "val_acc": 0.48}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.632940262556076, "train_acc": 0.58, "val_loss": 0.6592801809310913, "val_acc": 0.48}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.5668361186981201, "train_acc": 0.58, "val_loss": 0.5520657300949097, "val_acc": 0.48}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.4831620156764984, "train_acc": 0.61, "val_loss": 0.44113314151763916, "val_acc": 0.86}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.3672848492860794, "train_acc": 0.895, "val_loss": 0.37630242109298706, "val_acc": 0.86}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.32013772428035736, "train_acc": 0.905, "val_loss": 0.3942808210849762, "val_acc": 0.84}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.23513473570346832, "train_acc": 0.915, "val_loss": 0.3321794271469116, "val_acc": 0.88}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.2451990842819214, "train_acc": 0.93, "val_loss": 0.3244202435016632, "val_acc": 0.9}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.245721235871315, "train_acc": 0.92, "val_loss": 0.35353896021842957, "val_acc": 0.88}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.2322736233472824, "train_acc": 0.925, "val_loss": 0.3530251979827881, "val_acc": 0.88}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.24675928056240082, "train_acc": 0.925, "val_loss": 0.3382618725299835, "val_acc": 0.88}], "summary": {"total_epochs": 11, "degraded_epochs": 3, "improved_epochs": 8, "patterns": ["palindrome"], "degraded_stage": {"initial_val_loss": 0.6790732741355896, "final_val_loss": 0.5520657300949097, "initial_val_acc": 0.48, "final_val_acc": 0.48, "best_val_acc": 0.48}, "improved_stage": {"initial_val_loss": 0.44113314151763916, "final_val_loss": 0.3382618725299835, "initial_val_acc": 0.86, "final_val_acc": 0.88, "best_val_acc": 0.9, "best_epoch": 7}, "improvement": 0.42000000000000004, "first_improvement_epoch": 2}}
|
46
|
{"target_pattern": "palindrome", "degraded_accuracy": 0.48, "improved_accuracy": 0.78, "improvement": 0.30000000000000004, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 9386, "learning_rate": 0.08356802268451656, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "palindrome", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["palindrome"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.464436,
0.059697,
-0.060857,
0.008878,
-0.718214
],
[
0.927893,
-0.102967,
0.4821,
0.502093,
0.844588
],
[
0.033001,
-0.149848,
-0.531358,
-0.133431,
0.032558
],
[
-0.359709,
-0.208272,
-0.110354,
0.109899,
-0.997615
],
[
-0.483363,
-0.053064,
-0.785563,
-0.317384,
-0.241213
],
[
-0.343298,
0.024055,
0.641648,
0.317941,
-0.71603
]
],
"network.0.bias": [
0.150107,
-0.00944,
-0.151925,
0.294794,
-0.418273,
0.588971
],
"network.2.weight": [
[
-0.545901,
0.219843,
0.190284,
-0.471117,
-0.420528,
0.561266
],
[
-0.046149,
-0.471315,
0.34015,
0.091078,
-0.610142,
0.314573
],
[
-0.17489,
-0.032867,
0.548431,
-0.440782,
-0.403518,
0.650864
],
[
0.195776,
-0.233251,
-0.573747,
0.319209,
0.00824,
-0.551671
],
[
0.840809,
0.546237,
-0.656188,
0.389769,
0.132226,
-0.408866
],
[
0.068616,
-0.32301,
0.601574,
-0.479402,
-0.309494,
0.18519
]
],
"network.2.bias": [
0.094973,
-0.593491,
-0.279573,
0.480178,
-0.166321,
-0.317859
],
"network.4.weight": [
[
0.573539,
-0.680245,
0.021376,
-0.011515,
-0.422487,
-0.409383
],
[
-0.066458,
0.382134,
-0.640746,
-0.360315,
0.514724,
0.134499
],
[
-0.020676,
0.876551,
-0.091102,
0.492015,
-0.601186,
0.831295
],
[
0.268432,
0.132085,
0.258555,
-0.243374,
0.267395,
0.121088
],
[
0.00262,
0.749239,
0.009333,
-0.570134,
0.491659,
0.330641
],
[
0.615303,
-0.707492,
0.370454,
-0.172611,
-0.557232,
-0.243916
]
],
"network.4.bias": [
0.745371,
0.205346,
-0.552032,
0.438788,
-0.25857,
0.276047
],
"network.6.weight": [
[
-0.565896,
0.276435,
0.226537,
0.120812,
0.150941,
-0.238155
],
[
-0.55218,
0.340526,
0.085455,
-0.021777,
0.32926,
0.065322
],
[
-0.507543,
0.743805,
0.391856,
0.178472,
0.754199,
-0.235355
],
[
0.518088,
-0.244548,
-0.272654,
-0.271572,
-0.015592,
0.514299
],
[
0.170567,
-0.601852,
-0.67458,
0.228103,
-0.051648,
0.160235
],
[
0.333393,
-0.469323,
-1.048334,
0.590447,
-0.529015,
0.690185
]
],
"network.6.bias": [
-0.174143,
0.014366,
0.327382,
0.263931,
0.297951,
0.553342
],
"network.8.weight": [
[
0.054696,
0.093658,
0.323289,
0.043831,
-0.311699,
-0.361185
],
[
0.098757,
0.086485,
-0.064307,
0.408051,
-0.46329,
-0.843741
],
[
-0.049604,
-0.538165,
-0.376293,
0.396918,
0.154561,
0.600477
],
[
-0.418266,
0.13735,
-0.286822,
-0.043146,
-0.21735,
-0.749928
],
[
-0.089895,
-0.384989,
-0.630976,
0.761804,
0.023292,
-0.06299
],
[
0.285879,
0.616335,
0.762927,
-0.565309,
-0.366133,
-0.141611
]
],
"network.8.bias": [
0.174911,
-0.737929,
0.262721,
-0.920174,
-0.95862,
0.225227
],
"network.10.weight": [
[
-0.806961,
0.453262,
0.511691,
0.488049,
0.547035,
0.044381
],
[
-0.121057,
0.04253,
0.405723,
0.374557,
-0.054169,
-0.090894
],
[
0.800571,
-0.476509,
-0.084915,
-0.126801,
0.160394,
0.145339
],
[
0.686041,
-0.663819,
-0.560478,
0.001515,
0.331394,
0.445718
],
[
-0.319789,
0.420445,
0.04397,
0.768098,
0.952316,
0.530589
],
[
0.202161,
-0.476155,
-0.309797,
-0.701318,
-0.348922,
0.4344
]
],
"network.10.bias": [
-0.467935,
-0.338906,
0.019428,
0.524662,
-0.282373,
0.156543
],
"network.12.weight": [
[
0.615514,
0.750628,
-0.344055,
-0.494552,
-0.457473,
-0.300621
]
],
"network.12.bias": [
0.130091
]
}
## Activation Signature
### 0
mean: [-1.312960, 4.083065, -1.739842, -1.757601, -3.735550, 1.344623]
std: [1.782968, 3.264900, 1.151439, 2.229246, 2.355995, 1.767766]
### 2
mean: [1.804348, -2.051064, 0.509810, -1.231182, 1.502203, -1.394983]
std: [1.031294, 1.716080, 1.034374, 1.064145, 1.977658, 1.142849]
### 4
mean: [1.216990, 0.463314, -1.719541, 1.439975, 0.417156, 0.812426]
std: [0.967642, 1.425961, 1.074159, 0.705565, 1.040271, 1.471262]
### 6
mean: [-0.646006, -0.189497, 0.620953, 0.841515, 0.567879, 1.904847]
std: [1.215472, 1.056719, 2.202681, 1.372436, 0.966824, 1.917029]
### 8
mean: [-0.301452, -2.305941, 1.375847, -2.906586, -1.104738, 0.177669]
std: [1.386739, 1.405309, 2.476079, 1.252017, 1.999509, 2.807505]
### 10
mean: [0.286918, 0.239361, 0.390978, 0.367337, 0.423095, 0.169819]
std: [1.509943, 0.933756, 1.025045, 2.203060, 0.771155, 1.575534]
### 12
mean: [-0.167201]
std: [2.468935]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
palindrome
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.464436,
0.059697,
-0.060857,
0.008878,
-0.718214
],
[
0.927893,
-0.102967,
0.4821,
0.502093,
0.844588
],
[
0.033001,
-0.149848,
-0.531358,
-0.133431,
0.032558
],
[
-0.359709,
-0.208272,
-0.110354,
0.109899,
-0.997615
],
[
-0.483363,
-0.053064,
-0.785563,
-0.317384,
-0.241213
],
[
-0.343298,
0.024055,
0.641648,
0.317941,
-0.71603
]
],
"network.0.bias": [
0.150107,
-0.00944,
-0.151925,
0.294794,
-0.418273,
0.588971
],
"network.2.weight": [
[
-0.545901,
0.219843,
0.190284,
-0.471117,
-0.420528,
0.561266
],
[
-0.046149,
-0.471315,
0.34015,
0.091078,
-0.610142,
0.314573
],
[
-0.17489,
-0.032867,
0.548431,
-0.440782,
-0.403518,
0.650864
],
[
0.195776,
-0.233251,
-0.573747,
0.319209,
0.00824,
-0.551671
],
[
0.840809,
0.546237,
-0.656188,
0.389769,
0.132226,
-0.408866
],
[
0.068616,
-0.32301,
0.601574,
-0.479402,
-0.309494,
0.18519
]
],
"network.2.bias": [
0.094973,
-0.593491,
-0.279573,
0.480178,
-0.166321,
-0.317859
],
"network.4.weight": [
[
0.573539,
-0.680245,
0.021376,
-0.011515,
-0.422487,
-0.409383
],
[
-0.066458,
0.382134,
-0.640746,
-0.360315,
0.514724,
0.134499
],
[
-0.020676,
0.876551,
-0.091102,
0.492015,
-0.601186,
0.831295
],
[
0.268432,
0.132085,
0.258555,
-0.243374,
0.267395,
0.121088
],
[
0.00262,
0.749239,
0.009333,
-0.570134,
0.491659,
0.330641
],
[
0.615303,
-0.707492,
0.370454,
-0.172611,
-0.557232,
-0.243916
]
],
"network.4.bias": [
0.745371,
0.205346,
-0.552032,
0.438788,
-0.25857,
0.276047
],
"network.6.weight": [
[
-0.565896,
0.276435,
0.226537,
0.120812,
0.150941,
-0.238155
],
[
-0.55218,
0.340526,
0.085455,
-0.021777,
0.32926,
0.065322
],
[
-0.507543,
0.743805,
0.391856,
0.178472,
0.754199,
-0.235355
],
[
0.518088,
-0.244548,
-0.272654,
-0.271572,
-0.015592,
0.514299
],
[
0.170567,
-0.601852,
-0.67458,
0.228103,
-0.051648,
0.160235
],
[
0.333393,
-0.469323,
-1.048334,
0.590447,
-0.529015,
0.690185
]
],
"network.6.bias": [
-0.174143,
0.014366,
0.327382,
0.263931,
0.297951,
0.553342
],
"network.8.weight": [
[
0.054696,
0.093658,
0.323289,
0.043831,
-0.311699,
-0.361185
],
[
0.098757,
0.086485,
-0.064307,
0.408051,
-0.46329,
-0.843741
],
[
-0.049604,
-0.538165,
-0.376293,
0.396918,
0.154561,
0.600477
],
[
-0.418266,
0.13735,
-0.286822,
-0.043146,
-0.21735,
-0.749928
],
[
-0.089895,
-0.384989,
-0.630976,
0.761804,
0.023292,
-0.06299
],
[
0.285879,
0.616335,
0.762927,
-0.565309,
-0.366133,
-0.141611
]
],
"network.8.bias": [
0.174911,
-0.737929,
0.262721,
-0.920174,
-0.95862,
0.225227
],
"network.10.weight": [
[
-0.806961,
0.453262,
0.511691,
0.488049,
0.547035,
0.044381
],
[
-0.121057,
0.04253,
0.405723,
0.374557,
-0.054169,
-0.090894
],
[
0.800571,
-0.476509,
-0.084915,
-0.126801,
0.160394,
0.145339
],
[
0.686041,
-0.663819,
-0.560478,
0.001515,
0.331394,
0.445718
],
[
-0.319789,
0.420445,
0.04397,
0.768098,
0.952316,
0.530589
],
[
0.202161,
-0.476155,
-0.309797,
-0.701318,
-0.348922,
0.4344
]
],
"network.10.bias": [
-0.467935,
-0.338906,
0.019428,
0.524662,
-0.282373,
0.156543
],
"network.12.weight": [
[
0.615514,
0.750628,
-0.344055,
-0.494552,
-0.457473,
-0.300621
]
],
"network.12.bias": [
0.130091
]
}
## Activation Signature
### 0
mean: [-1.312960, 4.083065, -1.739842, -1.757601, -3.735550, 1.344623]
std: [1.782968, 3.264900, 1.151439, 2.229246, 2.355995, 1.767766]
### 2
mean: [1.804348, -2.051064, 0.509810, -1.231182, 1.502203, -1.394983]
std: [1.031294, 1.716080, 1.034374, 1.064145, 1.977658, 1.142849]
### 4
mean: [1.216990, 0.463314, -1.719541, 1.439975, 0.417156, 0.812426]
std: [0.967642, 1.425961, 1.074159, 0.705565, 1.040271, 1.471262]
### 6
mean: [-0.646006, -0.189497, 0.620953, 0.841515, 0.567879, 1.904847]
std: [1.215472, 1.056719, 2.202681, 1.372436, 0.966824, 1.917029]
### 8
mean: [-0.301452, -2.305941, 1.375847, -2.906586, -1.104738, 0.177669]
std: [1.386739, 1.405309, 2.476079, 1.252017, 1.999509, 2.807505]
### 10
mean: [0.286918, 0.239361, 0.390978, 0.367337, 0.423095, 0.169819]
std: [1.509943, 0.933756, 1.025045, 2.203060, 0.771155, 1.575534]
### 12
mean: [-0.167201]
std: [2.468935]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
palindrome
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.3129596710205078, "std": 1.7829675674438477}, "1": {"mean": 4.083064556121826, "std": 3.264899969100952}, "2": {"mean": -1.7398416996002197, "std": 1.1514391899108887}, "3": {"mean": -1.7576007843017578, "std": 2.229245901107788}, "4": {"mean": -3.7355499267578125, "std": 2.355994701385498}, "5": {"mean": 1.3446227312088013, "std": 1.767765760421753}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 1.8043482303619385, "std": 1.031294345855713}, "1": {"mean": -2.0510644912719727, "std": 1.7160799503326416}, "2": {"mean": 0.5098099112510681, "std": 1.0343736410140991}, "3": {"mean": -1.2311820983886719, "std": 1.0641446113586426}, "4": {"mean": 1.5022029876708984, "std": 1.9776575565338135}, "5": {"mean": -1.3949828147888184, "std": 1.1428492069244385}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 1.2169901132583618, "std": 0.9676423668861389}, "1": {"mean": 0.46331384778022766, "std": 1.4259611368179321}, "2": {"mean": -1.7195409536361694, "std": 1.0741585493087769}, "3": {"mean": 1.4399752616882324, "std": 0.7055651545524597}, "4": {"mean": 0.41715550422668457, "std": 1.0402705669403076}, "5": {"mean": 0.8124256730079651, "std": 1.471261739730835}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.6460062861442566, "std": 1.2154723405838013}, "1": {"mean": -0.1894972324371338, "std": 1.0567193031311035}, "2": {"mean": 0.6209525465965271, "std": 2.202680826187134}, "3": {"mean": 0.8415151238441467, "std": 1.3724360466003418}, "4": {"mean": 0.5678790211677551, "std": 0.9668241739273071}, "5": {"mean": 1.9048473834991455, "std": 1.917028546333313}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.301451712846756, "std": 1.3867390155792236}, "1": {"mean": -2.305941343307495, "std": 1.4053094387054443}, "2": {"mean": 1.3758469820022583, "std": 2.476078510284424}, "3": {"mean": -2.906585931777954, "std": 1.2520171403884888}, "4": {"mean": -1.1047377586364746, "std": 1.9995092153549194}, "5": {"mean": 0.17766894400119781, "std": 2.807504653930664}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": 0.2869175374507904, "std": 1.5099425315856934}, "1": {"mean": 0.23936127126216888, "std": 0.9337558150291443}, "2": {"mean": 0.39097848534584045, "std": 1.0250449180603027}, "3": {"mean": 0.36733710765838623, "std": 2.2030599117279053}, "4": {"mean": 0.42309531569480896, "std": 0.7711548209190369}, "5": {"mean": 0.16981926560401917, "std": 1.5755337476730347}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -0.16720148921012878, "std": 2.4689347743988037}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.464436, 0.059697, -0.060857, 0.008878, -0.718214], [0.927893, -0.102967, 0.4821, 0.502093, 0.844588], [0.033001, -0.149848, -0.531358, -0.133431, 0.032558], [-0.359709, -0.208272, -0.110354, 0.109899, -0.997615], [-0.483363, -0.053064, -0.785563, -0.317384, -0.241213], [-0.343298, 0.024055, 0.641648, 0.317941, -0.71603]], "network.0.bias": [0.150107, -0.00944, -0.151925, 0.294794, -0.418273, 0.588971], "network.2.weight": [[-0.545901, 0.219843, 0.190284, -0.471117, -0.420528, 0.561266], [-0.046149, -0.471315, 0.34015, 0.091078, -0.610142, 0.314573], [-0.17489, -0.032867, 0.548431, -0.440782, -0.403518, 0.650864], [0.195776, -0.233251, -0.573747, 0.319209, 0.00824, -0.551671], [0.840809, 0.546237, -0.656188, 0.389769, 0.132226, -0.408866], [0.068616, -0.32301, 0.601574, -0.479402, -0.309494, 0.18519]], "network.2.bias": [0.094973, -0.593491, -0.279573, 0.480178, -0.166321, -0.317859], "network.4.weight": [[0.573539, -0.680245, 0.021376, -0.011515, -0.422487, -0.409383], [-0.066458, 0.382134, -0.640746, -0.360315, 0.514724, 0.134499], [-0.020676, 0.876551, -0.091102, 0.492015, -0.601186, 0.831295], [0.268432, 0.132085, 0.258555, -0.243374, 0.267395, 0.121088], [0.00262, 0.749239, 0.009333, -0.570134, 0.491659, 0.330641], [0.615303, -0.707492, 0.370454, -0.172611, -0.557232, -0.243916]], "network.4.bias": [0.745371, 0.205346, -0.552032, 0.438788, -0.25857, 0.276047], "network.6.weight": [[-0.565896, 0.276435, 0.226537, 0.120812, 0.150941, -0.238155], [-0.55218, 0.340526, 0.085455, -0.021777, 0.32926, 0.065322], [-0.507543, 0.743805, 0.391856, 0.178472, 0.754199, -0.235355], [0.518088, -0.244548, -0.272654, -0.271572, -0.015592, 0.514299], [0.170567, -0.601852, -0.67458, 0.228103, -0.051648, 0.160235], [0.333393, -0.469323, -1.048334, 0.590447, -0.529015, 0.690185]], "network.6.bias": [-0.174143, 0.014366, 0.327382, 0.263931, 0.297951, 0.553342], "network.8.weight": [[0.054696, 0.093658, 0.323289, 0.043831, -0.311699, -0.361185], [0.098757, 0.086485, -0.064307, 0.408051, -0.46329, -0.843741], [-0.049604, -0.538165, -0.376293, 0.396918, 0.154561, 0.600477], [-0.418266, 0.13735, -0.286822, -0.043146, -0.21735, -0.749928], [-0.089895, -0.384989, -0.630976, 0.761804, 0.023292, -0.06299], [0.285879, 0.616335, 0.762927, -0.565309, -0.366133, -0.141611]], "network.8.bias": [0.174911, -0.737929, 0.262721, -0.920174, -0.95862, 0.225227], "network.10.weight": [[-0.806961, 0.453262, 0.511691, 0.488049, 0.547035, 0.044381], [-0.121057, 0.04253, 0.405723, 0.374557, -0.054169, -0.090894], [0.800571, -0.476509, -0.084915, -0.126801, 0.160394, 0.145339], [0.686041, -0.663819, -0.560478, 0.001515, 0.331394, 0.445718], [-0.319789, 0.420445, 0.04397, 0.768098, 0.952316, 0.530589], [0.202161, -0.476155, -0.309797, -0.701318, -0.348922, 0.4344]], "network.10.bias": [-0.467935, -0.338906, 0.019428, 0.524662, -0.282373, 0.156543], "network.12.weight": [[0.615514, 0.750628, -0.344055, -0.494552, -0.457473, -0.300621]], "network.12.bias": [0.130091]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7139422595500946, "train_acc": 0.425, "val_loss": 0.6979189515113831, "val_acc": 0.48}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6939050257205963, "train_acc": 0.575, "val_loss": 0.7200202941894531, "val_acc": 0.48}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6750213801860809, "train_acc": 0.575, "val_loss": 0.7078514099121094, "val_acc": 0.48}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6735662221908569, "train_acc": 0.575, "val_loss": 0.654421865940094, "val_acc": 0.48}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.5644418150186539, "train_acc": 0.605, "val_loss": 0.5362541079521179, "val_acc": 0.78}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.4868492931127548, "train_acc": 0.795, "val_loss": 0.6949689984321594, "val_acc": 0.74}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.6500429511070251, "train_acc": 0.665, "val_loss": 0.5019474625587463, "val_acc": 0.74}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.5028302371501923, "train_acc": 0.745, "val_loss": 0.5511242747306824, "val_acc": 0.72}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.5209252238273621, "train_acc": 0.73, "val_loss": 0.4984075427055359, "val_acc": 0.74}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.4779380261898041, "train_acc": 0.765, "val_loss": 0.46872684359550476, "val_acc": 0.76}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.4453800469636917, "train_acc": 0.78, "val_loss": 0.4475353956222534, "val_acc": 0.76}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.41000908613204956, "train_acc": 0.79, "val_loss": 0.4473569393157959, "val_acc": 0.78}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.4239147752523422, "train_acc": 0.795, "val_loss": 0.43936774134635925, "val_acc": 0.78}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.42949603497982025, "train_acc": 0.795, "val_loss": 0.41575583815574646, "val_acc": 0.78}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["palindrome"], "degraded_stage": {"initial_val_loss": 0.6979189515113831, "final_val_loss": 0.654421865940094, "initial_val_acc": 0.48, "final_val_acc": 0.48, "best_val_acc": 0.48}, "improved_stage": {"initial_val_loss": 0.5362541079521179, "final_val_loss": 0.41575583815574646, "initial_val_acc": 0.78, "final_val_acc": 0.78, "best_val_acc": 0.78, "best_epoch": 4}, "improvement": 0.30000000000000004, "first_improvement_epoch": 3}}
|
47
|
{"target_pattern": "sorted_descending", "degraded_accuracy": 0.58, "improved_accuracy": 0.94, "improvement": 0.36, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 8890, "learning_rate": 0.024940936964749232, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_descending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_descending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.356782,
-0.114368,
0.388695,
0.0225,
0.07697
],
[
-0.258201,
0.172004,
-0.400283,
-0.174749,
-0.148629
],
[
-0.424141,
0.033772,
-0.177687,
0.392342,
0.451003
],
[
-0.243412,
-0.184068,
0.089709,
-0.411058,
-0.291213
],
[
-0.592567,
0.148227,
0.029061,
0.248356,
-0.108177
],
[
-0.240633,
0.031292,
-0.080253,
-0.260962,
0.525364
]
],
"network.0.bias": [
0.411827,
-0.396781,
-0.111225,
-0.53357,
0.575339,
0.064377
],
"network.2.weight": [
[
0.315053,
-0.01305,
-0.081314,
0.080843,
0.127868,
0.492675
],
[
0.69171,
0.455601,
0.409934,
0.012752,
0.712808,
0.258743
],
[
0.256593,
-0.04099,
0.669469,
-0.257033,
0.277501,
0.35424
],
[
0.539187,
-0.036474,
0.416166,
0.253559,
0.68691,
0.56787
],
[
-0.027537,
0.348512,
-0.281783,
0.137982,
0.092078,
-0.252865
],
[
0.169158,
-0.093069,
0.342237,
-0.097016,
0.146391,
0.462733
]
],
"network.2.bias": [
-0.038517,
0.199279,
0.092833,
0.040946,
-0.36377,
-0.263126
],
"network.4.weight": [
[
0.451254,
0.586811,
0.428638,
0.723218,
0.408544,
0.423846
],
[
-0.218661,
0.282667,
0.346546,
0.394473,
0.108026,
0.176849
],
[
-0.014369,
0.540246,
0.318203,
0.769959,
-0.127441,
0.341876
],
[
-0.168781,
-0.291596,
-0.229519,
-0.041602,
-0.383007,
-0.315468
],
[
-0.509266,
-0.064677,
-0.065908,
0.001975,
-0.033708,
-0.139224
],
[
0.311669,
0.423818,
0.598522,
0.382115,
-0.164037,
0.402612
]
],
"network.4.bias": [
0.249965,
-0.052393,
-0.038265,
0.031034,
0.597992,
-0.473706
],
"network.6.weight": [
[
-0.308649,
-0.004422,
-0.406968,
-0.335836,
0.170091,
-0.185408
],
[
-0.366788,
-0.188306,
0.210597,
0.205696,
0.218181,
0.057221
],
[
0.644681,
0.463511,
0.355127,
-0.103441,
-0.569562,
0.749524
],
[
-0.102478,
0.303558,
-0.117362,
0.004879,
0.022575,
-0.083347
],
[
-0.244637,
-0.322644,
-0.130535,
0.403411,
-0.049113,
-0.289742
],
[
0.155518,
0.44968,
0.550862,
-0.102912,
-0.423359,
0.536511
]
],
"network.6.bias": [
-0.286294,
0.006503,
-0.486762,
0.489674,
-0.201882,
0.091018
],
"network.8.weight": [
[
-0.033688,
-0.167921,
0.06188,
0.392548,
0.253784,
-0.336089
],
[
0.32842,
-0.403077,
0.679236,
-0.532199,
-0.112339,
0.299232
],
[
0.348277,
-0.17135,
-0.198359,
0.099544,
-0.094617,
0.205367
],
[
-0.04037,
-0.314367,
0.390428,
-0.057744,
-0.254044,
0.513361
],
[
-0.098257,
-0.163206,
-0.058517,
-0.078942,
-0.383689,
0.360124
],
[
-0.292501,
-0.26428,
0.451976,
-0.42191,
0.053336,
0.274582
]
],
"network.8.bias": [
-0.252953,
-0.242472,
-0.151064,
-0.058772,
0.545435,
0.011753
],
"network.10.weight": [
[
0.246586,
-0.234395,
0.157002,
0.039836,
-0.309397,
-0.301392
],
[
-0.214676,
0.639252,
-0.114492,
0.621263,
0.116567,
0.403397
],
[
0.054172,
0.014607,
-0.103466,
-0.016177,
0.409209,
-0.222838
],
[
0.346692,
0.210409,
-0.010199,
-0.174851,
0.273365,
-0.247215
],
[
0.013331,
-0.223633,
0.014148,
-0.007528,
0.499425,
0.092755
],
[
0.304542,
-0.347307,
0.108237,
0.268554,
-0.193642,
0.066198
]
],
"network.10.bias": [
-0.004117,
-0.240454,
0.463467,
0.359456,
0.308284,
-0.292984
],
"network.12.weight": [
[
0.218058,
-0.486294,
0.516791,
0.325703,
0.50346,
0.116464
]
],
"network.12.bias": [
0.007671
]
}
## Activation Signature
### 0
mean: [0.714547, -1.798864, 0.448918, -2.225155, 0.567897, -0.257364]
std: [0.924508, 1.144781, 1.512526, 1.337756, 1.335449, 1.043614]
### 2
mean: [0.417321, 1.836123, 1.216758, 1.624305, -0.611215, 0.430355]
std: [0.393133, 1.176702, 1.037715, 1.180079, 0.391172, 0.656018]
### 4
mean: [3.418850, 1.521870, 2.750304, -1.074693, 0.119827, 1.979308]
std: [2.321629, 1.181866, 2.037041, 0.840829, 0.381606, 1.854123]
### 6
mean: [-2.804793, -0.790681, 4.789015, 0.114939, -2.485253, 3.813304]
std: [1.915330, 0.583269, 4.219172, 0.275759, 1.725504, 3.056029]
### 8
mean: [-1.163685, 4.098585, -0.312493, 3.782855, 1.620562, 3.175955]
std: [0.833278, 3.811296, 0.211058, 3.194903, 0.870731, 2.777306]
### 10
mean: [-2.297132, 6.263332, 0.412876, 0.225747, 0.452961, -0.825271]
std: [1.843543, 5.568081, 0.253965, 0.214592, 0.172343, 0.426885]
### 12
mean: [-2.521627]
std: [2.942414]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_descending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.356782,
-0.114368,
0.388695,
0.0225,
0.07697
],
[
-0.258201,
0.172004,
-0.400283,
-0.174749,
-0.148629
],
[
-0.424141,
0.033772,
-0.177687,
0.392342,
0.451003
],
[
-0.243412,
-0.184068,
0.089709,
-0.411058,
-0.291213
],
[
-0.592567,
0.148227,
0.029061,
0.248356,
-0.108177
],
[
-0.240633,
0.031292,
-0.080253,
-0.260962,
0.525364
]
],
"network.0.bias": [
0.411827,
-0.396781,
-0.111225,
-0.53357,
0.575339,
0.064377
],
"network.2.weight": [
[
0.315053,
-0.01305,
-0.081314,
0.080843,
0.127868,
0.492675
],
[
0.69171,
0.455601,
0.409934,
0.012752,
0.712808,
0.258743
],
[
0.256593,
-0.04099,
0.669469,
-0.257033,
0.277501,
0.35424
],
[
0.539187,
-0.036474,
0.416166,
0.253559,
0.68691,
0.56787
],
[
-0.027537,
0.348512,
-0.281783,
0.137982,
0.092078,
-0.252865
],
[
0.169158,
-0.093069,
0.342237,
-0.097016,
0.146391,
0.462733
]
],
"network.2.bias": [
-0.038517,
0.199279,
0.092833,
0.040946,
-0.36377,
-0.263126
],
"network.4.weight": [
[
0.451254,
0.586811,
0.428638,
0.723218,
0.408544,
0.423846
],
[
-0.218661,
0.282667,
0.346546,
0.394473,
0.108026,
0.176849
],
[
-0.014369,
0.540246,
0.318203,
0.769959,
-0.127441,
0.341876
],
[
-0.168781,
-0.291596,
-0.229519,
-0.041602,
-0.383007,
-0.315468
],
[
-0.509266,
-0.064677,
-0.065908,
0.001975,
-0.033708,
-0.139224
],
[
0.311669,
0.423818,
0.598522,
0.382115,
-0.164037,
0.402612
]
],
"network.4.bias": [
0.249965,
-0.052393,
-0.038265,
0.031034,
0.597992,
-0.473706
],
"network.6.weight": [
[
-0.308649,
-0.004422,
-0.406968,
-0.335836,
0.170091,
-0.185408
],
[
-0.366788,
-0.188306,
0.210597,
0.205696,
0.218181,
0.057221
],
[
0.644681,
0.463511,
0.355127,
-0.103441,
-0.569562,
0.749524
],
[
-0.102478,
0.303558,
-0.117362,
0.004879,
0.022575,
-0.083347
],
[
-0.244637,
-0.322644,
-0.130535,
0.403411,
-0.049113,
-0.289742
],
[
0.155518,
0.44968,
0.550862,
-0.102912,
-0.423359,
0.536511
]
],
"network.6.bias": [
-0.286294,
0.006503,
-0.486762,
0.489674,
-0.201882,
0.091018
],
"network.8.weight": [
[
-0.033688,
-0.167921,
0.06188,
0.392548,
0.253784,
-0.336089
],
[
0.32842,
-0.403077,
0.679236,
-0.532199,
-0.112339,
0.299232
],
[
0.348277,
-0.17135,
-0.198359,
0.099544,
-0.094617,
0.205367
],
[
-0.04037,
-0.314367,
0.390428,
-0.057744,
-0.254044,
0.513361
],
[
-0.098257,
-0.163206,
-0.058517,
-0.078942,
-0.383689,
0.360124
],
[
-0.292501,
-0.26428,
0.451976,
-0.42191,
0.053336,
0.274582
]
],
"network.8.bias": [
-0.252953,
-0.242472,
-0.151064,
-0.058772,
0.545435,
0.011753
],
"network.10.weight": [
[
0.246586,
-0.234395,
0.157002,
0.039836,
-0.309397,
-0.301392
],
[
-0.214676,
0.639252,
-0.114492,
0.621263,
0.116567,
0.403397
],
[
0.054172,
0.014607,
-0.103466,
-0.016177,
0.409209,
-0.222838
],
[
0.346692,
0.210409,
-0.010199,
-0.174851,
0.273365,
-0.247215
],
[
0.013331,
-0.223633,
0.014148,
-0.007528,
0.499425,
0.092755
],
[
0.304542,
-0.347307,
0.108237,
0.268554,
-0.193642,
0.066198
]
],
"network.10.bias": [
-0.004117,
-0.240454,
0.463467,
0.359456,
0.308284,
-0.292984
],
"network.12.weight": [
[
0.218058,
-0.486294,
0.516791,
0.325703,
0.50346,
0.116464
]
],
"network.12.bias": [
0.007671
]
}
## Activation Signature
### 0
mean: [0.714547, -1.798864, 0.448918, -2.225155, 0.567897, -0.257364]
std: [0.924508, 1.144781, 1.512526, 1.337756, 1.335449, 1.043614]
### 2
mean: [0.417321, 1.836123, 1.216758, 1.624305, -0.611215, 0.430355]
std: [0.393133, 1.176702, 1.037715, 1.180079, 0.391172, 0.656018]
### 4
mean: [3.418850, 1.521870, 2.750304, -1.074693, 0.119827, 1.979308]
std: [2.321629, 1.181866, 2.037041, 0.840829, 0.381606, 1.854123]
### 6
mean: [-2.804793, -0.790681, 4.789015, 0.114939, -2.485253, 3.813304]
std: [1.915330, 0.583269, 4.219172, 0.275759, 1.725504, 3.056029]
### 8
mean: [-1.163685, 4.098585, -0.312493, 3.782855, 1.620562, 3.175955]
std: [0.833278, 3.811296, 0.211058, 3.194903, 0.870731, 2.777306]
### 10
mean: [-2.297132, 6.263332, 0.412876, 0.225747, 0.452961, -0.825271]
std: [1.843543, 5.568081, 0.253965, 0.214592, 0.172343, 0.426885]
### 12
mean: [-2.521627]
std: [2.942414]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_descending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.7145472168922424, "std": 0.9245082139968872}, "1": {"mean": -1.7988641262054443, "std": 1.1447813510894775}, "2": {"mean": 0.4489181935787201, "std": 1.5125256776809692}, "3": {"mean": -2.2251551151275635, "std": 1.3377562761306763}, "4": {"mean": 0.5678970217704773, "std": 1.335448980331421}, "5": {"mean": -0.25736409425735474, "std": 1.0436137914657593}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.4173206388950348, "std": 0.3931327164173126}, "1": {"mean": 1.8361234664916992, "std": 1.1767024993896484}, "2": {"mean": 1.2167582511901855, "std": 1.0377154350280762}, "3": {"mean": 1.624305009841919, "std": 1.1800791025161743}, "4": {"mean": -0.6112152934074402, "std": 0.39117199182510376}, "5": {"mean": 0.4303554594516754, "std": 0.6560181379318237}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 3.4188499450683594, "std": 2.3216288089752197}, "1": {"mean": 1.5218702554702759, "std": 1.1818655729293823}, "2": {"mean": 2.7503037452697754, "std": 2.037041425704956}, "3": {"mean": -1.0746928453445435, "std": 0.8408288359642029}, "4": {"mean": 0.11982663720846176, "std": 0.38160616159439087}, "5": {"mean": 1.979308009147644, "std": 1.8541227579116821}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -2.804792881011963, "std": 1.9153295755386353}, "1": {"mean": -0.7906813025474548, "std": 0.5832691192626953}, "2": {"mean": 4.78901481628418, "std": 4.219171524047852}, "3": {"mean": 0.11493927985429764, "std": 0.27575868368148804}, "4": {"mean": -2.485252618789673, "std": 1.7255041599273682}, "5": {"mean": 3.8133039474487305, "std": 3.0560288429260254}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.1636853218078613, "std": 0.8332776427268982}, "1": {"mean": 4.09858512878418, "std": 3.811295509338379}, "2": {"mean": -0.31249305605888367, "std": 0.21105791628360748}, "3": {"mean": 3.782855272293091, "std": 3.1949033737182617}, "4": {"mean": 1.620562195777893, "std": 0.870730996131897}, "5": {"mean": 3.175954818725586, "std": 2.777305841445923}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -2.2971317768096924, "std": 1.8435429334640503}, "1": {"mean": 6.263331890106201, "std": 5.568080902099609}, "2": {"mean": 0.4128757119178772, "std": 0.2539650797843933}, "3": {"mean": 0.22574661672115326, "std": 0.21459153294563293}, "4": {"mean": 0.4529605507850647, "std": 0.1723431646823883}, "5": {"mean": -0.8252714276313782, "std": 0.4268849492073059}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -2.521627187728882, "std": 2.942413806915283}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.356782, -0.114368, 0.388695, 0.0225, 0.07697], [-0.258201, 0.172004, -0.400283, -0.174749, -0.148629], [-0.424141, 0.033772, -0.177687, 0.392342, 0.451003], [-0.243412, -0.184068, 0.089709, -0.411058, -0.291213], [-0.592567, 0.148227, 0.029061, 0.248356, -0.108177], [-0.240633, 0.031292, -0.080253, -0.260962, 0.525364]], "network.0.bias": [0.411827, -0.396781, -0.111225, -0.53357, 0.575339, 0.064377], "network.2.weight": [[0.315053, -0.01305, -0.081314, 0.080843, 0.127868, 0.492675], [0.69171, 0.455601, 0.409934, 0.012752, 0.712808, 0.258743], [0.256593, -0.04099, 0.669469, -0.257033, 0.277501, 0.35424], [0.539187, -0.036474, 0.416166, 0.253559, 0.68691, 0.56787], [-0.027537, 0.348512, -0.281783, 0.137982, 0.092078, -0.252865], [0.169158, -0.093069, 0.342237, -0.097016, 0.146391, 0.462733]], "network.2.bias": [-0.038517, 0.199279, 0.092833, 0.040946, -0.36377, -0.263126], "network.4.weight": [[0.451254, 0.586811, 0.428638, 0.723218, 0.408544, 0.423846], [-0.218661, 0.282667, 0.346546, 0.394473, 0.108026, 0.176849], [-0.014369, 0.540246, 0.318203, 0.769959, -0.127441, 0.341876], [-0.168781, -0.291596, -0.229519, -0.041602, -0.383007, -0.315468], [-0.509266, -0.064677, -0.065908, 0.001975, -0.033708, -0.139224], [0.311669, 0.423818, 0.598522, 0.382115, -0.164037, 0.402612]], "network.4.bias": [0.249965, -0.052393, -0.038265, 0.031034, 0.597992, -0.473706], "network.6.weight": [[-0.308649, -0.004422, -0.406968, -0.335836, 0.170091, -0.185408], [-0.366788, -0.188306, 0.210597, 0.205696, 0.218181, 0.057221], [0.644681, 0.463511, 0.355127, -0.103441, -0.569562, 0.749524], [-0.102478, 0.303558, -0.117362, 0.004879, 0.022575, -0.083347], [-0.244637, -0.322644, -0.130535, 0.403411, -0.049113, -0.289742], [0.155518, 0.44968, 0.550862, -0.102912, -0.423359, 0.536511]], "network.6.bias": [-0.286294, 0.006503, -0.486762, 0.489674, -0.201882, 0.091018], "network.8.weight": [[-0.033688, -0.167921, 0.06188, 0.392548, 0.253784, -0.336089], [0.32842, -0.403077, 0.679236, -0.532199, -0.112339, 0.299232], [0.348277, -0.17135, -0.198359, 0.099544, -0.094617, 0.205367], [-0.04037, -0.314367, 0.390428, -0.057744, -0.254044, 0.513361], [-0.098257, -0.163206, -0.058517, -0.078942, -0.383689, 0.360124], [-0.292501, -0.26428, 0.451976, -0.42191, 0.053336, 0.274582]], "network.8.bias": [-0.252953, -0.242472, -0.151064, -0.058772, 0.545435, 0.011753], "network.10.weight": [[0.246586, -0.234395, 0.157002, 0.039836, -0.309397, -0.301392], [-0.214676, 0.639252, -0.114492, 0.621263, 0.116567, 0.403397], [0.054172, 0.014607, -0.103466, -0.016177, 0.409209, -0.222838], [0.346692, 0.210409, -0.010199, -0.174851, 0.273365, -0.247215], [0.013331, -0.223633, 0.014148, -0.007528, 0.499425, 0.092755], [0.304542, -0.347307, 0.108237, 0.268554, -0.193642, 0.066198]], "network.10.bias": [-0.004117, -0.240454, 0.463467, 0.359456, 0.308284, -0.292984], "network.12.weight": [[0.218058, -0.486294, 0.516791, 0.325703, 0.50346, 0.116464]], "network.12.bias": [0.007671]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6983940899372101, "train_acc": 0.445, "val_loss": 0.6842834949493408, "val_acc": 0.58}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6819095015525818, "train_acc": 0.555, "val_loss": 0.6789032816886902, "val_acc": 0.58}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6887918710708618, "train_acc": 0.555, "val_loss": 0.6685178875923157, "val_acc": 0.58}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6714825332164764, "train_acc": 0.555, "val_loss": 0.6404300928115845, "val_acc": 0.58}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6451559960842133, "train_acc": 0.48, "val_loss": 0.5715429186820984, "val_acc": 0.58}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.5649116039276123, "train_acc": 0.48, "val_loss": 0.4622061252593994, "val_acc": 0.76}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.45433464646339417, "train_acc": 0.81, "val_loss": 0.3606742024421692, "val_acc": 0.9}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.37603385746479034, "train_acc": 0.905, "val_loss": 0.2900390923023224, "val_acc": 0.92}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.3053114265203476, "train_acc": 0.935, "val_loss": 0.24399425089359283, "val_acc": 0.94}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.25601428747177124, "train_acc": 0.95, "val_loss": 0.2400033175945282, "val_acc": 0.92}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.21074841171503067, "train_acc": 0.955, "val_loss": 0.2236010730266571, "val_acc": 0.92}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.19146382063627243, "train_acc": 0.95, "val_loss": 0.22229649126529694, "val_acc": 0.92}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.18340853601694107, "train_acc": 0.955, "val_loss": 0.23484604060649872, "val_acc": 0.92}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.13374072685837746, "train_acc": 0.96, "val_loss": 0.25680673122406006, "val_acc": 0.92}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["sorted_descending"], "degraded_stage": {"initial_val_loss": 0.6842834949493408, "final_val_loss": 0.6404300928115845, "initial_val_acc": 0.58, "final_val_acc": 0.58, "best_val_acc": 0.58}, "improved_stage": {"initial_val_loss": 0.5715429186820984, "final_val_loss": 0.25680673122406006, "initial_val_acc": 0.58, "final_val_acc": 0.92, "best_val_acc": 0.94, "best_epoch": 8}, "improvement": 0.36, "first_improvement_epoch": 3}}
|
48
|
{"target_pattern": "starts_with", "degraded_accuracy": 0.54, "improved_accuracy": 0.84, "improvement": 0.29999999999999993, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 1794, "learning_rate": 0.045307675361688976, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "starts_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["starts_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.518453,
0.348498,
0.149584,
0.18818,
-0.234792
],
[
-0.150008,
0.079428,
-0.965357,
0.526816,
-0.000901
],
[
0.059471,
-0.124936,
-0.001346,
-0.20693,
-0.722125
],
[
0.507711,
-0.050385,
0.499253,
0.391174,
-0.326648
],
[
0.661653,
0.410395,
-0.202681,
0.443128,
0.035083
],
[
-0.763264,
0.086048,
0.001268,
0.058293,
0.451107
],
[
-0.935944,
0.357581,
0.345467,
-0.196453,
0.510239
]
],
"network.0.bias": [
-0.289146,
0.155926,
-0.163738,
0.010114,
0.319281,
0.01113,
0.227316
],
"network.2.weight": [
[
0.400621,
-0.609714,
-0.456112,
-0.194599,
0.299933,
0.568818,
-0.038762
],
[
0.558499,
0.040502,
-0.919904,
0.33336,
0.245942,
-0.305056,
-0.240802
],
[
-0.033576,
0.181506,
0.008797,
-0.419204,
-0.351026,
0.031169,
0.394813
],
[
0.371338,
-0.866501,
-1.146922,
0.431615,
0.323808,
-0.617625,
-0.068534
],
[
0.174492,
-0.017196,
-0.923157,
-0.02803,
-0.017421,
-0.328529,
-0.037284
],
[
-0.201915,
-0.280831,
-0.691909,
0.250992,
-0.002043,
0.260099,
-0.326558
],
[
0.067683,
0.687909,
0.37413,
0.081219,
-0.173798,
0.27217,
0.649567
]
],
"network.2.bias": [
-0.114387,
0.187271,
0.055868,
-0.183958,
0.078739,
0.098607,
-0.034757
],
"network.4.weight": [
[
-0.499883,
-0.465327,
0.270218,
-0.058795,
-0.050893,
-0.507533,
0.346807
],
[
-0.674912,
-0.220082,
0.479812,
-0.409519,
-0.622506,
-0.54603,
0.251079
],
[
-0.263415,
0.282761,
-0.223946,
-0.056261,
-0.166901,
0.313941,
-0.271695
],
[
-0.343312,
-0.052163,
0.170436,
0.230198,
0.393837,
-0.415358,
-0.219208
],
[
-0.077205,
-0.229838,
0.110656,
-0.472846,
-0.415375,
-0.431208,
0.588861
],
[
0.470735,
-0.459169,
-0.093006,
-0.520355,
0.192627,
-0.317231,
-0.512878
],
[
-0.087625,
0.151517,
-0.513556,
0.457805,
0.069794,
0.30387,
0.024623
]
],
"network.4.bias": [
-0.013065,
-0.038966,
0.356093,
-0.093346,
0.484372,
-0.064617,
0.042125
],
"network.6.weight": [
[
-0.45858,
0.076768,
-8e-06,
-0.223189,
-0.253211,
0.505467,
0.031607
],
[
0.22525,
0.160807,
-0.40123,
-0.123889,
0.066317,
0.00511,
0.199316
],
[
0.289978,
0.098249,
-0.076922,
0.101123,
0.51772,
0.21559,
-0.431066
],
[
0.561735,
0.438903,
0.009627,
-0.103372,
0.774212,
-0.441492,
-0.143595
],
[
0.416224,
0.52617,
-0.256236,
-0.360663,
0.722321,
0.146929,
0.04814
],
[
-0.061762,
0.163537,
-0.143404,
-0.001202,
0.371629,
0.18019,
-0.252774
],
[
0.125738,
0.357417,
-0.197662,
0.187183,
0.405545,
0.527277,
-0.188577
]
],
"network.6.bias": [
-0.412153,
0.045196,
0.181302,
-0.13051,
-0.01444,
0.18062,
0.000726
],
"network.8.weight": [
[
-0.202452,
0.13042,
0.104038,
0.328458,
0.209057,
0.102052,
0.357211
],
[
-0.550533,
-0.47114,
-0.723945,
-0.722232,
-0.649395,
-0.713737,
-0.662245
],
[
-0.305103,
-0.379742,
-0.299419,
-0.616766,
-0.555768,
-0.7506,
-0.424703
],
[
-0.40082,
-0.230184,
-0.981498,
-0.884737,
-0.046239,
-0.958306,
-0.565381
],
[
0.219886,
0.298392,
0.041766,
0.328851,
0.503184,
0.130426,
0.357594
],
[
-0.130105,
-0.069777,
-0.260473,
-0.229986,
-0.156904,
-0.08927,
-0.043015
],
[
-0.656211,
-0.620076,
-0.838098,
-0.90056,
-0.267213,
-0.61801,
-0.702823
]
],
"network.8.bias": [
-0.164358,
0.679821,
0.387242,
0.575097,
0.088601,
-0.373183,
0.59699
],
"network.10.weight": [
[
0.269034,
-0.911412,
-0.631402,
-0.776908,
0.229792,
0.307208,
-0.924629
]
],
"network.10.bias": [
-0.407731
]
}
## Activation Signature
### 0
mean: [1.417511, -0.773923, -1.654573, 2.028987, 2.460657, -0.097889, 0.643833]
std: [1.591706, 2.268183, 1.481936, 1.683021, 1.994478, 1.600541, 1.831909]
### 2
mean: [0.747112, 1.967131, -1.199269, 1.376405, 0.122566, 0.032928, 0.838598]
std: [0.988110, 2.024141, 1.559529, 2.263649, 0.441786, 0.424933, 1.146067]
### 4
mean: [-1.136229, -1.510299, 0.404384, -0.239911, -0.366377, -1.887689, 1.037903]
std: [1.677410, 2.108935, 0.585741, 0.387242, 1.966011, 1.363667, 1.210856]
### 6
mean: [-0.546725, 0.145412, -0.030809, 0.179842, 0.330794, 0.044160, -0.095719]
std: [0.323269, 0.208621, 0.885069, 0.943786, 0.822729, 0.572582, 0.589439]
### 8
mean: [0.114236, -0.048137, -0.193481, -0.074649, 0.418121, -0.561283, -0.069222]
std: [0.671276, 2.037079, 1.557278, 1.877277, 0.889509, 0.506333, 1.947671]
### 10
mean: [-1.881466]
std: [1.548390]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
starts_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.518453,
0.348498,
0.149584,
0.18818,
-0.234792
],
[
-0.150008,
0.079428,
-0.965357,
0.526816,
-0.000901
],
[
0.059471,
-0.124936,
-0.001346,
-0.20693,
-0.722125
],
[
0.507711,
-0.050385,
0.499253,
0.391174,
-0.326648
],
[
0.661653,
0.410395,
-0.202681,
0.443128,
0.035083
],
[
-0.763264,
0.086048,
0.001268,
0.058293,
0.451107
],
[
-0.935944,
0.357581,
0.345467,
-0.196453,
0.510239
]
],
"network.0.bias": [
-0.289146,
0.155926,
-0.163738,
0.010114,
0.319281,
0.01113,
0.227316
],
"network.2.weight": [
[
0.400621,
-0.609714,
-0.456112,
-0.194599,
0.299933,
0.568818,
-0.038762
],
[
0.558499,
0.040502,
-0.919904,
0.33336,
0.245942,
-0.305056,
-0.240802
],
[
-0.033576,
0.181506,
0.008797,
-0.419204,
-0.351026,
0.031169,
0.394813
],
[
0.371338,
-0.866501,
-1.146922,
0.431615,
0.323808,
-0.617625,
-0.068534
],
[
0.174492,
-0.017196,
-0.923157,
-0.02803,
-0.017421,
-0.328529,
-0.037284
],
[
-0.201915,
-0.280831,
-0.691909,
0.250992,
-0.002043,
0.260099,
-0.326558
],
[
0.067683,
0.687909,
0.37413,
0.081219,
-0.173798,
0.27217,
0.649567
]
],
"network.2.bias": [
-0.114387,
0.187271,
0.055868,
-0.183958,
0.078739,
0.098607,
-0.034757
],
"network.4.weight": [
[
-0.499883,
-0.465327,
0.270218,
-0.058795,
-0.050893,
-0.507533,
0.346807
],
[
-0.674912,
-0.220082,
0.479812,
-0.409519,
-0.622506,
-0.54603,
0.251079
],
[
-0.263415,
0.282761,
-0.223946,
-0.056261,
-0.166901,
0.313941,
-0.271695
],
[
-0.343312,
-0.052163,
0.170436,
0.230198,
0.393837,
-0.415358,
-0.219208
],
[
-0.077205,
-0.229838,
0.110656,
-0.472846,
-0.415375,
-0.431208,
0.588861
],
[
0.470735,
-0.459169,
-0.093006,
-0.520355,
0.192627,
-0.317231,
-0.512878
],
[
-0.087625,
0.151517,
-0.513556,
0.457805,
0.069794,
0.30387,
0.024623
]
],
"network.4.bias": [
-0.013065,
-0.038966,
0.356093,
-0.093346,
0.484372,
-0.064617,
0.042125
],
"network.6.weight": [
[
-0.45858,
0.076768,
-8e-06,
-0.223189,
-0.253211,
0.505467,
0.031607
],
[
0.22525,
0.160807,
-0.40123,
-0.123889,
0.066317,
0.00511,
0.199316
],
[
0.289978,
0.098249,
-0.076922,
0.101123,
0.51772,
0.21559,
-0.431066
],
[
0.561735,
0.438903,
0.009627,
-0.103372,
0.774212,
-0.441492,
-0.143595
],
[
0.416224,
0.52617,
-0.256236,
-0.360663,
0.722321,
0.146929,
0.04814
],
[
-0.061762,
0.163537,
-0.143404,
-0.001202,
0.371629,
0.18019,
-0.252774
],
[
0.125738,
0.357417,
-0.197662,
0.187183,
0.405545,
0.527277,
-0.188577
]
],
"network.6.bias": [
-0.412153,
0.045196,
0.181302,
-0.13051,
-0.01444,
0.18062,
0.000726
],
"network.8.weight": [
[
-0.202452,
0.13042,
0.104038,
0.328458,
0.209057,
0.102052,
0.357211
],
[
-0.550533,
-0.47114,
-0.723945,
-0.722232,
-0.649395,
-0.713737,
-0.662245
],
[
-0.305103,
-0.379742,
-0.299419,
-0.616766,
-0.555768,
-0.7506,
-0.424703
],
[
-0.40082,
-0.230184,
-0.981498,
-0.884737,
-0.046239,
-0.958306,
-0.565381
],
[
0.219886,
0.298392,
0.041766,
0.328851,
0.503184,
0.130426,
0.357594
],
[
-0.130105,
-0.069777,
-0.260473,
-0.229986,
-0.156904,
-0.08927,
-0.043015
],
[
-0.656211,
-0.620076,
-0.838098,
-0.90056,
-0.267213,
-0.61801,
-0.702823
]
],
"network.8.bias": [
-0.164358,
0.679821,
0.387242,
0.575097,
0.088601,
-0.373183,
0.59699
],
"network.10.weight": [
[
0.269034,
-0.911412,
-0.631402,
-0.776908,
0.229792,
0.307208,
-0.924629
]
],
"network.10.bias": [
-0.407731
]
}
## Activation Signature
### 0
mean: [1.417511, -0.773923, -1.654573, 2.028987, 2.460657, -0.097889, 0.643833]
std: [1.591706, 2.268183, 1.481936, 1.683021, 1.994478, 1.600541, 1.831909]
### 2
mean: [0.747112, 1.967131, -1.199269, 1.376405, 0.122566, 0.032928, 0.838598]
std: [0.988110, 2.024141, 1.559529, 2.263649, 0.441786, 0.424933, 1.146067]
### 4
mean: [-1.136229, -1.510299, 0.404384, -0.239911, -0.366377, -1.887689, 1.037903]
std: [1.677410, 2.108935, 0.585741, 0.387242, 1.966011, 1.363667, 1.210856]
### 6
mean: [-0.546725, 0.145412, -0.030809, 0.179842, 0.330794, 0.044160, -0.095719]
std: [0.323269, 0.208621, 0.885069, 0.943786, 0.822729, 0.572582, 0.589439]
### 8
mean: [0.114236, -0.048137, -0.193481, -0.074649, 0.418121, -0.561283, -0.069222]
std: [0.671276, 2.037079, 1.557278, 1.877277, 0.889509, 0.506333, 1.947671]
### 10
mean: [-1.881466]
std: [1.548390]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
starts_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.4175111055374146, "std": 1.5917062759399414}, "1": {"mean": -0.7739226222038269, "std": 2.2681829929351807}, "2": {"mean": -1.6545730829238892, "std": 1.481935977935791}, "3": {"mean": 2.028987407684326, "std": 1.6830211877822876}, "4": {"mean": 2.4606566429138184, "std": 1.9944782257080078}, "5": {"mean": -0.09788919240236282, "std": 1.6005414724349976}, "6": {"mean": 0.6438329219818115, "std": 1.831908941268921}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.7471120357513428, "std": 0.9881104230880737}, "1": {"mean": 1.9671306610107422, "std": 2.024141311645508}, "2": {"mean": -1.199269413948059, "std": 1.559529185295105}, "3": {"mean": 1.3764052391052246, "std": 2.2636489868164062}, "4": {"mean": 0.12256602942943573, "std": 0.44178593158721924}, "5": {"mean": 0.03292805328965187, "std": 0.4249332547187805}, "6": {"mean": 0.8385975360870361, "std": 1.1460669040679932}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -1.1362290382385254, "std": 1.6774102449417114}, "1": {"mean": -1.5102986097335815, "std": 2.1089348793029785}, "2": {"mean": 0.4043835997581482, "std": 0.5857413411140442}, "3": {"mean": -0.23991072177886963, "std": 0.387241929769516}, "4": {"mean": -0.36637696623802185, "std": 1.9660109281539917}, "5": {"mean": -1.887689232826233, "std": 1.3636671304702759}, "6": {"mean": 1.0379027128219604, "std": 1.210856318473816}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.5467251539230347, "std": 0.3232686221599579}, "1": {"mean": 0.14541222155094147, "std": 0.20862098038196564}, "2": {"mean": -0.030808838084340096, "std": 0.8850694298744202}, "3": {"mean": 0.17984218895435333, "std": 0.9437856078147888}, "4": {"mean": 0.330793559551239, "std": 0.8227286338806152}, "5": {"mean": 0.044160280376672745, "std": 0.5725817084312439}, "6": {"mean": -0.09571931511163712, "std": 0.5894391536712646}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.1142364889383316, "std": 0.6712757349014282}, "1": {"mean": -0.04813683405518532, "std": 2.037078619003296}, "2": {"mean": -0.19348114728927612, "std": 1.5572783946990967}, "3": {"mean": -0.0746489018201828, "std": 1.8772772550582886}, "4": {"mean": 0.4181213080883026, "std": 0.8895086646080017}, "5": {"mean": -0.5612825155258179, "std": 0.5063334107398987}, "6": {"mean": -0.06922226399183273, "std": 1.947670578956604}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -1.881466031074524, "std": 1.54839026927948}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.518453, 0.348498, 0.149584, 0.18818, -0.234792], [-0.150008, 0.079428, -0.965357, 0.526816, -0.000901], [0.059471, -0.124936, -0.001346, -0.20693, -0.722125], [0.507711, -0.050385, 0.499253, 0.391174, -0.326648], [0.661653, 0.410395, -0.202681, 0.443128, 0.035083], [-0.763264, 0.086048, 0.001268, 0.058293, 0.451107], [-0.935944, 0.357581, 0.345467, -0.196453, 0.510239]], "network.0.bias": [-0.289146, 0.155926, -0.163738, 0.010114, 0.319281, 0.01113, 0.227316], "network.2.weight": [[0.400621, -0.609714, -0.456112, -0.194599, 0.299933, 0.568818, -0.038762], [0.558499, 0.040502, -0.919904, 0.33336, 0.245942, -0.305056, -0.240802], [-0.033576, 0.181506, 0.008797, -0.419204, -0.351026, 0.031169, 0.394813], [0.371338, -0.866501, -1.146922, 0.431615, 0.323808, -0.617625, -0.068534], [0.174492, -0.017196, -0.923157, -0.02803, -0.017421, -0.328529, -0.037284], [-0.201915, -0.280831, -0.691909, 0.250992, -0.002043, 0.260099, -0.326558], [0.067683, 0.687909, 0.37413, 0.081219, -0.173798, 0.27217, 0.649567]], "network.2.bias": [-0.114387, 0.187271, 0.055868, -0.183958, 0.078739, 0.098607, -0.034757], "network.4.weight": [[-0.499883, -0.465327, 0.270218, -0.058795, -0.050893, -0.507533, 0.346807], [-0.674912, -0.220082, 0.479812, -0.409519, -0.622506, -0.54603, 0.251079], [-0.263415, 0.282761, -0.223946, -0.056261, -0.166901, 0.313941, -0.271695], [-0.343312, -0.052163, 0.170436, 0.230198, 0.393837, -0.415358, -0.219208], [-0.077205, -0.229838, 0.110656, -0.472846, -0.415375, -0.431208, 0.588861], [0.470735, -0.459169, -0.093006, -0.520355, 0.192627, -0.317231, -0.512878], [-0.087625, 0.151517, -0.513556, 0.457805, 0.069794, 0.30387, 0.024623]], "network.4.bias": [-0.013065, -0.038966, 0.356093, -0.093346, 0.484372, -0.064617, 0.042125], "network.6.weight": [[-0.45858, 0.076768, -8e-06, -0.223189, -0.253211, 0.505467, 0.031607], [0.22525, 0.160807, -0.40123, -0.123889, 0.066317, 0.00511, 0.199316], [0.289978, 0.098249, -0.076922, 0.101123, 0.51772, 0.21559, -0.431066], [0.561735, 0.438903, 0.009627, -0.103372, 0.774212, -0.441492, -0.143595], [0.416224, 0.52617, -0.256236, -0.360663, 0.722321, 0.146929, 0.04814], [-0.061762, 0.163537, -0.143404, -0.001202, 0.371629, 0.18019, -0.252774], [0.125738, 0.357417, -0.197662, 0.187183, 0.405545, 0.527277, -0.188577]], "network.6.bias": [-0.412153, 0.045196, 0.181302, -0.13051, -0.01444, 0.18062, 0.000726], "network.8.weight": [[-0.202452, 0.13042, 0.104038, 0.328458, 0.209057, 0.102052, 0.357211], [-0.550533, -0.47114, -0.723945, -0.722232, -0.649395, -0.713737, -0.662245], [-0.305103, -0.379742, -0.299419, -0.616766, -0.555768, -0.7506, -0.424703], [-0.40082, -0.230184, -0.981498, -0.884737, -0.046239, -0.958306, -0.565381], [0.219886, 0.298392, 0.041766, 0.328851, 0.503184, 0.130426, 0.357594], [-0.130105, -0.069777, -0.260473, -0.229986, -0.156904, -0.08927, -0.043015], [-0.656211, -0.620076, -0.838098, -0.90056, -0.267213, -0.61801, -0.702823]], "network.8.bias": [-0.164358, 0.679821, 0.387242, 0.575097, 0.088601, -0.373183, 0.59699], "network.10.weight": [[0.269034, -0.911412, -0.631402, -0.776908, 0.229792, 0.307208, -0.924629]], "network.10.bias": [-0.407731]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6798089444637299, "train_acc": 0.58, "val_loss": 0.7518340945243835, "val_acc": 0.44}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.7109754383563995, "train_acc": 0.58, "val_loss": 0.7642511129379272, "val_acc": 0.44}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6812123358249664, "train_acc": 0.58, "val_loss": 0.7198202610015869, "val_acc": 0.44}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6589332520961761, "train_acc": 0.58, "val_loss": 0.684438169002533, "val_acc": 0.54}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6571219563484192, "train_acc": 0.66, "val_loss": 0.6297293305397034, "val_acc": 0.62}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.5822120606899261, "train_acc": 0.73, "val_loss": 0.590467631816864, "val_acc": 0.72}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.5826823711395264, "train_acc": 0.73, "val_loss": 0.6569777131080627, "val_acc": 0.64}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.5276804268360138, "train_acc": 0.745, "val_loss": 0.5430827140808105, "val_acc": 0.76}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.5107884407043457, "train_acc": 0.75, "val_loss": 0.5221885442733765, "val_acc": 0.78}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.5024775117635727, "train_acc": 0.745, "val_loss": 0.5052313804626465, "val_acc": 0.8}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.4863526225090027, "train_acc": 0.765, "val_loss": 0.43675899505615234, "val_acc": 0.8}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.483779639005661, "train_acc": 0.76, "val_loss": 0.40414467453956604, "val_acc": 0.84}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.45416484773159027, "train_acc": 0.76, "val_loss": 0.3758879005908966, "val_acc": 0.8}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.4263816177845001, "train_acc": 0.77, "val_loss": 0.3318997919559479, "val_acc": 0.82}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["starts_with"], "degraded_stage": {"initial_val_loss": 0.7518340945243835, "final_val_loss": 0.684438169002533, "initial_val_acc": 0.44, "final_val_acc": 0.54, "best_val_acc": 0.54}, "improved_stage": {"initial_val_loss": 0.6297293305397034, "final_val_loss": 0.3318997919559479, "initial_val_acc": 0.62, "final_val_acc": 0.82, "best_val_acc": 0.84, "best_epoch": 11}, "improvement": 0.29999999999999993, "first_improvement_epoch": 3}}
|
49
|
{"target_pattern": "alternating", "degraded_accuracy": 0.46, "improved_accuracy": 0.94, "improvement": 0.4799999999999999, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 3911, "learning_rate": 0.03942708661657638, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "alternating", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["alternating"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.225213,
-0.595051,
-0.593457,
0.099706,
-0.551871
],
[
-0.072744,
0.170267,
0.795061,
-0.233176,
0.077886
],
[
-0.093697,
0.432099,
0.748383,
-0.104163,
-0.15884
],
[
-0.165408,
-0.365347,
-0.110227,
0.048167,
0.671099
],
[
0.375296,
0.290414,
-0.529392,
0.128531,
0.422757
],
[
0.67084,
0.341363,
-0.216221,
-0.063743,
-0.145314
],
[
0.574614,
-1.200751,
0.031782,
-0.952072,
0.429675
]
],
"network.0.bias": [
-0.573225,
0.260186,
0.375742,
0.002737,
0.418112,
0.334011,
0.548444
],
"network.2.weight": [
[
0.019987,
0.728415,
0.542887,
-0.336374,
-0.263188,
0.223773,
-0.406385
],
[
0.23132,
-0.00197,
-0.397628,
0.514484,
0.843187,
-0.663505,
-0.361922
],
[
-0.176525,
-0.061979,
-0.490691,
-0.172159,
-0.062854,
-0.036306,
-0.370737
],
[
-0.210941,
0.568025,
0.414293,
-0.202623,
-0.271204,
-0.034746,
-0.3833
],
[
0.123682,
0.137596,
0.538902,
-0.073057,
-0.50579,
-0.186668,
-0.115042
],
[
0.056078,
-0.104216,
-0.126462,
-0.013606,
0.217423,
0.663964,
0.759508
],
[
0.003111,
-0.159389,
-0.268267,
-0.001261,
0.268529,
0.535588,
0.481923
]
],
"network.2.bias": [
0.196602,
0.206303,
0.02976,
-0.366135,
0.105231,
0.156541,
0.470728
],
"network.4.weight": [
[
0.116507,
0.150446,
-0.131707,
0.660427,
0.734444,
-0.528391,
0.089593
],
[
0.408236,
0.300497,
-0.230615,
0.581487,
0.443597,
-0.613986,
-0.448919
],
[
0.661346,
0.099115,
-0.095933,
0.429602,
0.347436,
-0.201066,
-0.0215
],
[
-0.24541,
-0.216023,
0.048785,
-0.237383,
0.363465,
-0.349075,
0.071439
],
[
0.031826,
-0.76777,
0.021837,
-0.317061,
-0.495543,
0.926388,
0.595837
],
[
0.36394,
0.660079,
0.212181,
0.782706,
0.304498,
-0.142497,
0.163584
],
[
-0.491878,
0.830948,
0.244746,
-0.329895,
-0.286305,
0.056675,
0.026162
]
],
"network.4.bias": [
0.20161,
0.331574,
-0.177783,
-0.099706,
0.306915,
-0.354251,
0.520792
],
"network.6.weight": [
[
-0.117047,
-0.408328,
-0.553073,
-0.243121,
0.651411,
-0.574401,
0.316326
],
[
-0.055699,
0.530535,
-0.495304,
0.125256,
-0.537689,
0.498182,
0.71463
],
[
0.390314,
0.414194,
0.134641,
0.16047,
-0.34506,
0.107771,
0.672886
],
[
0.026899,
0.205252,
0.463854,
-0.257887,
-0.058063,
0.611076,
0.295651
],
[
0.656264,
0.635016,
0.519434,
-0.082352,
-0.385793,
0.25861,
1.078685
],
[
0.42913,
0.483038,
0.526444,
0.333304,
-0.274462,
0.639149,
0.665797
],
[
0.423744,
0.966114,
0.567512,
0.362332,
-0.244936,
0.80399,
0.623242
]
],
"network.6.bias": [
0.268786,
0.480734,
0.232236,
-0.176147,
0.046481,
0.202342,
0.030762
],
"network.8.weight": [
[
-0.403891,
0.669557,
0.231159,
0.756024,
0.258461,
0.181127,
0.72677
],
[
-0.222956,
1.047586,
0.258106,
0.11548,
0.413038,
-0.21113,
0.125947
],
[
0.645694,
0.187168,
-0.047457,
-0.292743,
-0.31384,
-0.391416,
-0.27282
],
[
-0.281864,
0.247729,
-0.028922,
0.110738,
-0.377511,
-0.320402,
-0.264559
],
[
-0.275149,
1.156714,
0.864955,
0.490656,
0.907881,
0.565388,
0.74899
],
[
-0.298865,
0.221157,
-0.028247,
-0.006022,
-0.466777,
-0.239773,
-0.133891
],
[
0.043154,
-0.115708,
0.486866,
0.361943,
0.211014,
0.576926,
0.457319
]
],
"network.8.bias": [
0.017971,
-0.219794,
0.487656,
-0.29469,
-0.061545,
-0.370553,
-0.117475
],
"network.10.weight": [
[
-0.298054,
-0.179603,
0.779064,
-0.366201,
-0.665426,
0.051792,
-0.663424
]
],
"network.10.bias": [
0.369057
]
}
## Activation Signature
### 0
mean: [-3.644296, 1.736744, 2.190536, -0.168118, 1.105531, 1.023298, -2.369859]
std: [2.347322, 1.656745, 1.699424, 1.465802, 1.462555, 1.580667, 3.485458]
### 2
mean: [2.287924, -0.252919, -1.514778, 0.923903, 0.600179, 1.021348, 0.738680]
std: [2.240422, 1.672768, 1.019839, 1.697994, 1.388828, 1.470474, 1.275480]
### 4
mean: [1.448152, 1.386488, 1.995128, -1.088648, 0.827829, 2.020873, -0.776411]
std: [2.201042, 2.738138, 2.394827, 0.651874, 2.487697, 2.026699, 2.336141]
### 6
mean: [-1.881813, 1.003201, 2.018548, 2.534450, 3.945251, 4.168937, 5.315058]
std: [4.303631, 1.999123, 2.521858, 2.744536, 4.585760, 4.481009, 5.902308]
### 8
mean: [8.868411, 3.414546, -4.041357, -4.347809, 14.766307, -4.091723, 7.607160]
std: [9.688735, 3.992241, 5.864625, 3.864516, 15.470802, 3.469247, 7.965319]
### 10
mean: [-17.515579]
std: [19.369736]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
alternating
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.225213,
-0.595051,
-0.593457,
0.099706,
-0.551871
],
[
-0.072744,
0.170267,
0.795061,
-0.233176,
0.077886
],
[
-0.093697,
0.432099,
0.748383,
-0.104163,
-0.15884
],
[
-0.165408,
-0.365347,
-0.110227,
0.048167,
0.671099
],
[
0.375296,
0.290414,
-0.529392,
0.128531,
0.422757
],
[
0.67084,
0.341363,
-0.216221,
-0.063743,
-0.145314
],
[
0.574614,
-1.200751,
0.031782,
-0.952072,
0.429675
]
],
"network.0.bias": [
-0.573225,
0.260186,
0.375742,
0.002737,
0.418112,
0.334011,
0.548444
],
"network.2.weight": [
[
0.019987,
0.728415,
0.542887,
-0.336374,
-0.263188,
0.223773,
-0.406385
],
[
0.23132,
-0.00197,
-0.397628,
0.514484,
0.843187,
-0.663505,
-0.361922
],
[
-0.176525,
-0.061979,
-0.490691,
-0.172159,
-0.062854,
-0.036306,
-0.370737
],
[
-0.210941,
0.568025,
0.414293,
-0.202623,
-0.271204,
-0.034746,
-0.3833
],
[
0.123682,
0.137596,
0.538902,
-0.073057,
-0.50579,
-0.186668,
-0.115042
],
[
0.056078,
-0.104216,
-0.126462,
-0.013606,
0.217423,
0.663964,
0.759508
],
[
0.003111,
-0.159389,
-0.268267,
-0.001261,
0.268529,
0.535588,
0.481923
]
],
"network.2.bias": [
0.196602,
0.206303,
0.02976,
-0.366135,
0.105231,
0.156541,
0.470728
],
"network.4.weight": [
[
0.116507,
0.150446,
-0.131707,
0.660427,
0.734444,
-0.528391,
0.089593
],
[
0.408236,
0.300497,
-0.230615,
0.581487,
0.443597,
-0.613986,
-0.448919
],
[
0.661346,
0.099115,
-0.095933,
0.429602,
0.347436,
-0.201066,
-0.0215
],
[
-0.24541,
-0.216023,
0.048785,
-0.237383,
0.363465,
-0.349075,
0.071439
],
[
0.031826,
-0.76777,
0.021837,
-0.317061,
-0.495543,
0.926388,
0.595837
],
[
0.36394,
0.660079,
0.212181,
0.782706,
0.304498,
-0.142497,
0.163584
],
[
-0.491878,
0.830948,
0.244746,
-0.329895,
-0.286305,
0.056675,
0.026162
]
],
"network.4.bias": [
0.20161,
0.331574,
-0.177783,
-0.099706,
0.306915,
-0.354251,
0.520792
],
"network.6.weight": [
[
-0.117047,
-0.408328,
-0.553073,
-0.243121,
0.651411,
-0.574401,
0.316326
],
[
-0.055699,
0.530535,
-0.495304,
0.125256,
-0.537689,
0.498182,
0.71463
],
[
0.390314,
0.414194,
0.134641,
0.16047,
-0.34506,
0.107771,
0.672886
],
[
0.026899,
0.205252,
0.463854,
-0.257887,
-0.058063,
0.611076,
0.295651
],
[
0.656264,
0.635016,
0.519434,
-0.082352,
-0.385793,
0.25861,
1.078685
],
[
0.42913,
0.483038,
0.526444,
0.333304,
-0.274462,
0.639149,
0.665797
],
[
0.423744,
0.966114,
0.567512,
0.362332,
-0.244936,
0.80399,
0.623242
]
],
"network.6.bias": [
0.268786,
0.480734,
0.232236,
-0.176147,
0.046481,
0.202342,
0.030762
],
"network.8.weight": [
[
-0.403891,
0.669557,
0.231159,
0.756024,
0.258461,
0.181127,
0.72677
],
[
-0.222956,
1.047586,
0.258106,
0.11548,
0.413038,
-0.21113,
0.125947
],
[
0.645694,
0.187168,
-0.047457,
-0.292743,
-0.31384,
-0.391416,
-0.27282
],
[
-0.281864,
0.247729,
-0.028922,
0.110738,
-0.377511,
-0.320402,
-0.264559
],
[
-0.275149,
1.156714,
0.864955,
0.490656,
0.907881,
0.565388,
0.74899
],
[
-0.298865,
0.221157,
-0.028247,
-0.006022,
-0.466777,
-0.239773,
-0.133891
],
[
0.043154,
-0.115708,
0.486866,
0.361943,
0.211014,
0.576926,
0.457319
]
],
"network.8.bias": [
0.017971,
-0.219794,
0.487656,
-0.29469,
-0.061545,
-0.370553,
-0.117475
],
"network.10.weight": [
[
-0.298054,
-0.179603,
0.779064,
-0.366201,
-0.665426,
0.051792,
-0.663424
]
],
"network.10.bias": [
0.369057
]
}
## Activation Signature
### 0
mean: [-3.644296, 1.736744, 2.190536, -0.168118, 1.105531, 1.023298, -2.369859]
std: [2.347322, 1.656745, 1.699424, 1.465802, 1.462555, 1.580667, 3.485458]
### 2
mean: [2.287924, -0.252919, -1.514778, 0.923903, 0.600179, 1.021348, 0.738680]
std: [2.240422, 1.672768, 1.019839, 1.697994, 1.388828, 1.470474, 1.275480]
### 4
mean: [1.448152, 1.386488, 1.995128, -1.088648, 0.827829, 2.020873, -0.776411]
std: [2.201042, 2.738138, 2.394827, 0.651874, 2.487697, 2.026699, 2.336141]
### 6
mean: [-1.881813, 1.003201, 2.018548, 2.534450, 3.945251, 4.168937, 5.315058]
std: [4.303631, 1.999123, 2.521858, 2.744536, 4.585760, 4.481009, 5.902308]
### 8
mean: [8.868411, 3.414546, -4.041357, -4.347809, 14.766307, -4.091723, 7.607160]
std: [9.688735, 3.992241, 5.864625, 3.864516, 15.470802, 3.469247, 7.965319]
### 10
mean: [-17.515579]
std: [19.369736]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
alternating
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -3.6442956924438477, "std": 2.3473217487335205}, "1": {"mean": 1.7367438077926636, "std": 1.6567453145980835}, "2": {"mean": 2.1905357837677, "std": 1.699423909187317}, "3": {"mean": -0.16811810433864594, "std": 1.4658018350601196}, "4": {"mean": 1.1055306196212769, "std": 1.4625550508499146}, "5": {"mean": 1.0232980251312256, "std": 1.5806670188903809}, "6": {"mean": -2.369859457015991, "std": 3.4854581356048584}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 2.287924289703369, "std": 2.240421772003174}, "1": {"mean": -0.25291895866394043, "std": 1.6727683544158936}, "2": {"mean": -1.514777660369873, "std": 1.0198394060134888}, "3": {"mean": 0.9239026308059692, "std": 1.697993516921997}, "4": {"mean": 0.6001790165901184, "std": 1.3888275623321533}, "5": {"mean": 1.021348237991333, "std": 1.4704740047454834}, "6": {"mean": 0.7386796474456787, "std": 1.2754802703857422}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 1.448151707649231, "std": 2.2010416984558105}, "1": {"mean": 1.3864879608154297, "std": 2.738138198852539}, "2": {"mean": 1.9951276779174805, "std": 2.394826889038086}, "3": {"mean": -1.0886476039886475, "std": 0.651874303817749}, "4": {"mean": 0.8278287053108215, "std": 2.487697124481201}, "5": {"mean": 2.0208728313446045, "std": 2.0266993045806885}, "6": {"mean": -0.7764107584953308, "std": 2.3361408710479736}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -1.881812572479248, "std": 4.303630828857422}, "1": {"mean": 1.003200650215149, "std": 1.9991233348846436}, "2": {"mean": 2.018548011779785, "std": 2.521857500076294}, "3": {"mean": 2.5344502925872803, "std": 2.7445363998413086}, "4": {"mean": 3.94525146484375, "std": 4.585760116577148}, "5": {"mean": 4.168936729431152, "std": 4.481009006500244}, "6": {"mean": 5.31505823135376, "std": 5.902307987213135}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 8.86841106414795, "std": 9.688735008239746}, "1": {"mean": 3.414546489715576, "std": 3.9922409057617188}, "2": {"mean": -4.041356563568115, "std": 5.864624500274658}, "3": {"mean": -4.347809314727783, "std": 3.864515781402588}, "4": {"mean": 14.76630687713623, "std": 15.470802307128906}, "5": {"mean": -4.091723442077637, "std": 3.4692471027374268}, "6": {"mean": 7.6071600914001465, "std": 7.96531867980957}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -17.515579223632812, "std": 19.369735717773438}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.225213, -0.595051, -0.593457, 0.099706, -0.551871], [-0.072744, 0.170267, 0.795061, -0.233176, 0.077886], [-0.093697, 0.432099, 0.748383, -0.104163, -0.15884], [-0.165408, -0.365347, -0.110227, 0.048167, 0.671099], [0.375296, 0.290414, -0.529392, 0.128531, 0.422757], [0.67084, 0.341363, -0.216221, -0.063743, -0.145314], [0.574614, -1.200751, 0.031782, -0.952072, 0.429675]], "network.0.bias": [-0.573225, 0.260186, 0.375742, 0.002737, 0.418112, 0.334011, 0.548444], "network.2.weight": [[0.019987, 0.728415, 0.542887, -0.336374, -0.263188, 0.223773, -0.406385], [0.23132, -0.00197, -0.397628, 0.514484, 0.843187, -0.663505, -0.361922], [-0.176525, -0.061979, -0.490691, -0.172159, -0.062854, -0.036306, -0.370737], [-0.210941, 0.568025, 0.414293, -0.202623, -0.271204, -0.034746, -0.3833], [0.123682, 0.137596, 0.538902, -0.073057, -0.50579, -0.186668, -0.115042], [0.056078, -0.104216, -0.126462, -0.013606, 0.217423, 0.663964, 0.759508], [0.003111, -0.159389, -0.268267, -0.001261, 0.268529, 0.535588, 0.481923]], "network.2.bias": [0.196602, 0.206303, 0.02976, -0.366135, 0.105231, 0.156541, 0.470728], "network.4.weight": [[0.116507, 0.150446, -0.131707, 0.660427, 0.734444, -0.528391, 0.089593], [0.408236, 0.300497, -0.230615, 0.581487, 0.443597, -0.613986, -0.448919], [0.661346, 0.099115, -0.095933, 0.429602, 0.347436, -0.201066, -0.0215], [-0.24541, -0.216023, 0.048785, -0.237383, 0.363465, -0.349075, 0.071439], [0.031826, -0.76777, 0.021837, -0.317061, -0.495543, 0.926388, 0.595837], [0.36394, 0.660079, 0.212181, 0.782706, 0.304498, -0.142497, 0.163584], [-0.491878, 0.830948, 0.244746, -0.329895, -0.286305, 0.056675, 0.026162]], "network.4.bias": [0.20161, 0.331574, -0.177783, -0.099706, 0.306915, -0.354251, 0.520792], "network.6.weight": [[-0.117047, -0.408328, -0.553073, -0.243121, 0.651411, -0.574401, 0.316326], [-0.055699, 0.530535, -0.495304, 0.125256, -0.537689, 0.498182, 0.71463], [0.390314, 0.414194, 0.134641, 0.16047, -0.34506, 0.107771, 0.672886], [0.026899, 0.205252, 0.463854, -0.257887, -0.058063, 0.611076, 0.295651], [0.656264, 0.635016, 0.519434, -0.082352, -0.385793, 0.25861, 1.078685], [0.42913, 0.483038, 0.526444, 0.333304, -0.274462, 0.639149, 0.665797], [0.423744, 0.966114, 0.567512, 0.362332, -0.244936, 0.80399, 0.623242]], "network.6.bias": [0.268786, 0.480734, 0.232236, -0.176147, 0.046481, 0.202342, 0.030762], "network.8.weight": [[-0.403891, 0.669557, 0.231159, 0.756024, 0.258461, 0.181127, 0.72677], [-0.222956, 1.047586, 0.258106, 0.11548, 0.413038, -0.21113, 0.125947], [0.645694, 0.187168, -0.047457, -0.292743, -0.31384, -0.391416, -0.27282], [-0.281864, 0.247729, -0.028922, 0.110738, -0.377511, -0.320402, -0.264559], [-0.275149, 1.156714, 0.864955, 0.490656, 0.907881, 0.565388, 0.74899], [-0.298865, 0.221157, -0.028247, -0.006022, -0.466777, -0.239773, -0.133891], [0.043154, -0.115708, 0.486866, 0.361943, 0.211014, 0.576926, 0.457319]], "network.8.bias": [0.017971, -0.219794, 0.487656, -0.29469, -0.061545, -0.370553, -0.117475], "network.10.weight": [[-0.298054, -0.179603, 0.779064, -0.366201, -0.665426, 0.051792, -0.663424]], "network.10.bias": [0.369057]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6851120293140411, "train_acc": 0.58, "val_loss": 0.7199480533599854, "val_acc": 0.46}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6677788496017456, "train_acc": 0.58, "val_loss": 0.7104914784431458, "val_acc": 0.46}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6613682806491852, "train_acc": 0.58, "val_loss": 0.6884070038795471, "val_acc": 0.46}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6214916408061981, "train_acc": 0.58, "val_loss": 0.6227372884750366, "val_acc": 0.46}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6071749627590179, "train_acc": 0.51, "val_loss": 0.6322181820869446, "val_acc": 0.46}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.5954710245132446, "train_acc": 0.5, "val_loss": 0.6216787099838257, "val_acc": 0.8}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.5599729418754578, "train_acc": 0.845, "val_loss": 0.5324399471282959, "val_acc": 0.84}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.4593876153230667, "train_acc": 0.865, "val_loss": 0.4214504361152649, "val_acc": 0.86}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.37028956413269043, "train_acc": 0.885, "val_loss": 0.3593336045742035, "val_acc": 0.88}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.2907203733921051, "train_acc": 0.935, "val_loss": 0.3699520528316498, "val_acc": 0.86}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.255069836974144, "train_acc": 0.92, "val_loss": 0.32310837507247925, "val_acc": 0.9}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.20396523922681808, "train_acc": 0.935, "val_loss": 0.234858900308609, "val_acc": 0.94}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.20622289925813675, "train_acc": 0.93, "val_loss": 0.22933539748191833, "val_acc": 0.94}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.18041935563087463, "train_acc": 0.95, "val_loss": 0.27368080615997314, "val_acc": 0.92}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["alternating"], "degraded_stage": {"initial_val_loss": 0.7199480533599854, "final_val_loss": 0.6227372884750366, "initial_val_acc": 0.46, "final_val_acc": 0.46, "best_val_acc": 0.46}, "improved_stage": {"initial_val_loss": 0.6322181820869446, "final_val_loss": 0.27368080615997314, "initial_val_acc": 0.46, "final_val_acc": 0.92, "best_val_acc": 0.94, "best_epoch": 11}, "improvement": 0.4799999999999999, "first_improvement_epoch": 3}}
|
50
|
{"target_pattern": "palindrome", "degraded_accuracy": 0.38, "improved_accuracy": 0.98, "improvement": 0.6, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 8776, "learning_rate": 0.0354662552402799, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "palindrome", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["palindrome"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.074786,
-0.244822,
-0.106781,
0.225961,
0.729673
],
[
0.677415,
-0.051097,
0.156275,
0.082297,
-0.293735
],
[
-0.509436,
0.183798,
0.137996,
-0.183924,
0.781449
],
[
-0.320329,
-0.188685,
-0.299768,
0.266319,
0.531122
],
[
-0.246715,
0.23085,
-0.226758,
-0.060214,
0.316667
]
],
"network.0.bias": [
0.095907,
0.544845,
0.359637,
0.130028,
0.086226
],
"network.2.weight": [
[
0.190436,
-0.165086,
0.183709,
-0.250322,
0.106033
],
[
0.057306,
-0.155677,
-0.538713,
-0.06803,
0.031078
],
[
0.227343,
0.37762,
-0.619811,
0.16308,
-0.116428
],
[
-0.224537,
0.287693,
-0.252638,
-0.572311,
0.088906
],
[
0.724969,
-0.4583,
0.4739,
0.376914,
0.611394
]
],
"network.2.bias": [
0.54719,
-0.575409,
-0.02434,
0.404866,
0.021511
],
"network.4.weight": [
[
-0.370225,
-0.000599,
0.202601,
-0.124957,
0.023684
],
[
0.46887,
-0.396787,
-0.415779,
0.058903,
-0.542934
],
[
0.203079,
0.026217,
-0.104557,
0.428162,
-0.543742
],
[
0.095998,
0.018819,
0.48474,
0.368285,
0.609979
],
[
-0.076716,
-0.227983,
0.710465,
0.357875,
-0.269758
]
],
"network.4.bias": [
-0.549966,
0.456019,
0.548866,
-0.266141,
-0.321076
],
"network.6.weight": [
[
0.255518,
0.317129,
0.604464,
-0.247494,
-0.41826
],
[
-0.233847,
0.542255,
0.32927,
-0.287211,
-0.415191
],
[
-0.226222,
-0.49071,
0.040866,
0.566026,
0.43343
],
[
0.100214,
0.512142,
0.435051,
-0.116188,
-0.553135
],
[
-0.295792,
0.422076,
0.219668,
-0.29801,
-0.456596
]
],
"network.6.bias": [
0.114689,
0.171414,
0.295022,
0.07859,
0.188411
],
"network.8.weight": [
[
0.154495,
0.212369,
-0.140192,
0.397604,
0.157064
],
[
-0.593383,
-0.43248,
0.704529,
-0.253091,
-0.240118
],
[
0.795626,
0.633882,
-0.475843,
0.309569,
0.304649
],
[
0.004511,
-0.104193,
-0.405023,
-0.084174,
-0.109326
],
[
0.011043,
-0.239013,
0.004405,
-0.616752,
-0.320244
]
],
"network.8.bias": [
-0.302891,
0.577486,
0.407622,
-0.167737,
0.354943
],
"network.10.weight": [
[
-0.419557,
0.812011,
-0.285351,
0.144397,
0.363448
],
[
0.424788,
0.340644,
0.042656,
0.155035,
0.43788
],
[
0.042986,
-0.347058,
0.654252,
-0.035033,
-0.135653
],
[
-0.185048,
-0.178406,
0.169242,
0.288146,
0.096192
],
[
0.359315,
-0.327746,
-0.113562,
0.324712,
0.025822
]
],
"network.10.bias": [
0.402908,
-0.200347,
0.467208,
-0.303681,
-0.122048
],
"network.12.weight": [
[
-0.726744,
-0.128598,
0.618017,
-0.198846,
-0.062244
]
],
"network.12.bias": [
-0.047766
]
}
## Activation Signature
### 0
mean: [0.718157, 1.435392, 0.918222, -0.012455, -0.012380]
std: [1.507822, 1.441804, 1.564401, 1.490112, 0.786928]
### 2
mean: [0.586236, -1.389657, 0.096482, 0.062713, 0.929871]
std: [0.451241, 0.618032, 0.898670, 1.248674, 2.340188]
### 4
mean: [-0.724432, -0.058480, 0.161051, 0.894364, -0.261893]
std: [0.166440, 0.956007, 1.156030, 1.192477, 0.960501]
### 6
mean: [0.204360, 0.145233, 0.784763, 0.221245, 0.051341]
std: [0.532672, 0.575195, 0.815136, 0.442060, 0.546430]
### 8
mean: [-0.123684, 0.645226, 0.689519, -0.574188, 0.004648]
std: [0.392255, 1.008550, 0.999221, 0.263815, 0.368466]
### 10
mean: [0.838728, 0.239436, 0.709197, -0.314859, -0.434983]
std: [0.998312, 0.273917, 0.834992, 0.236337, 0.226499]
### 12
mean: [-0.278443]
std: [1.126889]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
palindrome
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.074786,
-0.244822,
-0.106781,
0.225961,
0.729673
],
[
0.677415,
-0.051097,
0.156275,
0.082297,
-0.293735
],
[
-0.509436,
0.183798,
0.137996,
-0.183924,
0.781449
],
[
-0.320329,
-0.188685,
-0.299768,
0.266319,
0.531122
],
[
-0.246715,
0.23085,
-0.226758,
-0.060214,
0.316667
]
],
"network.0.bias": [
0.095907,
0.544845,
0.359637,
0.130028,
0.086226
],
"network.2.weight": [
[
0.190436,
-0.165086,
0.183709,
-0.250322,
0.106033
],
[
0.057306,
-0.155677,
-0.538713,
-0.06803,
0.031078
],
[
0.227343,
0.37762,
-0.619811,
0.16308,
-0.116428
],
[
-0.224537,
0.287693,
-0.252638,
-0.572311,
0.088906
],
[
0.724969,
-0.4583,
0.4739,
0.376914,
0.611394
]
],
"network.2.bias": [
0.54719,
-0.575409,
-0.02434,
0.404866,
0.021511
],
"network.4.weight": [
[
-0.370225,
-0.000599,
0.202601,
-0.124957,
0.023684
],
[
0.46887,
-0.396787,
-0.415779,
0.058903,
-0.542934
],
[
0.203079,
0.026217,
-0.104557,
0.428162,
-0.543742
],
[
0.095998,
0.018819,
0.48474,
0.368285,
0.609979
],
[
-0.076716,
-0.227983,
0.710465,
0.357875,
-0.269758
]
],
"network.4.bias": [
-0.549966,
0.456019,
0.548866,
-0.266141,
-0.321076
],
"network.6.weight": [
[
0.255518,
0.317129,
0.604464,
-0.247494,
-0.41826
],
[
-0.233847,
0.542255,
0.32927,
-0.287211,
-0.415191
],
[
-0.226222,
-0.49071,
0.040866,
0.566026,
0.43343
],
[
0.100214,
0.512142,
0.435051,
-0.116188,
-0.553135
],
[
-0.295792,
0.422076,
0.219668,
-0.29801,
-0.456596
]
],
"network.6.bias": [
0.114689,
0.171414,
0.295022,
0.07859,
0.188411
],
"network.8.weight": [
[
0.154495,
0.212369,
-0.140192,
0.397604,
0.157064
],
[
-0.593383,
-0.43248,
0.704529,
-0.253091,
-0.240118
],
[
0.795626,
0.633882,
-0.475843,
0.309569,
0.304649
],
[
0.004511,
-0.104193,
-0.405023,
-0.084174,
-0.109326
],
[
0.011043,
-0.239013,
0.004405,
-0.616752,
-0.320244
]
],
"network.8.bias": [
-0.302891,
0.577486,
0.407622,
-0.167737,
0.354943
],
"network.10.weight": [
[
-0.419557,
0.812011,
-0.285351,
0.144397,
0.363448
],
[
0.424788,
0.340644,
0.042656,
0.155035,
0.43788
],
[
0.042986,
-0.347058,
0.654252,
-0.035033,
-0.135653
],
[
-0.185048,
-0.178406,
0.169242,
0.288146,
0.096192
],
[
0.359315,
-0.327746,
-0.113562,
0.324712,
0.025822
]
],
"network.10.bias": [
0.402908,
-0.200347,
0.467208,
-0.303681,
-0.122048
],
"network.12.weight": [
[
-0.726744,
-0.128598,
0.618017,
-0.198846,
-0.062244
]
],
"network.12.bias": [
-0.047766
]
}
## Activation Signature
### 0
mean: [0.718157, 1.435392, 0.918222, -0.012455, -0.012380]
std: [1.507822, 1.441804, 1.564401, 1.490112, 0.786928]
### 2
mean: [0.586236, -1.389657, 0.096482, 0.062713, 0.929871]
std: [0.451241, 0.618032, 0.898670, 1.248674, 2.340188]
### 4
mean: [-0.724432, -0.058480, 0.161051, 0.894364, -0.261893]
std: [0.166440, 0.956007, 1.156030, 1.192477, 0.960501]
### 6
mean: [0.204360, 0.145233, 0.784763, 0.221245, 0.051341]
std: [0.532672, 0.575195, 0.815136, 0.442060, 0.546430]
### 8
mean: [-0.123684, 0.645226, 0.689519, -0.574188, 0.004648]
std: [0.392255, 1.008550, 0.999221, 0.263815, 0.368466]
### 10
mean: [0.838728, 0.239436, 0.709197, -0.314859, -0.434983]
std: [0.998312, 0.273917, 0.834992, 0.236337, 0.226499]
### 12
mean: [-0.278443]
std: [1.126889]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
palindrome
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.7181571125984192, "std": 1.5078222751617432}, "1": {"mean": 1.4353917837142944, "std": 1.44180428981781}, "2": {"mean": 0.9182223081588745, "std": 1.5644007921218872}, "3": {"mean": -0.012455434538424015, "std": 1.4901118278503418}, "4": {"mean": -0.012380009517073631, "std": 0.7869278192520142}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.5862355828285217, "std": 0.45124146342277527}, "1": {"mean": -1.3896573781967163, "std": 0.6180319786071777}, "2": {"mean": 0.0964815765619278, "std": 0.898669958114624}, "3": {"mean": 0.06271331757307053, "std": 1.2486735582351685}, "4": {"mean": 0.929871141910553, "std": 2.3401877880096436}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.7244316339492798, "std": 0.16643951833248138}, "1": {"mean": -0.058480072766542435, "std": 0.9560067057609558}, "2": {"mean": 0.16105066239833832, "std": 1.1560300588607788}, "3": {"mean": 0.8943639397621155, "std": 1.192476511001587}, "4": {"mean": -0.26189306378364563, "std": 0.9605007171630859}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.20435985922813416, "std": 0.5326719284057617}, "1": {"mean": 0.1452333778142929, "std": 0.5751945972442627}, "2": {"mean": 0.7847633361816406, "std": 0.8151357173919678}, "3": {"mean": 0.22124534845352173, "std": 0.4420604109764099}, "4": {"mean": 0.051340892910957336, "std": 0.5464303493499756}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.12368357181549072, "std": 0.3922547399997711}, "1": {"mean": 0.6452263593673706, "std": 1.0085495710372925}, "2": {"mean": 0.6895190477371216, "std": 0.9992206692695618}, "3": {"mean": -0.5741884708404541, "std": 0.2638150453567505}, "4": {"mean": 0.004647995810955763, "std": 0.3684658408164978}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": 0.8387278318405151, "std": 0.9983118772506714}, "1": {"mean": 0.23943570256233215, "std": 0.2739166021347046}, "2": {"mean": 0.7091968655586243, "std": 0.8349918723106384}, "3": {"mean": -0.3148590624332428, "std": 0.23633742332458496}, "4": {"mean": -0.43498319387435913, "std": 0.2264990508556366}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -0.2784431278705597, "std": 1.1268888711929321}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.074786, -0.244822, -0.106781, 0.225961, 0.729673], [0.677415, -0.051097, 0.156275, 0.082297, -0.293735], [-0.509436, 0.183798, 0.137996, -0.183924, 0.781449], [-0.320329, -0.188685, -0.299768, 0.266319, 0.531122], [-0.246715, 0.23085, -0.226758, -0.060214, 0.316667]], "network.0.bias": [0.095907, 0.544845, 0.359637, 0.130028, 0.086226], "network.2.weight": [[0.190436, -0.165086, 0.183709, -0.250322, 0.106033], [0.057306, -0.155677, -0.538713, -0.06803, 0.031078], [0.227343, 0.37762, -0.619811, 0.16308, -0.116428], [-0.224537, 0.287693, -0.252638, -0.572311, 0.088906], [0.724969, -0.4583, 0.4739, 0.376914, 0.611394]], "network.2.bias": [0.54719, -0.575409, -0.02434, 0.404866, 0.021511], "network.4.weight": [[-0.370225, -0.000599, 0.202601, -0.124957, 0.023684], [0.46887, -0.396787, -0.415779, 0.058903, -0.542934], [0.203079, 0.026217, -0.104557, 0.428162, -0.543742], [0.095998, 0.018819, 0.48474, 0.368285, 0.609979], [-0.076716, -0.227983, 0.710465, 0.357875, -0.269758]], "network.4.bias": [-0.549966, 0.456019, 0.548866, -0.266141, -0.321076], "network.6.weight": [[0.255518, 0.317129, 0.604464, -0.247494, -0.41826], [-0.233847, 0.542255, 0.32927, -0.287211, -0.415191], [-0.226222, -0.49071, 0.040866, 0.566026, 0.43343], [0.100214, 0.512142, 0.435051, -0.116188, -0.553135], [-0.295792, 0.422076, 0.219668, -0.29801, -0.456596]], "network.6.bias": [0.114689, 0.171414, 0.295022, 0.07859, 0.188411], "network.8.weight": [[0.154495, 0.212369, -0.140192, 0.397604, 0.157064], [-0.593383, -0.43248, 0.704529, -0.253091, -0.240118], [0.795626, 0.633882, -0.475843, 0.309569, 0.304649], [0.004511, -0.104193, -0.405023, -0.084174, -0.109326], [0.011043, -0.239013, 0.004405, -0.616752, -0.320244]], "network.8.bias": [-0.302891, 0.577486, 0.407622, -0.167737, 0.354943], "network.10.weight": [[-0.419557, 0.812011, -0.285351, 0.144397, 0.363448], [0.424788, 0.340644, 0.042656, 0.155035, 0.43788], [0.042986, -0.347058, 0.654252, -0.035033, -0.135653], [-0.185048, -0.178406, 0.169242, 0.288146, 0.096192], [0.359315, -0.327746, -0.113562, 0.324712, 0.025822]], "network.10.bias": [0.402908, -0.200347, 0.467208, -0.303681, -0.122048], "network.12.weight": [[-0.726744, -0.128598, 0.618017, -0.198846, -0.062244]], "network.12.bias": [-0.047766]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6802124381065369, "train_acc": 0.605, "val_loss": 0.7295581698417664, "val_acc": 0.38}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.666690319776535, "train_acc": 0.605, "val_loss": 0.7070908546447754, "val_acc": 0.38}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6462147533893585, "train_acc": 0.605, "val_loss": 0.66290283203125, "val_acc": 0.38}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.6405598223209381, "train_acc": 0.53, "val_loss": 0.6825972199440002, "val_acc": 0.38}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.5814093947410583, "train_acc": 0.685, "val_loss": 0.5374184250831604, "val_acc": 0.88}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.5127487331628799, "train_acc": 0.87, "val_loss": 0.4843367636203766, "val_acc": 0.86}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.4325132369995117, "train_acc": 0.855, "val_loss": 0.3181614577770233, "val_acc": 0.98}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.3119715005159378, "train_acc": 0.905, "val_loss": 0.47881823778152466, "val_acc": 0.76}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.26153166592121124, "train_acc": 0.895, "val_loss": 0.23678909242153168, "val_acc": 0.92}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.2706978917121887, "train_acc": 0.905, "val_loss": 0.33986592292785645, "val_acc": 0.9}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.25956422835588455, "train_acc": 0.9, "val_loss": 0.34799420833587646, "val_acc": 0.88}, {"stage": "improved", "epoch": 8, "global_epoch": 11, "train_loss": 0.19367840141057968, "train_acc": 0.92, "val_loss": 0.36796846985816956, "val_acc": 0.86}], "summary": {"total_epochs": 12, "degraded_epochs": 3, "improved_epochs": 9, "patterns": ["palindrome"], "degraded_stage": {"initial_val_loss": 0.7295581698417664, "final_val_loss": 0.66290283203125, "initial_val_acc": 0.38, "final_val_acc": 0.38, "best_val_acc": 0.38}, "improved_stage": {"initial_val_loss": 0.6825972199440002, "final_val_loss": 0.36796846985816956, "initial_val_acc": 0.38, "final_val_acc": 0.86, "best_val_acc": 0.98, "best_epoch": 6}, "improvement": 0.6, "first_improvement_epoch": 2}}
|
51
|
{"target_pattern": "mountain_pattern", "degraded_accuracy": 0.66, "improved_accuracy": 0.84, "improvement": 0.17999999999999994, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 1955, "learning_rate": 0.09855088108934446, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "mountain_pattern", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["mountain_pattern"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.505764,
-0.750362,
0.341107,
0.516758,
-0.484932
],
[
-0.299606,
0.150042,
-0.855076,
-0.238442,
0.115035
],
[
1.213086,
0.939246,
-0.351775,
-0.448921,
-0.305036
],
[
-2.133677,
-0.484616,
0.192832,
0.265941,
0.050579
],
[
-0.303943,
-0.160779,
-0.472903,
-0.066146,
-0.982549
],
[
0.065523,
-0.458249,
-0.620241,
-0.325351,
-0.166758
]
],
"network.0.bias": [
-0.356872,
-0.182546,
0.48386,
0.120823,
-0.538946,
-0.666656
],
"network.2.weight": [
[
-0.096089,
0.189391,
-0.183268,
-0.266589,
0.061279,
1.117049
],
[
-0.323362,
-2.117579,
1.035096,
-0.166854,
-1.377107,
-1.755424
],
[
-0.617458,
-2.280591,
0.749607,
0.103544,
-0.775297,
-1.322881
],
[
0.650707,
1.356761,
-0.513325,
0.256697,
0.433939,
0.821065
],
[
-0.656506,
-1.681828,
0.678636,
-0.176137,
-1.358211,
-1.289514
],
[
0.465051,
-2.225478,
0.941041,
-0.997728,
-1.66695,
-0.87236
]
],
"network.2.bias": [
-0.619703,
-0.065607,
0.159141,
0.156489,
-0.08031,
0.056739
],
"network.4.weight": [
[
-0.428281,
-0.382536,
-0.089742,
-0.773726,
-0.190547,
-0.110099
],
[
0.200424,
-0.01202,
0.060448,
0.424082,
0.655519,
-0.803489
],
[
-1.302328,
0.884683,
1.095603,
-0.083776,
0.969155,
0.675231
],
[
0.23222,
-0.247028,
-0.764664,
-0.599856,
0.129885,
-0.618131
],
[
0.298615,
0.036701,
0.370123,
-0.076982,
0.405846,
-0.794504
],
[
-0.245021,
-0.300633,
-0.215305,
0.489535,
0.116958,
-1.767424
]
],
"network.4.bias": [
-0.62357,
0.05841,
0.217477,
-0.673801,
-0.035524,
0.470202
],
"network.6.weight": [
[
0.565611,
0.107785,
-0.834364,
0.207837,
0.616477,
0.923266
],
[
0.582375,
0.34877,
-0.401407,
-0.155737,
0.028743,
0.454837
],
[
-0.665243,
-0.972321,
0.810309,
-1.171498,
0.38115,
0.079041
],
[
0.241212,
-0.5514,
1.136428,
0.055349,
0.464124,
-0.144367
],
[
0.176069,
0.309,
-1.15441,
-0.143974,
0.610616,
0.742362
],
[
-0.635481,
-1.244131,
0.77824,
-0.45388,
-0.039128,
-0.434947
]
],
"network.6.bias": [
0.310775,
-0.056411,
0.325488,
-0.365912,
0.559837,
0.117533
],
"network.8.weight": [
[
0.647499,
0.257072,
-1.102874,
-0.367844,
0.772253,
-0.177306
],
[
0.077868,
-0.254476,
0.940077,
0.822472,
-0.06593,
0.510004
],
[
-0.604658,
-0.161286,
0.878217,
0.931598,
-0.433394,
0.623618
],
[
-0.142399,
-0.623161,
0.383243,
1.062355,
0.183346,
0.60676
],
[
-0.696352,
-0.345925,
-0.489457,
-0.187558,
-1.122684,
-0.459244
],
[
0.471453,
0.067717,
-1.163074,
-0.613638,
0.538598,
-0.429204
]
],
"network.8.bias": [
0.024412,
0.362389,
-0.108576,
0.620548,
-1.027157,
0.128446
],
"network.10.weight": [
[
0.473648,
-0.66931,
-0.543632,
-0.602893,
-0.452571,
0.281963
]
],
"network.10.bias": [
-0.095737
]
}
## Activation Signature
### 0
mean: [-1.130984, -2.437588, 1.631278, -2.382018, -3.551658, -3.619127]
std: [2.099892, 1.867301, 3.311677, 4.603977, 2.535317, 1.904657]
### 2
mean: [-1.115693, 2.055714, 1.664544, -0.720615, 1.209212, 2.008956]
std: [0.494203, 3.125301, 2.338343, 1.765005, 2.203844, 2.835941]
### 4
mean: [-2.197675, -0.527549, 6.871439, -3.765188, -0.438573, -3.843443]
std: [1.964818, 0.928793, 8.902348, 3.803063, 0.475584, 6.172052]
### 6
mean: [-5.322303, -2.750631, 5.997141, 7.312260, -7.239088, 5.435455]
std: [7.600005, 3.668836, 7.170979, 10.198218, 10.466903, 7.054097]
### 8
mean: [-9.898289, 14.878080, 15.233572, 14.147126, -8.338842, -13.476859]
std: [13.204537, 18.641594, 20.289927, 17.727560, 8.303020, 17.771898]
### 10
mean: [-26.741732]
std: [34.289993]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
mountain_pattern
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.505764,
-0.750362,
0.341107,
0.516758,
-0.484932
],
[
-0.299606,
0.150042,
-0.855076,
-0.238442,
0.115035
],
[
1.213086,
0.939246,
-0.351775,
-0.448921,
-0.305036
],
[
-2.133677,
-0.484616,
0.192832,
0.265941,
0.050579
],
[
-0.303943,
-0.160779,
-0.472903,
-0.066146,
-0.982549
],
[
0.065523,
-0.458249,
-0.620241,
-0.325351,
-0.166758
]
],
"network.0.bias": [
-0.356872,
-0.182546,
0.48386,
0.120823,
-0.538946,
-0.666656
],
"network.2.weight": [
[
-0.096089,
0.189391,
-0.183268,
-0.266589,
0.061279,
1.117049
],
[
-0.323362,
-2.117579,
1.035096,
-0.166854,
-1.377107,
-1.755424
],
[
-0.617458,
-2.280591,
0.749607,
0.103544,
-0.775297,
-1.322881
],
[
0.650707,
1.356761,
-0.513325,
0.256697,
0.433939,
0.821065
],
[
-0.656506,
-1.681828,
0.678636,
-0.176137,
-1.358211,
-1.289514
],
[
0.465051,
-2.225478,
0.941041,
-0.997728,
-1.66695,
-0.87236
]
],
"network.2.bias": [
-0.619703,
-0.065607,
0.159141,
0.156489,
-0.08031,
0.056739
],
"network.4.weight": [
[
-0.428281,
-0.382536,
-0.089742,
-0.773726,
-0.190547,
-0.110099
],
[
0.200424,
-0.01202,
0.060448,
0.424082,
0.655519,
-0.803489
],
[
-1.302328,
0.884683,
1.095603,
-0.083776,
0.969155,
0.675231
],
[
0.23222,
-0.247028,
-0.764664,
-0.599856,
0.129885,
-0.618131
],
[
0.298615,
0.036701,
0.370123,
-0.076982,
0.405846,
-0.794504
],
[
-0.245021,
-0.300633,
-0.215305,
0.489535,
0.116958,
-1.767424
]
],
"network.4.bias": [
-0.62357,
0.05841,
0.217477,
-0.673801,
-0.035524,
0.470202
],
"network.6.weight": [
[
0.565611,
0.107785,
-0.834364,
0.207837,
0.616477,
0.923266
],
[
0.582375,
0.34877,
-0.401407,
-0.155737,
0.028743,
0.454837
],
[
-0.665243,
-0.972321,
0.810309,
-1.171498,
0.38115,
0.079041
],
[
0.241212,
-0.5514,
1.136428,
0.055349,
0.464124,
-0.144367
],
[
0.176069,
0.309,
-1.15441,
-0.143974,
0.610616,
0.742362
],
[
-0.635481,
-1.244131,
0.77824,
-0.45388,
-0.039128,
-0.434947
]
],
"network.6.bias": [
0.310775,
-0.056411,
0.325488,
-0.365912,
0.559837,
0.117533
],
"network.8.weight": [
[
0.647499,
0.257072,
-1.102874,
-0.367844,
0.772253,
-0.177306
],
[
0.077868,
-0.254476,
0.940077,
0.822472,
-0.06593,
0.510004
],
[
-0.604658,
-0.161286,
0.878217,
0.931598,
-0.433394,
0.623618
],
[
-0.142399,
-0.623161,
0.383243,
1.062355,
0.183346,
0.60676
],
[
-0.696352,
-0.345925,
-0.489457,
-0.187558,
-1.122684,
-0.459244
],
[
0.471453,
0.067717,
-1.163074,
-0.613638,
0.538598,
-0.429204
]
],
"network.8.bias": [
0.024412,
0.362389,
-0.108576,
0.620548,
-1.027157,
0.128446
],
"network.10.weight": [
[
0.473648,
-0.66931,
-0.543632,
-0.602893,
-0.452571,
0.281963
]
],
"network.10.bias": [
-0.095737
]
}
## Activation Signature
### 0
mean: [-1.130984, -2.437588, 1.631278, -2.382018, -3.551658, -3.619127]
std: [2.099892, 1.867301, 3.311677, 4.603977, 2.535317, 1.904657]
### 2
mean: [-1.115693, 2.055714, 1.664544, -0.720615, 1.209212, 2.008956]
std: [0.494203, 3.125301, 2.338343, 1.765005, 2.203844, 2.835941]
### 4
mean: [-2.197675, -0.527549, 6.871439, -3.765188, -0.438573, -3.843443]
std: [1.964818, 0.928793, 8.902348, 3.803063, 0.475584, 6.172052]
### 6
mean: [-5.322303, -2.750631, 5.997141, 7.312260, -7.239088, 5.435455]
std: [7.600005, 3.668836, 7.170979, 10.198218, 10.466903, 7.054097]
### 8
mean: [-9.898289, 14.878080, 15.233572, 14.147126, -8.338842, -13.476859]
std: [13.204537, 18.641594, 20.289927, 17.727560, 8.303020, 17.771898]
### 10
mean: [-26.741732]
std: [34.289993]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
mountain_pattern
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.130983829498291, "std": 2.0998923778533936}, "1": {"mean": -2.4375882148742676, "std": 1.8673007488250732}, "2": {"mean": 1.631278157234192, "std": 3.3116767406463623}, "3": {"mean": -2.3820176124572754, "std": 4.603977203369141}, "4": {"mean": -3.5516581535339355, "std": 2.5353174209594727}, "5": {"mean": -3.619127035140991, "std": 1.9046568870544434}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -1.11569344997406, "std": 0.4942033886909485}, "1": {"mean": 2.055713653564453, "std": 3.1253011226654053}, "2": {"mean": 1.664543628692627, "std": 2.3383426666259766}, "3": {"mean": -0.7206147313117981, "std": 1.7650049924850464}, "4": {"mean": 1.2092121839523315, "std": 2.2038443088531494}, "5": {"mean": 2.008955955505371, "std": 2.8359413146972656}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -2.197674512863159, "std": 1.9648178815841675}, "1": {"mean": -0.5275487899780273, "std": 0.9287927746772766}, "2": {"mean": 6.871439456939697, "std": 8.902347564697266}, "3": {"mean": -3.7651875019073486, "std": 3.803062677383423}, "4": {"mean": -0.4385731518268585, "std": 0.47558358311653137}, "5": {"mean": -3.843442678451538, "std": 6.172051906585693}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -5.32230281829834, "std": 7.60000467300415}, "1": {"mean": -2.7506308555603027, "std": 3.6688356399536133}, "2": {"mean": 5.997141361236572, "std": 7.170979022979736}, "3": {"mean": 7.312260150909424, "std": 10.19821834564209}, "4": {"mean": -7.2390875816345215, "std": 10.466902732849121}, "5": {"mean": 5.435455322265625, "std": 7.0540971755981445}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -9.89828872680664, "std": 13.204537391662598}, "1": {"mean": 14.878080368041992, "std": 18.64159393310547}, "2": {"mean": 15.233572006225586, "std": 20.289926528930664}, "3": {"mean": 14.147126197814941, "std": 17.72756004333496}, "4": {"mean": -8.338842391967773, "std": 8.303020477294922}, "5": {"mean": -13.476859092712402, "std": 17.77189826965332}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -26.741731643676758, "std": 34.28999328613281}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.505764, -0.750362, 0.341107, 0.516758, -0.484932], [-0.299606, 0.150042, -0.855076, -0.238442, 0.115035], [1.213086, 0.939246, -0.351775, -0.448921, -0.305036], [-2.133677, -0.484616, 0.192832, 0.265941, 0.050579], [-0.303943, -0.160779, -0.472903, -0.066146, -0.982549], [0.065523, -0.458249, -0.620241, -0.325351, -0.166758]], "network.0.bias": [-0.356872, -0.182546, 0.48386, 0.120823, -0.538946, -0.666656], "network.2.weight": [[-0.096089, 0.189391, -0.183268, -0.266589, 0.061279, 1.117049], [-0.323362, -2.117579, 1.035096, -0.166854, -1.377107, -1.755424], [-0.617458, -2.280591, 0.749607, 0.103544, -0.775297, -1.322881], [0.650707, 1.356761, -0.513325, 0.256697, 0.433939, 0.821065], [-0.656506, -1.681828, 0.678636, -0.176137, -1.358211, -1.289514], [0.465051, -2.225478, 0.941041, -0.997728, -1.66695, -0.87236]], "network.2.bias": [-0.619703, -0.065607, 0.159141, 0.156489, -0.08031, 0.056739], "network.4.weight": [[-0.428281, -0.382536, -0.089742, -0.773726, -0.190547, -0.110099], [0.200424, -0.01202, 0.060448, 0.424082, 0.655519, -0.803489], [-1.302328, 0.884683, 1.095603, -0.083776, 0.969155, 0.675231], [0.23222, -0.247028, -0.764664, -0.599856, 0.129885, -0.618131], [0.298615, 0.036701, 0.370123, -0.076982, 0.405846, -0.794504], [-0.245021, -0.300633, -0.215305, 0.489535, 0.116958, -1.767424]], "network.4.bias": [-0.62357, 0.05841, 0.217477, -0.673801, -0.035524, 0.470202], "network.6.weight": [[0.565611, 0.107785, -0.834364, 0.207837, 0.616477, 0.923266], [0.582375, 0.34877, -0.401407, -0.155737, 0.028743, 0.454837], [-0.665243, -0.972321, 0.810309, -1.171498, 0.38115, 0.079041], [0.241212, -0.5514, 1.136428, 0.055349, 0.464124, -0.144367], [0.176069, 0.309, -1.15441, -0.143974, 0.610616, 0.742362], [-0.635481, -1.244131, 0.77824, -0.45388, -0.039128, -0.434947]], "network.6.bias": [0.310775, -0.056411, 0.325488, -0.365912, 0.559837, 0.117533], "network.8.weight": [[0.647499, 0.257072, -1.102874, -0.367844, 0.772253, -0.177306], [0.077868, -0.254476, 0.940077, 0.822472, -0.06593, 0.510004], [-0.604658, -0.161286, 0.878217, 0.931598, -0.433394, 0.623618], [-0.142399, -0.623161, 0.383243, 1.062355, 0.183346, 0.60676], [-0.696352, -0.345925, -0.489457, -0.187558, -1.122684, -0.459244], [0.471453, 0.067717, -1.163074, -0.613638, 0.538598, -0.429204]], "network.8.bias": [0.024412, 0.362389, -0.108576, 0.620548, -1.027157, 0.128446], "network.10.weight": [[0.473648, -0.66931, -0.543632, -0.602893, -0.452571, 0.281963]], "network.10.bias": [-0.095737]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6877061426639557, "train_acc": 0.565, "val_loss": 0.6857731342315674, "val_acc": 0.56}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6883341073989868, "train_acc": 0.565, "val_loss": 0.6857035160064697, "val_acc": 0.56}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6859287023544312, "train_acc": 0.59, "val_loss": 0.6767683625221252, "val_acc": 0.56}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6632654666900635, "train_acc": 0.59, "val_loss": 0.6645850539207458, "val_acc": 0.72}, {"stage": "degraded", "epoch": 4, "global_epoch": 4, "train_loss": 0.6052843034267426, "train_acc": 0.69, "val_loss": 0.5482322573661804, "val_acc": 0.66}, {"stage": "improved", "epoch": 0, "global_epoch": 5, "train_loss": 0.6815657615661621, "train_acc": 0.655, "val_loss": 0.6492695808410645, "val_acc": 0.7}, {"stage": "improved", "epoch": 1, "global_epoch": 6, "train_loss": 0.5913737416267395, "train_acc": 0.745, "val_loss": 0.606422483921051, "val_acc": 0.54}, {"stage": "improved", "epoch": 2, "global_epoch": 7, "train_loss": 0.611136794090271, "train_acc": 0.495, "val_loss": 0.5209540724754333, "val_acc": 0.76}, {"stage": "improved", "epoch": 3, "global_epoch": 8, "train_loss": 0.49702753126621246, "train_acc": 0.775, "val_loss": 0.48000648617744446, "val_acc": 0.72}, {"stage": "improved", "epoch": 4, "global_epoch": 9, "train_loss": 0.44883711636066437, "train_acc": 0.79, "val_loss": 0.5353825092315674, "val_acc": 0.72}, {"stage": "improved", "epoch": 5, "global_epoch": 10, "train_loss": 0.413448691368103, "train_acc": 0.795, "val_loss": 0.5883917808532715, "val_acc": 0.7}, {"stage": "improved", "epoch": 6, "global_epoch": 11, "train_loss": 0.3963129222393036, "train_acc": 0.8, "val_loss": 0.47099676728248596, "val_acc": 0.74}, {"stage": "improved", "epoch": 7, "global_epoch": 12, "train_loss": 0.40389537811279297, "train_acc": 0.79, "val_loss": 0.5884816646575928, "val_acc": 0.72}, {"stage": "improved", "epoch": 8, "global_epoch": 13, "train_loss": 0.3629572093486786, "train_acc": 0.83, "val_loss": 0.528208315372467, "val_acc": 0.82}, {"stage": "improved", "epoch": 9, "global_epoch": 14, "train_loss": 0.2940787896513939, "train_acc": 0.885, "val_loss": 0.3866063058376312, "val_acc": 0.84}], "summary": {"total_epochs": 15, "degraded_epochs": 5, "improved_epochs": 10, "patterns": ["mountain_pattern"], "degraded_stage": {"initial_val_loss": 0.6857731342315674, "final_val_loss": 0.5482322573661804, "initial_val_acc": 0.56, "final_val_acc": 0.66, "best_val_acc": 0.66}, "improved_stage": {"initial_val_loss": 0.6492695808410645, "final_val_loss": 0.3866063058376312, "initial_val_acc": 0.7, "final_val_acc": 0.84, "best_val_acc": 0.84, "best_epoch": 14}, "improvement": 0.17999999999999994, "first_improvement_epoch": 4}}
|
52
|
{"target_pattern": "contains_abc", "degraded_accuracy": 0.5, "improved_accuracy": 0.94, "improvement": 0.43999999999999995, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 5360, "learning_rate": 0.0885264373224075, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "contains_abc", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["contains_abc"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
1.629613,
0.388033,
0.436823,
0.421994,
-0.050135
],
[
1.792086,
0.378106,
0.626559,
0.531388,
0.147028
],
[
1.065834,
1.143501,
0.341078,
-0.360935,
-0.035245
],
[
1.921439,
0.09089,
-0.12209,
-0.014013,
-0.513727
],
[
0.153987,
-0.397447,
-0.385392,
-0.199481,
-0.30293
]
],
"network.0.bias": [
-0.889312,
-0.929206,
-0.55849,
-0.52083,
0.248679
],
"network.2.weight": [
[
-0.298389,
-0.218035,
-0.407338,
-0.531167,
-0.153903
],
[
0.208449,
-0.379101,
-0.287817,
0.226517,
-0.268346
],
[
0.712287,
0.574106,
1.313093,
0.417985,
-0.085384
],
[
-0.826194,
-0.367166,
-0.350328,
0.084968,
0.144953
],
[
-0.522894,
-0.41062,
-0.357813,
0.047618,
-0.329223
]
],
"network.2.bias": [
0.017586,
-0.287787,
-0.739241,
-0.48164,
-0.19606
],
"network.4.weight": [
[
0.135392,
0.131135,
-0.079763,
-0.025427,
0.037063
],
[
-0.137726,
-0.015066,
0.854399,
0.627491,
-0.0681
],
[
-0.172503,
-0.33418,
-0.503058,
0.040458,
-0.442478
],
[
0.188212,
0.253486,
-0.033568,
-0.424176,
-0.311681
],
[
-0.038526,
0.312718,
0.766932,
0.171136,
-0.08398
]
],
"network.4.bias": [
-0.206461,
-0.391151,
-0.08156,
-0.285212,
-0.293319
],
"network.6.weight": [
[
0.407293,
0.534342,
0.186076,
0.025177,
0.62598
],
[
0.362574,
-0.481999,
-0.005146,
0.280783,
-0.20035
],
[
-0.112728,
-0.261836,
-0.133338,
0.20084,
0.049836
],
[
0.006535,
0.523248,
-0.132554,
-0.207683,
0.720169
],
[
0.432219,
-0.349369,
-0.360552,
0.422018,
-0.276283
]
],
"network.6.bias": [
-0.181328,
-0.369831,
-0.36793,
-0.823074,
-0.148315
],
"network.8.weight": [
[
-0.728034,
0.283899,
0.400094,
-0.193774,
0.102237
],
[
-0.345331,
0.266681,
0.019296,
0.007627,
-0.336387
],
[
-0.150653,
-0.043089,
-0.445484,
0.140107,
-0.415579
],
[
0.263616,
-0.260024,
-0.392833,
0.71311,
0.211595
],
[
0.052277,
-0.349784,
-0.093528,
0.739114,
0.101322
]
],
"network.8.bias": [
-0.064017,
-0.33205,
-0.537624,
-0.818245,
-0.584147
],
"network.10.weight": [
[
-0.129677,
0.088166,
0.004919,
-0.169452,
-0.282585
],
[
-0.209759,
-0.385072,
-0.008127,
-0.350437,
0.229409
],
[
0.070583,
0.035198,
0.130482,
0.703988,
0.903468
],
[
0.350095,
-0.320903,
-0.313061,
-0.017138,
-0.188867
],
[
-0.033997,
-0.382445,
0.263066,
0.935368,
0.567408
]
],
"network.10.bias": [
-0.134491,
-0.312821,
-0.601583,
-0.312297,
-0.469227
],
"network.12.weight": [
[
0.33649,
0.175292,
-0.879659,
-0.165043,
-0.356907
]
],
"network.12.bias": [
1.108924
]
}
## Activation Signature
### 0
mean: [3.601317, 4.619619, 2.746597, 1.117239, -1.890359]
std: [4.068793, 4.700958, 3.708075, 3.846622, 1.425291]
### 2
mean: [-4.317779, -1.745489, 9.275125, -6.153624, -5.055650]
std: [5.160367, 1.249608, 11.054904, 5.830149, 4.955956]
### 4
mean: [-0.953550, 7.611765, -4.793607, -0.599603, 6.890329]
std: [0.875356, 9.376412, 5.520649, 0.368420, 8.416506]
### 6
mean: [8.251509, -5.452226, -2.029362, 8.177385, -4.740360]
std: [10.234889, 6.177978, 2.025774, 10.921013, 5.576810]
### 8
mean: [-7.702116, -3.127144, -0.619360, 7.295052, 5.996034]
std: [9.530118, 3.444616, 0.039063, 10.400609, 8.523481]
### 10
mean: [-3.143167, -1.530205, 10.234256, -1.602311, 10.035218]
std: [4.112444, 1.661885, 14.810390, 1.763687, 14.355552]
### 12
mean: [-11.724571]
std: [17.971619]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
contains_abc
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
1.629613,
0.388033,
0.436823,
0.421994,
-0.050135
],
[
1.792086,
0.378106,
0.626559,
0.531388,
0.147028
],
[
1.065834,
1.143501,
0.341078,
-0.360935,
-0.035245
],
[
1.921439,
0.09089,
-0.12209,
-0.014013,
-0.513727
],
[
0.153987,
-0.397447,
-0.385392,
-0.199481,
-0.30293
]
],
"network.0.bias": [
-0.889312,
-0.929206,
-0.55849,
-0.52083,
0.248679
],
"network.2.weight": [
[
-0.298389,
-0.218035,
-0.407338,
-0.531167,
-0.153903
],
[
0.208449,
-0.379101,
-0.287817,
0.226517,
-0.268346
],
[
0.712287,
0.574106,
1.313093,
0.417985,
-0.085384
],
[
-0.826194,
-0.367166,
-0.350328,
0.084968,
0.144953
],
[
-0.522894,
-0.41062,
-0.357813,
0.047618,
-0.329223
]
],
"network.2.bias": [
0.017586,
-0.287787,
-0.739241,
-0.48164,
-0.19606
],
"network.4.weight": [
[
0.135392,
0.131135,
-0.079763,
-0.025427,
0.037063
],
[
-0.137726,
-0.015066,
0.854399,
0.627491,
-0.0681
],
[
-0.172503,
-0.33418,
-0.503058,
0.040458,
-0.442478
],
[
0.188212,
0.253486,
-0.033568,
-0.424176,
-0.311681
],
[
-0.038526,
0.312718,
0.766932,
0.171136,
-0.08398
]
],
"network.4.bias": [
-0.206461,
-0.391151,
-0.08156,
-0.285212,
-0.293319
],
"network.6.weight": [
[
0.407293,
0.534342,
0.186076,
0.025177,
0.62598
],
[
0.362574,
-0.481999,
-0.005146,
0.280783,
-0.20035
],
[
-0.112728,
-0.261836,
-0.133338,
0.20084,
0.049836
],
[
0.006535,
0.523248,
-0.132554,
-0.207683,
0.720169
],
[
0.432219,
-0.349369,
-0.360552,
0.422018,
-0.276283
]
],
"network.6.bias": [
-0.181328,
-0.369831,
-0.36793,
-0.823074,
-0.148315
],
"network.8.weight": [
[
-0.728034,
0.283899,
0.400094,
-0.193774,
0.102237
],
[
-0.345331,
0.266681,
0.019296,
0.007627,
-0.336387
],
[
-0.150653,
-0.043089,
-0.445484,
0.140107,
-0.415579
],
[
0.263616,
-0.260024,
-0.392833,
0.71311,
0.211595
],
[
0.052277,
-0.349784,
-0.093528,
0.739114,
0.101322
]
],
"network.8.bias": [
-0.064017,
-0.33205,
-0.537624,
-0.818245,
-0.584147
],
"network.10.weight": [
[
-0.129677,
0.088166,
0.004919,
-0.169452,
-0.282585
],
[
-0.209759,
-0.385072,
-0.008127,
-0.350437,
0.229409
],
[
0.070583,
0.035198,
0.130482,
0.703988,
0.903468
],
[
0.350095,
-0.320903,
-0.313061,
-0.017138,
-0.188867
],
[
-0.033997,
-0.382445,
0.263066,
0.935368,
0.567408
]
],
"network.10.bias": [
-0.134491,
-0.312821,
-0.601583,
-0.312297,
-0.469227
],
"network.12.weight": [
[
0.33649,
0.175292,
-0.879659,
-0.165043,
-0.356907
]
],
"network.12.bias": [
1.108924
]
}
## Activation Signature
### 0
mean: [3.601317, 4.619619, 2.746597, 1.117239, -1.890359]
std: [4.068793, 4.700958, 3.708075, 3.846622, 1.425291]
### 2
mean: [-4.317779, -1.745489, 9.275125, -6.153624, -5.055650]
std: [5.160367, 1.249608, 11.054904, 5.830149, 4.955956]
### 4
mean: [-0.953550, 7.611765, -4.793607, -0.599603, 6.890329]
std: [0.875356, 9.376412, 5.520649, 0.368420, 8.416506]
### 6
mean: [8.251509, -5.452226, -2.029362, 8.177385, -4.740360]
std: [10.234889, 6.177978, 2.025774, 10.921013, 5.576810]
### 8
mean: [-7.702116, -3.127144, -0.619360, 7.295052, 5.996034]
std: [9.530118, 3.444616, 0.039063, 10.400609, 8.523481]
### 10
mean: [-3.143167, -1.530205, 10.234256, -1.602311, 10.035218]
std: [4.112444, 1.661885, 14.810390, 1.763687, 14.355552]
### 12
mean: [-11.724571]
std: [17.971619]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
contains_abc
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 3.6013174057006836, "std": 4.068793296813965}, "1": {"mean": 4.619618892669678, "std": 4.700958251953125}, "2": {"mean": 2.7465968132019043, "std": 3.7080745697021484}, "3": {"mean": 1.1172387599945068, "std": 3.8466224670410156}, "4": {"mean": -1.890358805656433, "std": 1.4252911806106567}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -4.317779064178467, "std": 5.160367488861084}, "1": {"mean": -1.7454890012741089, "std": 1.249608039855957}, "2": {"mean": 9.275124549865723, "std": 11.054903984069824}, "3": {"mean": -6.153624057769775, "std": 5.830149173736572}, "4": {"mean": -5.055649757385254, "std": 4.955955982208252}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.9535495042800903, "std": 0.8753560781478882}, "1": {"mean": 7.611765384674072, "std": 9.376412391662598}, "2": {"mean": -4.793607234954834, "std": 5.520648956298828}, "3": {"mean": -0.5996025800704956, "std": 0.36841994524002075}, "4": {"mean": 6.890329360961914, "std": 8.416505813598633}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 8.251508712768555, "std": 10.234889030456543}, "1": {"mean": -5.452225685119629, "std": 6.177978038787842}, "2": {"mean": -2.029362201690674, "std": 2.025773763656616}, "3": {"mean": 8.177385330200195, "std": 10.921012878417969}, "4": {"mean": -4.740360260009766, "std": 5.576809883117676}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -7.7021164894104, "std": 9.530117988586426}, "1": {"mean": -3.1271440982818604, "std": 3.4446163177490234}, "2": {"mean": -0.6193603277206421, "std": 0.039063069969415665}, "3": {"mean": 7.2950520515441895, "std": 10.400609016418457}, "4": {"mean": 5.996034145355225, "std": 8.523481369018555}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -3.1431667804718018, "std": 4.1124444007873535}, "1": {"mean": -1.530205488204956, "std": 1.661885380744934}, "2": {"mean": 10.23425579071045, "std": 14.81039047241211}, "3": {"mean": -1.6023108959197998, "std": 1.7636865377426147}, "4": {"mean": 10.035218238830566, "std": 14.355551719665527}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -11.724571228027344, "std": 17.97161865234375}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[1.629613, 0.388033, 0.436823, 0.421994, -0.050135], [1.792086, 0.378106, 0.626559, 0.531388, 0.147028], [1.065834, 1.143501, 0.341078, -0.360935, -0.035245], [1.921439, 0.09089, -0.12209, -0.014013, -0.513727], [0.153987, -0.397447, -0.385392, -0.199481, -0.30293]], "network.0.bias": [-0.889312, -0.929206, -0.55849, -0.52083, 0.248679], "network.2.weight": [[-0.298389, -0.218035, -0.407338, -0.531167, -0.153903], [0.208449, -0.379101, -0.287817, 0.226517, -0.268346], [0.712287, 0.574106, 1.313093, 0.417985, -0.085384], [-0.826194, -0.367166, -0.350328, 0.084968, 0.144953], [-0.522894, -0.41062, -0.357813, 0.047618, -0.329223]], "network.2.bias": [0.017586, -0.287787, -0.739241, -0.48164, -0.19606], "network.4.weight": [[0.135392, 0.131135, -0.079763, -0.025427, 0.037063], [-0.137726, -0.015066, 0.854399, 0.627491, -0.0681], [-0.172503, -0.33418, -0.503058, 0.040458, -0.442478], [0.188212, 0.253486, -0.033568, -0.424176, -0.311681], [-0.038526, 0.312718, 0.766932, 0.171136, -0.08398]], "network.4.bias": [-0.206461, -0.391151, -0.08156, -0.285212, -0.293319], "network.6.weight": [[0.407293, 0.534342, 0.186076, 0.025177, 0.62598], [0.362574, -0.481999, -0.005146, 0.280783, -0.20035], [-0.112728, -0.261836, -0.133338, 0.20084, 0.049836], [0.006535, 0.523248, -0.132554, -0.207683, 0.720169], [0.432219, -0.349369, -0.360552, 0.422018, -0.276283]], "network.6.bias": [-0.181328, -0.369831, -0.36793, -0.823074, -0.148315], "network.8.weight": [[-0.728034, 0.283899, 0.400094, -0.193774, 0.102237], [-0.345331, 0.266681, 0.019296, 0.007627, -0.336387], [-0.150653, -0.043089, -0.445484, 0.140107, -0.415579], [0.263616, -0.260024, -0.392833, 0.71311, 0.211595], [0.052277, -0.349784, -0.093528, 0.739114, 0.101322]], "network.8.bias": [-0.064017, -0.33205, -0.537624, -0.818245, -0.584147], "network.10.weight": [[-0.129677, 0.088166, 0.004919, -0.169452, -0.282585], [-0.209759, -0.385072, -0.008127, -0.350437, 0.229409], [0.070583, 0.035198, 0.130482, 0.703988, 0.903468], [0.350095, -0.320903, -0.313061, -0.017138, -0.188867], [-0.033997, -0.382445, 0.263066, 0.935368, 0.567408]], "network.10.bias": [-0.134491, -0.312821, -0.601583, -0.312297, -0.469227], "network.12.weight": [[0.33649, 0.175292, -0.879659, -0.165043, -0.356907]], "network.12.bias": [1.108924]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.700906902551651, "train_acc": 0.435, "val_loss": 0.692647397518158, "val_acc": 0.5}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6685549914836884, "train_acc": 0.565, "val_loss": 0.7511438131332397, "val_acc": 0.5}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6455483734607697, "train_acc": 0.565, "val_loss": 0.649824321269989, "val_acc": 0.5}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.5913639664649963, "train_acc": 0.5, "val_loss": 0.5097365379333496, "val_acc": 0.66}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.4904482066631317, "train_acc": 0.69, "val_loss": 0.46585914492607117, "val_acc": 0.88}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.44949135184288025, "train_acc": 0.865, "val_loss": 0.3993456959724426, "val_acc": 0.84}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.39273446798324585, "train_acc": 0.835, "val_loss": 0.3575840890407562, "val_acc": 0.86}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.3256186544895172, "train_acc": 0.88, "val_loss": 0.3428063690662384, "val_acc": 0.9}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.29003994166851044, "train_acc": 0.91, "val_loss": 0.25508180260658264, "val_acc": 0.9}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.3070850968360901, "train_acc": 0.875, "val_loss": 0.22789207100868225, "val_acc": 0.94}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.2327297031879425, "train_acc": 0.955, "val_loss": 0.23940841853618622, "val_acc": 0.94}, {"stage": "improved", "epoch": 8, "global_epoch": 11, "train_loss": 0.2391916811466217, "train_acc": 0.915, "val_loss": 0.25214603543281555, "val_acc": 0.94}, {"stage": "improved", "epoch": 9, "global_epoch": 12, "train_loss": 0.25671058893203735, "train_acc": 0.915, "val_loss": 0.21754279732704163, "val_acc": 0.94}], "summary": {"total_epochs": 13, "degraded_epochs": 3, "improved_epochs": 10, "patterns": ["contains_abc"], "degraded_stage": {"initial_val_loss": 0.692647397518158, "final_val_loss": 0.649824321269989, "initial_val_acc": 0.5, "final_val_acc": 0.5, "best_val_acc": 0.5}, "improved_stage": {"initial_val_loss": 0.5097365379333496, "final_val_loss": 0.21754279732704163, "initial_val_acc": 0.66, "final_val_acc": 0.94, "best_val_acc": 0.94, "best_epoch": 9}, "improvement": 0.43999999999999995, "first_improvement_epoch": 2}}
|
53
|
{"target_pattern": "no_repeats", "degraded_accuracy": 0.74, "improved_accuracy": 0.92, "improvement": 0.18000000000000005, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 9169, "learning_rate": 0.02856298291523507, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "no_repeats", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["no_repeats"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.065299,
-0.074194,
-0.10302,
-0.29292,
-0.460209
],
[
-0.626203,
0.238692,
0.32479,
0.305247,
0.235595
],
[
-1.018884,
0.06704,
0.162449,
0.057626,
0.398798
],
[
0.213343,
-0.94947,
-0.054145,
0.248576,
0.24854
],
[
-0.453497,
-0.340124,
0.674925,
0.315594,
0.261434
],
[
0.648151,
0.627438,
-0.253199,
-0.044029,
-0.101481
]
],
"network.0.bias": [
0.214394,
-0.080583,
0.318612,
0.666324,
-0.23896,
0.385612
],
"network.2.weight": [
[
0.895415,
0.45356,
0.446296,
-0.87101,
0.283021,
-0.364376
],
[
0.380109,
0.149942,
0.084301,
-0.722832,
0.383324,
-0.039986
],
[
-0.302287,
-0.203987,
-0.316437,
0.281583,
-0.000419,
-0.478354
],
[
0.516264,
-0.488158,
-0.258936,
-0.35952,
0.188294,
-0.374703
],
[
-0.39077,
-0.112921,
-0.373319,
0.476498,
0.498261,
0.108444
],
[
-0.399192,
-0.185852,
-0.345165,
0.17665,
-0.626121,
0.724628
]
],
"network.2.bias": [
-0.144893,
-0.342232,
-0.063302,
-0.084924,
0.414184,
0.595165
],
"network.4.weight": [
[
0.645109,
0.520766,
0.711798,
-0.333005,
-0.582937,
-0.016525
],
[
0.012132,
-0.078083,
-0.546575,
-0.601058,
0.624437,
0.238484
],
[
-0.53423,
0.191339,
0.687657,
0.361907,
0.40735,
-0.165116
],
[
-0.090785,
-0.342122,
0.238198,
0.252467,
0.0235,
-0.181383
],
[
0.427716,
-0.025043,
0.856041,
0.001399,
-0.086648,
-0.244094
],
[
-0.092681,
0.0038,
-0.804652,
-0.121037,
0.420945,
0.638179
]
],
"network.4.bias": [
-0.295043,
0.217569,
-0.508103,
-0.669409,
-0.185252,
0.619996
],
"network.6.weight": [
[
0.714241,
-0.016929,
0.294509,
-0.178798,
0.246724,
-0.618202
],
[
0.402298,
0.143961,
0.300636,
0.342805,
-0.352793,
-0.311283
],
[
-0.181191,
0.556888,
-0.27229,
-0.282377,
-0.163654,
0.722601
],
[
-0.293287,
0.439559,
-0.789109,
-0.343593,
0.083204,
0.325654
],
[
-0.094737,
0.263477,
-0.430333,
-0.515599,
-0.224386,
0.687994
],
[
0.384811,
-0.432924,
0.453176,
0.021374,
0.503194,
-0.318916
]
],
"network.6.bias": [
0.032466,
-0.307002,
0.393688,
0.617282,
0.374123,
0.226356
],
"network.8.weight": [
[
0.654615,
0.361356,
-0.452991,
-0.607923,
-0.815383,
0.452916
]
],
"network.8.bias": [
-0.665059
]
}
## Activation Signature
### 0
mean: [-1.250482, 1.198696, 0.127593, -0.071841, 0.985958, 1.587050]
std: [1.186582, 1.522862, 1.997800, 1.675079, 1.701035, 1.899388]
### 2
mean: [0.060736, -0.071824, -1.155880, -1.492457, 1.012668, 0.597499]
std: [1.584538, 0.828538, 0.956503, 0.812166, 0.717616, 2.193154]
### 4
mean: [-0.390608, 1.143836, -0.697821, -1.012460, -0.362124, 1.738058]
std: [0.945005, 0.543884, 0.500166, 0.279211, 0.653938, 1.035011]
### 6
mean: [-0.923547, -0.711191, 2.197964, 1.702628, 1.883704, -0.744456]
std: [0.967586, 0.354921, 1.122081, 0.630787, 0.921229, 0.762753]
### 8
mean: [-4.173319]
std: [1.875767]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
no_repeats
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.065299,
-0.074194,
-0.10302,
-0.29292,
-0.460209
],
[
-0.626203,
0.238692,
0.32479,
0.305247,
0.235595
],
[
-1.018884,
0.06704,
0.162449,
0.057626,
0.398798
],
[
0.213343,
-0.94947,
-0.054145,
0.248576,
0.24854
],
[
-0.453497,
-0.340124,
0.674925,
0.315594,
0.261434
],
[
0.648151,
0.627438,
-0.253199,
-0.044029,
-0.101481
]
],
"network.0.bias": [
0.214394,
-0.080583,
0.318612,
0.666324,
-0.23896,
0.385612
],
"network.2.weight": [
[
0.895415,
0.45356,
0.446296,
-0.87101,
0.283021,
-0.364376
],
[
0.380109,
0.149942,
0.084301,
-0.722832,
0.383324,
-0.039986
],
[
-0.302287,
-0.203987,
-0.316437,
0.281583,
-0.000419,
-0.478354
],
[
0.516264,
-0.488158,
-0.258936,
-0.35952,
0.188294,
-0.374703
],
[
-0.39077,
-0.112921,
-0.373319,
0.476498,
0.498261,
0.108444
],
[
-0.399192,
-0.185852,
-0.345165,
0.17665,
-0.626121,
0.724628
]
],
"network.2.bias": [
-0.144893,
-0.342232,
-0.063302,
-0.084924,
0.414184,
0.595165
],
"network.4.weight": [
[
0.645109,
0.520766,
0.711798,
-0.333005,
-0.582937,
-0.016525
],
[
0.012132,
-0.078083,
-0.546575,
-0.601058,
0.624437,
0.238484
],
[
-0.53423,
0.191339,
0.687657,
0.361907,
0.40735,
-0.165116
],
[
-0.090785,
-0.342122,
0.238198,
0.252467,
0.0235,
-0.181383
],
[
0.427716,
-0.025043,
0.856041,
0.001399,
-0.086648,
-0.244094
],
[
-0.092681,
0.0038,
-0.804652,
-0.121037,
0.420945,
0.638179
]
],
"network.4.bias": [
-0.295043,
0.217569,
-0.508103,
-0.669409,
-0.185252,
0.619996
],
"network.6.weight": [
[
0.714241,
-0.016929,
0.294509,
-0.178798,
0.246724,
-0.618202
],
[
0.402298,
0.143961,
0.300636,
0.342805,
-0.352793,
-0.311283
],
[
-0.181191,
0.556888,
-0.27229,
-0.282377,
-0.163654,
0.722601
],
[
-0.293287,
0.439559,
-0.789109,
-0.343593,
0.083204,
0.325654
],
[
-0.094737,
0.263477,
-0.430333,
-0.515599,
-0.224386,
0.687994
],
[
0.384811,
-0.432924,
0.453176,
0.021374,
0.503194,
-0.318916
]
],
"network.6.bias": [
0.032466,
-0.307002,
0.393688,
0.617282,
0.374123,
0.226356
],
"network.8.weight": [
[
0.654615,
0.361356,
-0.452991,
-0.607923,
-0.815383,
0.452916
]
],
"network.8.bias": [
-0.665059
]
}
## Activation Signature
### 0
mean: [-1.250482, 1.198696, 0.127593, -0.071841, 0.985958, 1.587050]
std: [1.186582, 1.522862, 1.997800, 1.675079, 1.701035, 1.899388]
### 2
mean: [0.060736, -0.071824, -1.155880, -1.492457, 1.012668, 0.597499]
std: [1.584538, 0.828538, 0.956503, 0.812166, 0.717616, 2.193154]
### 4
mean: [-0.390608, 1.143836, -0.697821, -1.012460, -0.362124, 1.738058]
std: [0.945005, 0.543884, 0.500166, 0.279211, 0.653938, 1.035011]
### 6
mean: [-0.923547, -0.711191, 2.197964, 1.702628, 1.883704, -0.744456]
std: [0.967586, 0.354921, 1.122081, 0.630787, 0.921229, 0.762753]
### 8
mean: [-4.173319]
std: [1.875767]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
no_repeats
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.2504818439483643, "std": 1.1865822076797485}, "1": {"mean": 1.1986956596374512, "std": 1.5228615999221802}, "2": {"mean": 0.12759336829185486, "std": 1.9977995157241821}, "3": {"mean": -0.0718405544757843, "std": 1.6750794649124146}, "4": {"mean": 0.9859582781791687, "std": 1.7010352611541748}, "5": {"mean": 1.5870500802993774, "std": 1.8993884325027466}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.06073581799864769, "std": 1.5845376253128052}, "1": {"mean": -0.07182357460260391, "std": 0.8285382986068726}, "2": {"mean": -1.1558802127838135, "std": 0.9565026164054871}, "3": {"mean": -1.492457389831543, "std": 0.8121656179428101}, "4": {"mean": 1.0126675367355347, "std": 0.717616081237793}, "5": {"mean": 0.5974991321563721, "std": 2.1931538581848145}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.3906082212924957, "std": 0.9450045228004456}, "1": {"mean": 1.1438356637954712, "std": 0.543883740901947}, "2": {"mean": -0.6978206634521484, "std": 0.5001664161682129}, "3": {"mean": -1.0124601125717163, "std": 0.27921077609062195}, "4": {"mean": -0.3621237874031067, "std": 0.6539379954338074}, "5": {"mean": 1.7380579710006714, "std": 1.0350106954574585}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.9235466122627258, "std": 0.967586100101471}, "1": {"mean": -0.7111911177635193, "std": 0.3549211323261261}, "2": {"mean": 2.1979639530181885, "std": 1.1220811605453491}, "3": {"mean": 1.7026275396347046, "std": 0.6307870149612427}, "4": {"mean": 1.8837040662765503, "std": 0.9212290048599243}, "5": {"mean": -0.744455873966217, "std": 0.7627533078193665}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -4.173319339752197, "std": 1.8757671117782593}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.065299, -0.074194, -0.10302, -0.29292, -0.460209], [-0.626203, 0.238692, 0.32479, 0.305247, 0.235595], [-1.018884, 0.06704, 0.162449, 0.057626, 0.398798], [0.213343, -0.94947, -0.054145, 0.248576, 0.24854], [-0.453497, -0.340124, 0.674925, 0.315594, 0.261434], [0.648151, 0.627438, -0.253199, -0.044029, -0.101481]], "network.0.bias": [0.214394, -0.080583, 0.318612, 0.666324, -0.23896, 0.385612], "network.2.weight": [[0.895415, 0.45356, 0.446296, -0.87101, 0.283021, -0.364376], [0.380109, 0.149942, 0.084301, -0.722832, 0.383324, -0.039986], [-0.302287, -0.203987, -0.316437, 0.281583, -0.000419, -0.478354], [0.516264, -0.488158, -0.258936, -0.35952, 0.188294, -0.374703], [-0.39077, -0.112921, -0.373319, 0.476498, 0.498261, 0.108444], [-0.399192, -0.185852, -0.345165, 0.17665, -0.626121, 0.724628]], "network.2.bias": [-0.144893, -0.342232, -0.063302, -0.084924, 0.414184, 0.595165], "network.4.weight": [[0.645109, 0.520766, 0.711798, -0.333005, -0.582937, -0.016525], [0.012132, -0.078083, -0.546575, -0.601058, 0.624437, 0.238484], [-0.53423, 0.191339, 0.687657, 0.361907, 0.40735, -0.165116], [-0.090785, -0.342122, 0.238198, 0.252467, 0.0235, -0.181383], [0.427716, -0.025043, 0.856041, 0.001399, -0.086648, -0.244094], [-0.092681, 0.0038, -0.804652, -0.121037, 0.420945, 0.638179]], "network.4.bias": [-0.295043, 0.217569, -0.508103, -0.669409, -0.185252, 0.619996], "network.6.weight": [[0.714241, -0.016929, 0.294509, -0.178798, 0.246724, -0.618202], [0.402298, 0.143961, 0.300636, 0.342805, -0.352793, -0.311283], [-0.181191, 0.556888, -0.27229, -0.282377, -0.163654, 0.722601], [-0.293287, 0.439559, -0.789109, -0.343593, 0.083204, 0.325654], [-0.094737, 0.263477, -0.430333, -0.515599, -0.224386, 0.687994], [0.384811, -0.432924, 0.453176, 0.021374, 0.503194, -0.318916]], "network.6.bias": [0.032466, -0.307002, 0.393688, 0.617282, 0.374123, 0.226356], "network.8.weight": [[0.654615, 0.361356, -0.452991, -0.607923, -0.815383, 0.452916]], "network.8.bias": [-0.665059]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6870701909065247, "train_acc": 0.555, "val_loss": 0.6929000616073608, "val_acc": 0.52}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6878697276115417, "train_acc": 0.555, "val_loss": 0.687052309513092, "val_acc": 0.52}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.674715667963028, "train_acc": 0.58, "val_loss": 0.6615461111068726, "val_acc": 0.7}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6674387753009796, "train_acc": 0.625, "val_loss": 0.6426857709884644, "val_acc": 0.74}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6264405250549316, "train_acc": 0.705, "val_loss": 0.6130397319793701, "val_acc": 0.76}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.5919950306415558, "train_acc": 0.715, "val_loss": 0.5789476037025452, "val_acc": 0.78}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.5503651797771454, "train_acc": 0.8, "val_loss": 0.5126643180847168, "val_acc": 0.8}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.4808283895254135, "train_acc": 0.805, "val_loss": 0.42870521545410156, "val_acc": 0.86}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.40968042612075806, "train_acc": 0.845, "val_loss": 0.38353049755096436, "val_acc": 0.82}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.3757261633872986, "train_acc": 0.84, "val_loss": 0.2886218726634979, "val_acc": 0.88}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.33537209033966064, "train_acc": 0.865, "val_loss": 0.2539495825767517, "val_acc": 0.86}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.37758147716522217, "train_acc": 0.855, "val_loss": 0.25443676114082336, "val_acc": 0.88}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.28493520617485046, "train_acc": 0.89, "val_loss": 0.24582570791244507, "val_acc": 0.92}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.32686471939086914, "train_acc": 0.845, "val_loss": 0.2604953646659851, "val_acc": 0.86}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["no_repeats"], "degraded_stage": {"initial_val_loss": 0.6929000616073608, "final_val_loss": 0.6426857709884644, "initial_val_acc": 0.52, "final_val_acc": 0.74, "best_val_acc": 0.74}, "improved_stage": {"initial_val_loss": 0.6130397319793701, "final_val_loss": 0.2604953646659851, "initial_val_acc": 0.76, "final_val_acc": 0.86, "best_val_acc": 0.92, "best_epoch": 12}, "improvement": 0.18000000000000005, "first_improvement_epoch": 3}}
|
54
|
{"target_pattern": "has_majority", "degraded_accuracy": 0.48, "improved_accuracy": 0.78, "improvement": 0.30000000000000004, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 7506, "learning_rate": 0.05975136536946539, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "has_majority", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["has_majority"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.327549,
-1.083559,
0.380645,
0.18442,
-0.91056
],
[
0.089496,
0.497805,
0.564761,
-0.005037,
-0.016307
],
[
-0.387039,
0.3142,
0.232852,
0.502417,
-0.005017
],
[
-0.437767,
0.276942,
-0.370443,
-0.641839,
-0.696545
],
[
0.115424,
0.030307,
0.372086,
-0.960215,
0.102389
],
[
-0.198311,
0.479039,
-0.620284,
-0.479769,
0.179563
],
[
-0.230279,
-0.002171,
-0.096226,
-0.257413,
-0.392771
]
],
"network.0.bias": [
0.177993,
-0.064067,
-0.59667,
0.022977,
0.09618,
-0.17186,
-0.078431
],
"network.2.weight": [
[
0.850331,
-0.356137,
-0.053621,
0.590554,
0.352509,
0.231733,
0.253385
],
[
-0.359012,
0.243978,
0.278778,
-0.166438,
-0.844235,
-0.080629,
-0.377375
],
[
-0.077039,
-0.159758,
0.332805,
-0.182454,
-0.272325,
-0.151041,
-0.327258
],
[
-0.094511,
0.213801,
0.636691,
0.018813,
-0.497936,
-0.337174,
-0.008319
],
[
0.33482,
0.014921,
0.089224,
0.344422,
-0.739899,
-0.222934,
-0.275323
],
[
0.644009,
-0.388801,
0.139979,
-0.26698,
0.148427,
-0.234605,
-0.361689
],
[
0.420103,
-0.122903,
-0.124313,
0.228086,
0.174821,
0.763329,
-0.272892
]
],
"network.2.bias": [
0.52699,
-0.005364,
0.131117,
-0.234199,
-0.626884,
0.264795,
0.419041
],
"network.4.weight": [
[
-0.155262,
0.043378,
-0.072589,
-0.17726,
-0.370149,
-0.024485,
-0.018568
],
[
-0.566723,
0.426357,
0.335675,
0.459227,
-0.279242,
-0.636927,
-0.23372
],
[
-0.352579,
-0.023429,
-0.361056,
-0.343874,
-0.324154,
-0.322074,
-0.055951
],
[
0.925938,
-0.159139,
0.23821,
-0.286497,
-0.446345,
0.757425,
0.196756
],
[
0.439972,
0.392831,
0.081561,
0.255903,
-0.052684,
-0.341379,
-0.406873
],
[
0.115731,
0.11816,
0.006576,
0.346141,
0.46774,
-0.314147,
-0.387132
],
[
0.015314,
-0.045227,
-0.459786,
-0.211884,
-0.149655,
-0.183065,
-0.312843
]
],
"network.4.bias": [
-0.330239,
0.07841,
0.004666,
0.249772,
-0.116027,
0.146287,
-0.035074
],
"network.6.weight": [
[
-0.156665,
0.273307,
0.352111,
0.081119,
0.422342,
0.403154,
-0.032775
],
[
0.287991,
-0.539494,
-0.173898,
0.619834,
-0.063475,
-0.297495,
-0.141716
],
[
0.04425,
0.277439,
-0.195548,
-0.299513,
0.494984,
0.333486,
-0.262537
],
[
-0.059028,
0.428109,
0.313416,
-0.083315,
0.474458,
0.392011,
-0.115764
],
[
-0.224431,
-0.219104,
0.162262,
0.672488,
-0.124051,
0.313514,
-0.030655
],
[
0.221852,
0.180987,
0.321551,
-0.109962,
0.299032,
0.235764,
0.507471
],
[
0.118877,
-0.37524,
0.01308,
-0.012405,
0.090499,
-0.303624,
-0.455166
]
],
"network.6.bias": [
-0.300557,
0.14966,
-0.238246,
-0.118568,
0.454585,
-0.091558,
-0.287717
],
"network.8.weight": [
[
-0.150842,
-0.004108,
0.089471,
-0.147752,
-0.108343,
-0.564326,
-0.264081
],
[
-0.510134,
0.549692,
-0.472999,
-0.440898,
0.347982,
-0.127408,
0.138025
],
[
0.439903,
0.150382,
0.307352,
0.285713,
-0.240525,
0.289841,
-0.460809
],
[
-0.201581,
0.838425,
-0.429012,
-0.602082,
0.50572,
-0.062251,
0.282692
],
[
0.131776,
-0.264891,
0.367583,
-0.111057,
-0.733935,
0.112829,
0.112512
],
[
-0.207102,
0.622262,
0.039772,
-0.239525,
0.379169,
-0.367617,
0.312141
],
[
0.335528,
0.237609,
0.407358,
0.463525,
-0.169748,
-0.00757,
-0.412237
]
],
"network.8.bias": [
-0.593881,
-0.200657,
0.108122,
-0.342561,
-0.11874,
-0.129998,
0.142864
],
"network.10.weight": [
[
0.226575,
0.308482,
-0.454197,
0.502802,
-0.304443,
0.626089,
-0.460543
]
],
"network.10.bias": [
0.096274
]
}
## Activation Signature
### 0
mean: [-1.321296, 2.103226, 1.051842, -3.026427, -0.860554, -1.648813, -1.606387]
std: [2.252778, 1.656081, 1.528937, 2.395588, 2.162817, 1.480875, 1.183597]
### 2
mean: [0.137568, 0.359741, 0.055811, 0.747581, -0.739960, -0.139920, 0.252873]
std: [0.830926, 0.970533, 0.560517, 1.133587, 0.641405, 0.700328, 0.456072]
### 4
mean: [-0.555794, 0.380913, -0.629163, 0.544028, 0.321068, 0.376537, -0.503660]
std: [0.171855, 1.175311, 0.471273, 0.977229, 0.548578, 0.517794, 0.331829]
### 6
mean: [0.261540, 0.063608, 0.066067, 0.450200, 0.848315, 0.163676, -0.638915]
std: [0.608178, 1.032353, 0.813811, 0.838612, 0.622835, 0.477112, 0.420317]
### 8
mean: [-0.936806, -0.274081, 0.467139, -0.097010, -0.714057, 0.191332, 0.618701]
std: [0.329578, 1.317418, 0.805304, 1.456252, 0.774991, 0.902768, 0.771294]
### 10
mean: [0.242791]
std: [1.412363]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
has_majority
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.327549,
-1.083559,
0.380645,
0.18442,
-0.91056
],
[
0.089496,
0.497805,
0.564761,
-0.005037,
-0.016307
],
[
-0.387039,
0.3142,
0.232852,
0.502417,
-0.005017
],
[
-0.437767,
0.276942,
-0.370443,
-0.641839,
-0.696545
],
[
0.115424,
0.030307,
0.372086,
-0.960215,
0.102389
],
[
-0.198311,
0.479039,
-0.620284,
-0.479769,
0.179563
],
[
-0.230279,
-0.002171,
-0.096226,
-0.257413,
-0.392771
]
],
"network.0.bias": [
0.177993,
-0.064067,
-0.59667,
0.022977,
0.09618,
-0.17186,
-0.078431
],
"network.2.weight": [
[
0.850331,
-0.356137,
-0.053621,
0.590554,
0.352509,
0.231733,
0.253385
],
[
-0.359012,
0.243978,
0.278778,
-0.166438,
-0.844235,
-0.080629,
-0.377375
],
[
-0.077039,
-0.159758,
0.332805,
-0.182454,
-0.272325,
-0.151041,
-0.327258
],
[
-0.094511,
0.213801,
0.636691,
0.018813,
-0.497936,
-0.337174,
-0.008319
],
[
0.33482,
0.014921,
0.089224,
0.344422,
-0.739899,
-0.222934,
-0.275323
],
[
0.644009,
-0.388801,
0.139979,
-0.26698,
0.148427,
-0.234605,
-0.361689
],
[
0.420103,
-0.122903,
-0.124313,
0.228086,
0.174821,
0.763329,
-0.272892
]
],
"network.2.bias": [
0.52699,
-0.005364,
0.131117,
-0.234199,
-0.626884,
0.264795,
0.419041
],
"network.4.weight": [
[
-0.155262,
0.043378,
-0.072589,
-0.17726,
-0.370149,
-0.024485,
-0.018568
],
[
-0.566723,
0.426357,
0.335675,
0.459227,
-0.279242,
-0.636927,
-0.23372
],
[
-0.352579,
-0.023429,
-0.361056,
-0.343874,
-0.324154,
-0.322074,
-0.055951
],
[
0.925938,
-0.159139,
0.23821,
-0.286497,
-0.446345,
0.757425,
0.196756
],
[
0.439972,
0.392831,
0.081561,
0.255903,
-0.052684,
-0.341379,
-0.406873
],
[
0.115731,
0.11816,
0.006576,
0.346141,
0.46774,
-0.314147,
-0.387132
],
[
0.015314,
-0.045227,
-0.459786,
-0.211884,
-0.149655,
-0.183065,
-0.312843
]
],
"network.4.bias": [
-0.330239,
0.07841,
0.004666,
0.249772,
-0.116027,
0.146287,
-0.035074
],
"network.6.weight": [
[
-0.156665,
0.273307,
0.352111,
0.081119,
0.422342,
0.403154,
-0.032775
],
[
0.287991,
-0.539494,
-0.173898,
0.619834,
-0.063475,
-0.297495,
-0.141716
],
[
0.04425,
0.277439,
-0.195548,
-0.299513,
0.494984,
0.333486,
-0.262537
],
[
-0.059028,
0.428109,
0.313416,
-0.083315,
0.474458,
0.392011,
-0.115764
],
[
-0.224431,
-0.219104,
0.162262,
0.672488,
-0.124051,
0.313514,
-0.030655
],
[
0.221852,
0.180987,
0.321551,
-0.109962,
0.299032,
0.235764,
0.507471
],
[
0.118877,
-0.37524,
0.01308,
-0.012405,
0.090499,
-0.303624,
-0.455166
]
],
"network.6.bias": [
-0.300557,
0.14966,
-0.238246,
-0.118568,
0.454585,
-0.091558,
-0.287717
],
"network.8.weight": [
[
-0.150842,
-0.004108,
0.089471,
-0.147752,
-0.108343,
-0.564326,
-0.264081
],
[
-0.510134,
0.549692,
-0.472999,
-0.440898,
0.347982,
-0.127408,
0.138025
],
[
0.439903,
0.150382,
0.307352,
0.285713,
-0.240525,
0.289841,
-0.460809
],
[
-0.201581,
0.838425,
-0.429012,
-0.602082,
0.50572,
-0.062251,
0.282692
],
[
0.131776,
-0.264891,
0.367583,
-0.111057,
-0.733935,
0.112829,
0.112512
],
[
-0.207102,
0.622262,
0.039772,
-0.239525,
0.379169,
-0.367617,
0.312141
],
[
0.335528,
0.237609,
0.407358,
0.463525,
-0.169748,
-0.00757,
-0.412237
]
],
"network.8.bias": [
-0.593881,
-0.200657,
0.108122,
-0.342561,
-0.11874,
-0.129998,
0.142864
],
"network.10.weight": [
[
0.226575,
0.308482,
-0.454197,
0.502802,
-0.304443,
0.626089,
-0.460543
]
],
"network.10.bias": [
0.096274
]
}
## Activation Signature
### 0
mean: [-1.321296, 2.103226, 1.051842, -3.026427, -0.860554, -1.648813, -1.606387]
std: [2.252778, 1.656081, 1.528937, 2.395588, 2.162817, 1.480875, 1.183597]
### 2
mean: [0.137568, 0.359741, 0.055811, 0.747581, -0.739960, -0.139920, 0.252873]
std: [0.830926, 0.970533, 0.560517, 1.133587, 0.641405, 0.700328, 0.456072]
### 4
mean: [-0.555794, 0.380913, -0.629163, 0.544028, 0.321068, 0.376537, -0.503660]
std: [0.171855, 1.175311, 0.471273, 0.977229, 0.548578, 0.517794, 0.331829]
### 6
mean: [0.261540, 0.063608, 0.066067, 0.450200, 0.848315, 0.163676, -0.638915]
std: [0.608178, 1.032353, 0.813811, 0.838612, 0.622835, 0.477112, 0.420317]
### 8
mean: [-0.936806, -0.274081, 0.467139, -0.097010, -0.714057, 0.191332, 0.618701]
std: [0.329578, 1.317418, 0.805304, 1.456252, 0.774991, 0.902768, 0.771294]
### 10
mean: [0.242791]
std: [1.412363]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
has_majority
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.321295976638794, "std": 2.2527780532836914}, "1": {"mean": 2.1032261848449707, "std": 1.6560808420181274}, "2": {"mean": 1.0518416166305542, "std": 1.5289374589920044}, "3": {"mean": -3.0264267921447754, "std": 2.395587921142578}, "4": {"mean": -0.8605543971061707, "std": 2.1628165245056152}, "5": {"mean": -1.6488133668899536, "std": 1.4808751344680786}, "6": {"mean": -1.606386661529541, "std": 1.1835973262786865}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.13756844401359558, "std": 0.8309258222579956}, "1": {"mean": 0.3597412109375, "std": 0.9705332517623901}, "2": {"mean": 0.055811330676078796, "std": 0.5605169534683228}, "3": {"mean": 0.7475806474685669, "std": 1.1335870027542114}, "4": {"mean": -0.7399599552154541, "std": 0.6414052248001099}, "5": {"mean": -0.1399197280406952, "std": 0.7003283500671387}, "6": {"mean": 0.2528728246688843, "std": 0.45607155561447144}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.555793821811676, "std": 0.17185524106025696}, "1": {"mean": 0.38091281056404114, "std": 1.1753112077713013}, "2": {"mean": -0.6291626691818237, "std": 0.47127312421798706}, "3": {"mean": 0.54402756690979, "std": 0.9772289991378784}, "4": {"mean": 0.32106754183769226, "std": 0.5485780239105225}, "5": {"mean": 0.3765365183353424, "std": 0.5177939534187317}, "6": {"mean": -0.5036604404449463, "std": 0.3318285644054413}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.2615398168563843, "std": 0.6081777215003967}, "1": {"mean": 0.06360840797424316, "std": 1.0323528051376343}, "2": {"mean": 0.06606713682413101, "std": 0.8138105273246765}, "3": {"mean": 0.45020049810409546, "std": 0.8386121988296509}, "4": {"mean": 0.8483145236968994, "std": 0.6228349208831787}, "5": {"mean": 0.16367609798908234, "std": 0.47711196541786194}, "6": {"mean": -0.6389151811599731, "std": 0.42031723260879517}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.9368064999580383, "std": 0.32957807183265686}, "1": {"mean": -0.2740810811519623, "std": 1.3174175024032593}, "2": {"mean": 0.4671388268470764, "std": 0.8053041100502014}, "3": {"mean": -0.09701011329889297, "std": 1.4562519788742065}, "4": {"mean": -0.7140569686889648, "std": 0.774990975856781}, "5": {"mean": 0.1913321316242218, "std": 0.9027684330940247}, "6": {"mean": 0.6187005639076233, "std": 0.771294116973877}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": 0.24279113113880157, "std": 1.4123626947402954}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.327549, -1.083559, 0.380645, 0.18442, -0.91056], [0.089496, 0.497805, 0.564761, -0.005037, -0.016307], [-0.387039, 0.3142, 0.232852, 0.502417, -0.005017], [-0.437767, 0.276942, -0.370443, -0.641839, -0.696545], [0.115424, 0.030307, 0.372086, -0.960215, 0.102389], [-0.198311, 0.479039, -0.620284, -0.479769, 0.179563], [-0.230279, -0.002171, -0.096226, -0.257413, -0.392771]], "network.0.bias": [0.177993, -0.064067, -0.59667, 0.022977, 0.09618, -0.17186, -0.078431], "network.2.weight": [[0.850331, -0.356137, -0.053621, 0.590554, 0.352509, 0.231733, 0.253385], [-0.359012, 0.243978, 0.278778, -0.166438, -0.844235, -0.080629, -0.377375], [-0.077039, -0.159758, 0.332805, -0.182454, -0.272325, -0.151041, -0.327258], [-0.094511, 0.213801, 0.636691, 0.018813, -0.497936, -0.337174, -0.008319], [0.33482, 0.014921, 0.089224, 0.344422, -0.739899, -0.222934, -0.275323], [0.644009, -0.388801, 0.139979, -0.26698, 0.148427, -0.234605, -0.361689], [0.420103, -0.122903, -0.124313, 0.228086, 0.174821, 0.763329, -0.272892]], "network.2.bias": [0.52699, -0.005364, 0.131117, -0.234199, -0.626884, 0.264795, 0.419041], "network.4.weight": [[-0.155262, 0.043378, -0.072589, -0.17726, -0.370149, -0.024485, -0.018568], [-0.566723, 0.426357, 0.335675, 0.459227, -0.279242, -0.636927, -0.23372], [-0.352579, -0.023429, -0.361056, -0.343874, -0.324154, -0.322074, -0.055951], [0.925938, -0.159139, 0.23821, -0.286497, -0.446345, 0.757425, 0.196756], [0.439972, 0.392831, 0.081561, 0.255903, -0.052684, -0.341379, -0.406873], [0.115731, 0.11816, 0.006576, 0.346141, 0.46774, -0.314147, -0.387132], [0.015314, -0.045227, -0.459786, -0.211884, -0.149655, -0.183065, -0.312843]], "network.4.bias": [-0.330239, 0.07841, 0.004666, 0.249772, -0.116027, 0.146287, -0.035074], "network.6.weight": [[-0.156665, 0.273307, 0.352111, 0.081119, 0.422342, 0.403154, -0.032775], [0.287991, -0.539494, -0.173898, 0.619834, -0.063475, -0.297495, -0.141716], [0.04425, 0.277439, -0.195548, -0.299513, 0.494984, 0.333486, -0.262537], [-0.059028, 0.428109, 0.313416, -0.083315, 0.474458, 0.392011, -0.115764], [-0.224431, -0.219104, 0.162262, 0.672488, -0.124051, 0.313514, -0.030655], [0.221852, 0.180987, 0.321551, -0.109962, 0.299032, 0.235764, 0.507471], [0.118877, -0.37524, 0.01308, -0.012405, 0.090499, -0.303624, -0.455166]], "network.6.bias": [-0.300557, 0.14966, -0.238246, -0.118568, 0.454585, -0.091558, -0.287717], "network.8.weight": [[-0.150842, -0.004108, 0.089471, -0.147752, -0.108343, -0.564326, -0.264081], [-0.510134, 0.549692, -0.472999, -0.440898, 0.347982, -0.127408, 0.138025], [0.439903, 0.150382, 0.307352, 0.285713, -0.240525, 0.289841, -0.460809], [-0.201581, 0.838425, -0.429012, -0.602082, 0.50572, -0.062251, 0.282692], [0.131776, -0.264891, 0.367583, -0.111057, -0.733935, 0.112829, 0.112512], [-0.207102, 0.622262, 0.039772, -0.239525, 0.379169, -0.367617, 0.312141], [0.335528, 0.237609, 0.407358, 0.463525, -0.169748, -0.00757, -0.412237]], "network.8.bias": [-0.593881, -0.200657, 0.108122, -0.342561, -0.11874, -0.129998, 0.142864], "network.10.weight": [[0.226575, 0.308482, -0.454197, 0.502802, -0.304443, 0.626089, -0.460543]], "network.10.bias": [0.096274]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.682339608669281, "train_acc": 0.56, "val_loss": 0.7184015512466431, "val_acc": 0.48}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6837696433067322, "train_acc": 0.56, "val_loss": 0.6920127868652344, "val_acc": 0.48}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6568727791309357, "train_acc": 0.56, "val_loss": 0.6730180382728577, "val_acc": 0.48}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.6438421905040741, "train_acc": 0.53, "val_loss": 0.9443336725234985, "val_acc": 0.52}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.7032889425754547, "train_acc": 0.645, "val_loss": 0.6582567095756531, "val_acc": 0.58}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.6529999375343323, "train_acc": 0.58, "val_loss": 0.6346135139465332, "val_acc": 0.62}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.621658444404602, "train_acc": 0.625, "val_loss": 0.5809366106987, "val_acc": 0.68}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.5568651854991913, "train_acc": 0.7, "val_loss": 0.5617126226425171, "val_acc": 0.76}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.5424421429634094, "train_acc": 0.685, "val_loss": 0.5559018850326538, "val_acc": 0.74}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.5372087955474854, "train_acc": 0.71, "val_loss": 0.4885575771331787, "val_acc": 0.78}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.4877663403749466, "train_acc": 0.75, "val_loss": 0.493984192609787, "val_acc": 0.76}, {"stage": "improved", "epoch": 8, "global_epoch": 11, "train_loss": 0.4939248710870743, "train_acc": 0.735, "val_loss": 0.5070913434028625, "val_acc": 0.72}, {"stage": "improved", "epoch": 9, "global_epoch": 12, "train_loss": 0.4724911153316498, "train_acc": 0.745, "val_loss": 0.5035784244537354, "val_acc": 0.7}], "summary": {"total_epochs": 13, "degraded_epochs": 3, "improved_epochs": 10, "patterns": ["has_majority"], "degraded_stage": {"initial_val_loss": 0.7184015512466431, "final_val_loss": 0.6730180382728577, "initial_val_acc": 0.48, "final_val_acc": 0.48, "best_val_acc": 0.48}, "improved_stage": {"initial_val_loss": 0.9443336725234985, "final_val_loss": 0.5035784244537354, "initial_val_acc": 0.52, "final_val_acc": 0.7, "best_val_acc": 0.78, "best_epoch": 9}, "improvement": 0.30000000000000004, "first_improvement_epoch": 2}}
|
55
|
{"target_pattern": "sorted_descending", "degraded_accuracy": 0.74, "improved_accuracy": 0.94, "improvement": 0.19999999999999996, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 2026, "learning_rate": 0.03338577609673197, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_descending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_descending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.462254,
-0.47006,
-0.014951,
-0.48518,
-0.430504
],
[
-0.020587,
0.348213,
-0.671486,
0.353866,
-0.257769
],
[
0.234915,
-0.228687,
0.451584,
0.300487,
0.254216
],
[
0.129163,
0.4747,
-0.092261,
-0.134925,
0.031672
],
[
-0.365028,
0.111318,
0.42817,
0.19403,
0.542192
],
[
-0.31739,
0.126688,
-0.391787,
-0.237109,
-0.445633
],
[
0.726824,
0.047418,
0.144674,
-0.008557,
-0.26643
],
[
-0.033729,
0.401033,
0.002053,
0.083421,
0.249306
]
],
"network.0.bias": [
0.060676,
0.157121,
-0.234959,
-0.124789,
-0.034979,
-0.527187,
0.227787,
0.444835
],
"network.2.weight": [
[
-0.071417,
0.518032,
-0.046706,
0.401239,
-0.01786,
0.234886,
-0.285564,
0.156041
],
[
-0.317515,
-0.17621,
0.369757,
0.29465,
-0.254378,
0.102579,
0.516856,
0.207786
],
[
0.000438,
0.34152,
0.15348,
-0.173602,
0.545589,
0.021001,
-0.288874,
-0.044668
],
[
-0.308109,
0.025596,
-0.016742,
-0.203356,
-0.326627,
-0.027454,
-0.052189,
-0.124576
],
[
-0.076626,
0.010424,
0.375791,
-0.123423,
0.191939,
0.05514,
0.567424,
0.389159
],
[
-0.088543,
0.246174,
0.353697,
-0.116174,
0.508008,
0.094645,
-0.52461,
0.142899
],
[
0.247442,
0.354704,
0.114346,
-0.519661,
0.01265,
-0.064576,
0.177132,
-0.051725
],
[
-0.230195,
-0.181915,
-0.207882,
-0.086052,
0.289687,
0.15021,
0.434101,
0.42594
]
],
"network.2.bias": [
-0.211451,
0.218652,
0.146719,
-0.351126,
0.238156,
0.378868,
-0.1426,
0.072392
],
"network.4.weight": [
[
0.34067,
0.0372,
0.696037,
0.12519,
0.103173,
0.525336,
0.462976,
0.036815
],
[
-0.370389,
0.373603,
-0.293629,
-0.288586,
-0.002499,
-0.030302,
-0.305362,
0.254665
],
[
-0.494822,
0.555188,
-0.207136,
0.123847,
0.310666,
-0.351654,
-0.165996,
0.119225
],
[
-0.357116,
-0.26921,
0.004988,
-0.150578,
-0.271559,
0.023085,
0.159916,
-0.396517
],
[
-0.202233,
0.217002,
-0.069587,
-0.028912,
0.367976,
-0.299563,
-0.21309,
0.020521
],
[
0.438132,
-0.044742,
-0.030684,
0.053187,
0.139097,
0.526742,
0.022863,
0.496889
],
[
-0.319588,
0.496577,
-0.526142,
-0.344913,
0.144213,
-0.05356,
-0.262009,
0.583193
],
[
0.102792,
0.132441,
0.375398,
-0.1041,
0.469523,
-0.091725,
0.429815,
-0.283108
]
],
"network.4.bias": [
0.086733,
0.199558,
0.464804,
-0.417907,
0.350659,
0.099062,
0.206415,
-0.140765
],
"network.6.weight": [
[
0.001153,
0.238598,
0.255203,
-0.492912,
0.368678,
0.129532,
0.363936,
-0.425403
],
[
-0.175798,
0.389186,
0.330764,
-0.112786,
0.230469,
-0.33745,
0.558908,
-0.495301
],
[
0.603808,
-0.071022,
-0.304835,
-0.006666,
-0.287882,
0.132355,
-0.01716,
0.331845
],
[
0.5545,
-0.229232,
-0.465058,
0.005838,
-0.444199,
0.397159,
-0.07095,
0.058472
],
[
0.452731,
-0.180143,
-0.003108,
0.156833,
-0.123454,
0.518988,
0.071915,
0.356746
],
[
-0.228126,
0.352602,
0.105652,
-0.081566,
-0.18236,
-0.126499,
-0.240213,
-0.291854
],
[
-0.257855,
0.284168,
0.499263,
-0.563838,
0.222473,
-0.088445,
0.203931,
-0.397352
],
[
-0.140893,
0.057962,
0.178475,
0.19052,
-0.036745,
0.36679,
0.253099,
0.14236
]
],
"network.6.bias": [
-0.242355,
0.408477,
0.119757,
0.531378,
0.302933,
-0.200708,
0.384087,
-0.219557
],
"network.8.weight": [
[
0.08625,
0.38367,
-0.368916,
-0.734345,
-0.318585,
0.001178,
0.422326,
-0.337922
]
],
"network.8.bias": [
-0.235577
]
}
## Activation Signature
### 0
mean: [-2.973755, -0.195705, 1.541881, 0.457958, 1.692801, -2.567789, 1.173948, 1.624290]
std: [2.206944, 1.731400, 1.381870, 0.920595, 1.527256, 1.607245, 1.595345, 0.928350]
### 2
mean: [0.075448, 1.433798, 0.977752, -1.328553, 2.449485, 1.454946, 0.070165, 1.345433]
std: [0.760432, 1.297919, 1.084742, 0.621112, 1.400088, 1.389487, 0.378903, 0.779793]
### 4
mean: [2.213379, 0.538247, 1.227934, -2.036488, 0.946093, 1.970136, 1.254622, 1.193971]
std: [1.392471, 0.856076, 1.484269, 0.975232, 0.904075, 0.964329, 1.431512, 0.724449]
### 6
mean: [0.811310, 0.379090, 1.375942, 1.348112, 2.612111, -1.432272, 0.466158, 0.921376]
std: [1.229738, 1.847002, 1.457168, 1.939834, 1.322431, 0.627525, 1.533034, 0.774059]
### 8
mean: [-2.371964]
std: [2.637109]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_descending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.462254,
-0.47006,
-0.014951,
-0.48518,
-0.430504
],
[
-0.020587,
0.348213,
-0.671486,
0.353866,
-0.257769
],
[
0.234915,
-0.228687,
0.451584,
0.300487,
0.254216
],
[
0.129163,
0.4747,
-0.092261,
-0.134925,
0.031672
],
[
-0.365028,
0.111318,
0.42817,
0.19403,
0.542192
],
[
-0.31739,
0.126688,
-0.391787,
-0.237109,
-0.445633
],
[
0.726824,
0.047418,
0.144674,
-0.008557,
-0.26643
],
[
-0.033729,
0.401033,
0.002053,
0.083421,
0.249306
]
],
"network.0.bias": [
0.060676,
0.157121,
-0.234959,
-0.124789,
-0.034979,
-0.527187,
0.227787,
0.444835
],
"network.2.weight": [
[
-0.071417,
0.518032,
-0.046706,
0.401239,
-0.01786,
0.234886,
-0.285564,
0.156041
],
[
-0.317515,
-0.17621,
0.369757,
0.29465,
-0.254378,
0.102579,
0.516856,
0.207786
],
[
0.000438,
0.34152,
0.15348,
-0.173602,
0.545589,
0.021001,
-0.288874,
-0.044668
],
[
-0.308109,
0.025596,
-0.016742,
-0.203356,
-0.326627,
-0.027454,
-0.052189,
-0.124576
],
[
-0.076626,
0.010424,
0.375791,
-0.123423,
0.191939,
0.05514,
0.567424,
0.389159
],
[
-0.088543,
0.246174,
0.353697,
-0.116174,
0.508008,
0.094645,
-0.52461,
0.142899
],
[
0.247442,
0.354704,
0.114346,
-0.519661,
0.01265,
-0.064576,
0.177132,
-0.051725
],
[
-0.230195,
-0.181915,
-0.207882,
-0.086052,
0.289687,
0.15021,
0.434101,
0.42594
]
],
"network.2.bias": [
-0.211451,
0.218652,
0.146719,
-0.351126,
0.238156,
0.378868,
-0.1426,
0.072392
],
"network.4.weight": [
[
0.34067,
0.0372,
0.696037,
0.12519,
0.103173,
0.525336,
0.462976,
0.036815
],
[
-0.370389,
0.373603,
-0.293629,
-0.288586,
-0.002499,
-0.030302,
-0.305362,
0.254665
],
[
-0.494822,
0.555188,
-0.207136,
0.123847,
0.310666,
-0.351654,
-0.165996,
0.119225
],
[
-0.357116,
-0.26921,
0.004988,
-0.150578,
-0.271559,
0.023085,
0.159916,
-0.396517
],
[
-0.202233,
0.217002,
-0.069587,
-0.028912,
0.367976,
-0.299563,
-0.21309,
0.020521
],
[
0.438132,
-0.044742,
-0.030684,
0.053187,
0.139097,
0.526742,
0.022863,
0.496889
],
[
-0.319588,
0.496577,
-0.526142,
-0.344913,
0.144213,
-0.05356,
-0.262009,
0.583193
],
[
0.102792,
0.132441,
0.375398,
-0.1041,
0.469523,
-0.091725,
0.429815,
-0.283108
]
],
"network.4.bias": [
0.086733,
0.199558,
0.464804,
-0.417907,
0.350659,
0.099062,
0.206415,
-0.140765
],
"network.6.weight": [
[
0.001153,
0.238598,
0.255203,
-0.492912,
0.368678,
0.129532,
0.363936,
-0.425403
],
[
-0.175798,
0.389186,
0.330764,
-0.112786,
0.230469,
-0.33745,
0.558908,
-0.495301
],
[
0.603808,
-0.071022,
-0.304835,
-0.006666,
-0.287882,
0.132355,
-0.01716,
0.331845
],
[
0.5545,
-0.229232,
-0.465058,
0.005838,
-0.444199,
0.397159,
-0.07095,
0.058472
],
[
0.452731,
-0.180143,
-0.003108,
0.156833,
-0.123454,
0.518988,
0.071915,
0.356746
],
[
-0.228126,
0.352602,
0.105652,
-0.081566,
-0.18236,
-0.126499,
-0.240213,
-0.291854
],
[
-0.257855,
0.284168,
0.499263,
-0.563838,
0.222473,
-0.088445,
0.203931,
-0.397352
],
[
-0.140893,
0.057962,
0.178475,
0.19052,
-0.036745,
0.36679,
0.253099,
0.14236
]
],
"network.6.bias": [
-0.242355,
0.408477,
0.119757,
0.531378,
0.302933,
-0.200708,
0.384087,
-0.219557
],
"network.8.weight": [
[
0.08625,
0.38367,
-0.368916,
-0.734345,
-0.318585,
0.001178,
0.422326,
-0.337922
]
],
"network.8.bias": [
-0.235577
]
}
## Activation Signature
### 0
mean: [-2.973755, -0.195705, 1.541881, 0.457958, 1.692801, -2.567789, 1.173948, 1.624290]
std: [2.206944, 1.731400, 1.381870, 0.920595, 1.527256, 1.607245, 1.595345, 0.928350]
### 2
mean: [0.075448, 1.433798, 0.977752, -1.328553, 2.449485, 1.454946, 0.070165, 1.345433]
std: [0.760432, 1.297919, 1.084742, 0.621112, 1.400088, 1.389487, 0.378903, 0.779793]
### 4
mean: [2.213379, 0.538247, 1.227934, -2.036488, 0.946093, 1.970136, 1.254622, 1.193971]
std: [1.392471, 0.856076, 1.484269, 0.975232, 0.904075, 0.964329, 1.431512, 0.724449]
### 6
mean: [0.811310, 0.379090, 1.375942, 1.348112, 2.612111, -1.432272, 0.466158, 0.921376]
std: [1.229738, 1.847002, 1.457168, 1.939834, 1.322431, 0.627525, 1.533034, 0.774059]
### 8
mean: [-2.371964]
std: [2.637109]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_descending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -2.973755121231079, "std": 2.206943988800049}, "1": {"mean": -0.19570527970790863, "std": 1.7314003705978394}, "2": {"mean": 1.5418806076049805, "std": 1.381870150566101}, "3": {"mean": 0.4579584002494812, "std": 0.9205945134162903}, "4": {"mean": 1.692800760269165, "std": 1.527255654335022}, "5": {"mean": -2.567789077758789, "std": 1.6072447299957275}, "6": {"mean": 1.173947811126709, "std": 1.5953450202941895}, "7": {"mean": 1.6242897510528564, "std": 0.9283497929573059}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.07544776797294617, "std": 0.7604323029518127}, "1": {"mean": 1.4337981939315796, "std": 1.2979192733764648}, "2": {"mean": 0.9777524471282959, "std": 1.0847424268722534}, "3": {"mean": -1.328553318977356, "std": 0.6211123466491699}, "4": {"mean": 2.4494845867156982, "std": 1.4000880718231201}, "5": {"mean": 1.4549462795257568, "std": 1.3894866704940796}, "6": {"mean": 0.07016538828611374, "std": 0.37890297174453735}, "7": {"mean": 1.3454327583312988, "std": 0.7797927856445312}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 2.213379144668579, "std": 1.392471432685852}, "1": {"mean": 0.5382468700408936, "std": 0.8560757040977478}, "2": {"mean": 1.2279342412948608, "std": 1.4842692613601685}, "3": {"mean": -2.0364882946014404, "std": 0.9752323031425476}, "4": {"mean": 0.9460926651954651, "std": 0.9040750861167908}, "5": {"mean": 1.970136046409607, "std": 0.9643288254737854}, "6": {"mean": 1.2546215057373047, "std": 1.4315115213394165}, "7": {"mean": 1.193970799446106, "std": 0.7244489789009094}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.8113104701042175, "std": 1.229738473892212}, "1": {"mean": 0.37909016013145447, "std": 1.8470016717910767}, "2": {"mean": 1.3759419918060303, "std": 1.4571679830551147}, "3": {"mean": 1.3481115102767944, "std": 1.9398343563079834}, "4": {"mean": 2.6121110916137695, "std": 1.3224313259124756}, "5": {"mean": -1.432271957397461, "std": 0.6275252103805542}, "6": {"mean": 0.4661575257778168, "std": 1.533034086227417}, "7": {"mean": 0.9213758707046509, "std": 0.7740594148635864}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -2.3719635009765625, "std": 2.637108564376831}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.462254, -0.47006, -0.014951, -0.48518, -0.430504], [-0.020587, 0.348213, -0.671486, 0.353866, -0.257769], [0.234915, -0.228687, 0.451584, 0.300487, 0.254216], [0.129163, 0.4747, -0.092261, -0.134925, 0.031672], [-0.365028, 0.111318, 0.42817, 0.19403, 0.542192], [-0.31739, 0.126688, -0.391787, -0.237109, -0.445633], [0.726824, 0.047418, 0.144674, -0.008557, -0.26643], [-0.033729, 0.401033, 0.002053, 0.083421, 0.249306]], "network.0.bias": [0.060676, 0.157121, -0.234959, -0.124789, -0.034979, -0.527187, 0.227787, 0.444835], "network.2.weight": [[-0.071417, 0.518032, -0.046706, 0.401239, -0.01786, 0.234886, -0.285564, 0.156041], [-0.317515, -0.17621, 0.369757, 0.29465, -0.254378, 0.102579, 0.516856, 0.207786], [0.000438, 0.34152, 0.15348, -0.173602, 0.545589, 0.021001, -0.288874, -0.044668], [-0.308109, 0.025596, -0.016742, -0.203356, -0.326627, -0.027454, -0.052189, -0.124576], [-0.076626, 0.010424, 0.375791, -0.123423, 0.191939, 0.05514, 0.567424, 0.389159], [-0.088543, 0.246174, 0.353697, -0.116174, 0.508008, 0.094645, -0.52461, 0.142899], [0.247442, 0.354704, 0.114346, -0.519661, 0.01265, -0.064576, 0.177132, -0.051725], [-0.230195, -0.181915, -0.207882, -0.086052, 0.289687, 0.15021, 0.434101, 0.42594]], "network.2.bias": [-0.211451, 0.218652, 0.146719, -0.351126, 0.238156, 0.378868, -0.1426, 0.072392], "network.4.weight": [[0.34067, 0.0372, 0.696037, 0.12519, 0.103173, 0.525336, 0.462976, 0.036815], [-0.370389, 0.373603, -0.293629, -0.288586, -0.002499, -0.030302, -0.305362, 0.254665], [-0.494822, 0.555188, -0.207136, 0.123847, 0.310666, -0.351654, -0.165996, 0.119225], [-0.357116, -0.26921, 0.004988, -0.150578, -0.271559, 0.023085, 0.159916, -0.396517], [-0.202233, 0.217002, -0.069587, -0.028912, 0.367976, -0.299563, -0.21309, 0.020521], [0.438132, -0.044742, -0.030684, 0.053187, 0.139097, 0.526742, 0.022863, 0.496889], [-0.319588, 0.496577, -0.526142, -0.344913, 0.144213, -0.05356, -0.262009, 0.583193], [0.102792, 0.132441, 0.375398, -0.1041, 0.469523, -0.091725, 0.429815, -0.283108]], "network.4.bias": [0.086733, 0.199558, 0.464804, -0.417907, 0.350659, 0.099062, 0.206415, -0.140765], "network.6.weight": [[0.001153, 0.238598, 0.255203, -0.492912, 0.368678, 0.129532, 0.363936, -0.425403], [-0.175798, 0.389186, 0.330764, -0.112786, 0.230469, -0.33745, 0.558908, -0.495301], [0.603808, -0.071022, -0.304835, -0.006666, -0.287882, 0.132355, -0.01716, 0.331845], [0.5545, -0.229232, -0.465058, 0.005838, -0.444199, 0.397159, -0.07095, 0.058472], [0.452731, -0.180143, -0.003108, 0.156833, -0.123454, 0.518988, 0.071915, 0.356746], [-0.228126, 0.352602, 0.105652, -0.081566, -0.18236, -0.126499, -0.240213, -0.291854], [-0.257855, 0.284168, 0.499263, -0.563838, 0.222473, -0.088445, 0.203931, -0.397352], [-0.140893, 0.057962, 0.178475, 0.19052, -0.036745, 0.36679, 0.253099, 0.14236]], "network.6.bias": [-0.242355, 0.408477, 0.119757, 0.531378, 0.302933, -0.200708, 0.384087, -0.219557], "network.8.weight": [[0.08625, 0.38367, -0.368916, -0.734345, -0.318585, 0.001178, 0.422326, -0.337922]], "network.8.bias": [-0.235577]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6975021660327911, "train_acc": 0.465, "val_loss": 0.6671959161758423, "val_acc": 0.72}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6571891009807587, "train_acc": 0.655, "val_loss": 0.6217471957206726, "val_acc": 0.74}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.6036905646324158, "train_acc": 0.735, "val_loss": 0.48350411653518677, "val_acc": 0.88}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.44696804881095886, "train_acc": 0.87, "val_loss": 0.3810785710811615, "val_acc": 0.82}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.2962309867143631, "train_acc": 0.915, "val_loss": 0.33219677209854126, "val_acc": 0.9}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.32484276592731476, "train_acc": 0.93, "val_loss": 0.3583541214466095, "val_acc": 0.9}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.28492702543735504, "train_acc": 0.93, "val_loss": 0.2814509868621826, "val_acc": 0.9}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.24412710219621658, "train_acc": 0.93, "val_loss": 0.23809976875782013, "val_acc": 0.94}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.27845434844493866, "train_acc": 0.915, "val_loss": 0.22403313219547272, "val_acc": 0.92}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.2414628565311432, "train_acc": 0.93, "val_loss": 0.2215646356344223, "val_acc": 0.94}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.22060757130384445, "train_acc": 0.945, "val_loss": 0.23432494699954987, "val_acc": 0.94}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.21491804718971252, "train_acc": 0.94, "val_loss": 0.213835671544075, "val_acc": 0.92}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["sorted_descending"], "degraded_stage": {"initial_val_loss": 0.6671959161758423, "final_val_loss": 0.6217471957206726, "initial_val_acc": 0.72, "final_val_acc": 0.74, "best_val_acc": 0.74}, "improved_stage": {"initial_val_loss": 0.48350411653518677, "final_val_loss": 0.213835671544075, "initial_val_acc": 0.88, "final_val_acc": 0.92, "best_val_acc": 0.94, "best_epoch": 7}, "improvement": 0.19999999999999996, "first_improvement_epoch": 1}}
|
56
|
{"target_pattern": "has_majority", "degraded_accuracy": 0.54, "improved_accuracy": 0.72, "improvement": 0.17999999999999994, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 5859, "learning_rate": 0.048405008971870465, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "has_majority", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["has_majority"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.873055,
0.2277,
0.006831,
0.683613,
0.127053
],
[
-0.16345,
0.402889,
0.186683,
0.836778,
-0.086725
],
[
-1.052957,
-0.032885,
0.12437,
0.408571,
0.005701
],
[
-0.305762,
-0.370699,
-0.153204,
-0.393847,
0.045501
],
[
-0.217134,
0.22188,
0.446161,
0.842933,
-0.004783
],
[
-0.936061,
-0.28378,
-0.036383,
0.020158,
0.219462
]
],
"network.0.bias": [
-0.616137,
0.021327,
0.097528,
0.212677,
-0.112363,
-0.088916
],
"network.2.weight": [
[
-0.036156,
-0.249774,
0.312787,
-0.050697,
-0.109959,
-0.079507
],
[
-0.58085,
-0.259657,
0.267755,
-0.129082,
0.169693,
0.27446
],
[
0.355153,
0.482691,
0.288579,
-0.024552,
0.581852,
0.289234
],
[
-0.102095,
-0.003659,
-0.352073,
0.39378,
-0.562468,
0.162847
],
[
-0.470459,
-0.46656,
-0.228371,
0.08577,
0.134685,
0.130265
],
[
0.542746,
0.364429,
-0.03225,
-0.232039,
0.394074,
0.330195
]
],
"network.2.bias": [
-0.180422,
-0.268744,
0.160204,
-0.601689,
-0.550912,
-0.046158
],
"network.4.weight": [
[
-0.186463,
-0.238222,
0.51911,
-0.025518,
0.085426,
0.228173
],
[
0.171915,
0.777169,
-0.179746,
-0.120801,
0.211964,
-0.106255
],
[
-0.07135,
-0.218959,
-0.069309,
-0.298526,
-0.009147,
-0.308131
],
[
0.326663,
0.153147,
-0.234346,
-0.397793,
-0.386925,
-0.400129
],
[
0.243423,
-0.482595,
0.594374,
0.285554,
-0.085902,
0.439815
],
[
-0.392705,
-0.392868,
-0.349516,
-0.253497,
-0.121729,
0.038845
]
],
"network.4.bias": [
0.070777,
0.441525,
-0.020704,
0.230382,
-0.197573,
-0.235062
],
"network.6.weight": [
[
-0.178819,
-0.158792,
-0.082261,
-0.193444,
-0.21084,
-0.397684
],
[
-0.008304,
-0.390475,
-0.350187,
0.35692,
-0.156173,
-0.059601
],
[
0.091,
-0.383484,
0.199269,
0.238265,
0.469047,
0.002542
],
[
0.49888,
-0.540233,
0.159517,
0.058835,
0.414729,
0.158765
],
[
0.152961,
0.172936,
-0.043045,
-0.215173,
-0.552308,
-0.176479
],
[
-0.065998,
-0.24332,
0.161672,
-0.214601,
-0.350792,
0.094832
]
],
"network.6.bias": [
-0.184246,
-0.35436,
-0.423295,
-0.176991,
-0.351426,
-0.078331
],
"network.8.weight": [
[
0.161832,
-0.309396,
-0.296497,
-0.446938,
0.050138,
-0.21733
],
[
0.131862,
-0.389263,
-0.333733,
-0.414008,
-0.241299,
-0.240381
],
[
-0.142757,
-0.281696,
0.355424,
0.194096,
0.133108,
0.392347
],
[
-0.369154,
-0.384708,
0.29927,
0.480367,
0.197338,
0.326589
],
[
-0.031285,
-0.355476,
0.174128,
0.119126,
0.49388,
0.295381
],
[
0.170511,
-0.311937,
0.388348,
0.337755,
-0.018124,
-0.07282
]
],
"network.8.bias": [
-0.162755,
-0.117822,
-0.296121,
-0.428857,
0.796202,
-0.306676
],
"network.10.weight": [
[
-0.219225,
0.042506,
0.441958,
0.084446,
-0.42971,
-0.63953
],
[
-0.378811,
0.045716,
-0.377449,
0.065202,
-0.212774,
0.089002
],
[
0.40654,
-0.290139,
0.056327,
-0.095519,
-0.427049,
0.189815
],
[
0.29977,
0.044322,
0.801501,
0.611847,
-0.134929,
0.375322
],
[
0.133505,
-0.050421,
0.146717,
-0.132659,
-0.587076,
-0.148046
],
[
-0.162952,
0.212508,
-0.19398,
-0.355738,
-0.395565,
8e-05
]
],
"network.10.bias": [
-0.409824,
-0.318609,
-0.287025,
-0.087703,
0.034179,
-0.161222
],
"network.12.weight": [
[
-0.311426,
-0.268567,
0.040462,
-0.297144,
-0.049424,
0.096265
]
],
"network.12.bias": [
0.546403
]
}
## Activation Signature
### 0
mean: [0.349251, 2.629498, -0.129757, -1.951815, 2.755446, -1.533005]
std: [2.343874, 2.143537, 2.291334, 1.546924, 2.141741, 2.126932]
### 2
mean: [-0.943267, -0.898157, 3.701360, -2.554107, -2.104748, 2.607172]
std: [0.618040, 0.787871, 2.863079, 1.561476, 1.490687, 2.247766]
### 4
mean: [2.589130, -0.500339, -1.084517, -1.684390, 3.153003, -1.427875]
std: [1.995482, 0.752555, 0.886350, 1.564328, 2.683592, 0.913771]
### 6
mean: [-1.334699, -0.894924, 1.269377, 2.380306, -1.694353, -1.387936]
std: [0.896743, 0.406954, 1.464860, 2.156877, 1.179098, 1.036092]
### 8
mean: [-1.656252, -1.582520, 0.664142, 1.149666, 1.325043, 1.047869]
std: [1.336263, 1.317299, 0.885421, 1.410107, 0.484161, 1.230904]
### 10
mean: [-1.263579, -0.698043, -0.718138, 1.511301, -0.967159, -1.273550]
std: [0.479466, 0.220811, 0.067869, 1.818094, 0.509447, 0.812582]
### 12
mean: [0.080981]
std: [0.525309]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
has_majority
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.873055,
0.2277,
0.006831,
0.683613,
0.127053
],
[
-0.16345,
0.402889,
0.186683,
0.836778,
-0.086725
],
[
-1.052957,
-0.032885,
0.12437,
0.408571,
0.005701
],
[
-0.305762,
-0.370699,
-0.153204,
-0.393847,
0.045501
],
[
-0.217134,
0.22188,
0.446161,
0.842933,
-0.004783
],
[
-0.936061,
-0.28378,
-0.036383,
0.020158,
0.219462
]
],
"network.0.bias": [
-0.616137,
0.021327,
0.097528,
0.212677,
-0.112363,
-0.088916
],
"network.2.weight": [
[
-0.036156,
-0.249774,
0.312787,
-0.050697,
-0.109959,
-0.079507
],
[
-0.58085,
-0.259657,
0.267755,
-0.129082,
0.169693,
0.27446
],
[
0.355153,
0.482691,
0.288579,
-0.024552,
0.581852,
0.289234
],
[
-0.102095,
-0.003659,
-0.352073,
0.39378,
-0.562468,
0.162847
],
[
-0.470459,
-0.46656,
-0.228371,
0.08577,
0.134685,
0.130265
],
[
0.542746,
0.364429,
-0.03225,
-0.232039,
0.394074,
0.330195
]
],
"network.2.bias": [
-0.180422,
-0.268744,
0.160204,
-0.601689,
-0.550912,
-0.046158
],
"network.4.weight": [
[
-0.186463,
-0.238222,
0.51911,
-0.025518,
0.085426,
0.228173
],
[
0.171915,
0.777169,
-0.179746,
-0.120801,
0.211964,
-0.106255
],
[
-0.07135,
-0.218959,
-0.069309,
-0.298526,
-0.009147,
-0.308131
],
[
0.326663,
0.153147,
-0.234346,
-0.397793,
-0.386925,
-0.400129
],
[
0.243423,
-0.482595,
0.594374,
0.285554,
-0.085902,
0.439815
],
[
-0.392705,
-0.392868,
-0.349516,
-0.253497,
-0.121729,
0.038845
]
],
"network.4.bias": [
0.070777,
0.441525,
-0.020704,
0.230382,
-0.197573,
-0.235062
],
"network.6.weight": [
[
-0.178819,
-0.158792,
-0.082261,
-0.193444,
-0.21084,
-0.397684
],
[
-0.008304,
-0.390475,
-0.350187,
0.35692,
-0.156173,
-0.059601
],
[
0.091,
-0.383484,
0.199269,
0.238265,
0.469047,
0.002542
],
[
0.49888,
-0.540233,
0.159517,
0.058835,
0.414729,
0.158765
],
[
0.152961,
0.172936,
-0.043045,
-0.215173,
-0.552308,
-0.176479
],
[
-0.065998,
-0.24332,
0.161672,
-0.214601,
-0.350792,
0.094832
]
],
"network.6.bias": [
-0.184246,
-0.35436,
-0.423295,
-0.176991,
-0.351426,
-0.078331
],
"network.8.weight": [
[
0.161832,
-0.309396,
-0.296497,
-0.446938,
0.050138,
-0.21733
],
[
0.131862,
-0.389263,
-0.333733,
-0.414008,
-0.241299,
-0.240381
],
[
-0.142757,
-0.281696,
0.355424,
0.194096,
0.133108,
0.392347
],
[
-0.369154,
-0.384708,
0.29927,
0.480367,
0.197338,
0.326589
],
[
-0.031285,
-0.355476,
0.174128,
0.119126,
0.49388,
0.295381
],
[
0.170511,
-0.311937,
0.388348,
0.337755,
-0.018124,
-0.07282
]
],
"network.8.bias": [
-0.162755,
-0.117822,
-0.296121,
-0.428857,
0.796202,
-0.306676
],
"network.10.weight": [
[
-0.219225,
0.042506,
0.441958,
0.084446,
-0.42971,
-0.63953
],
[
-0.378811,
0.045716,
-0.377449,
0.065202,
-0.212774,
0.089002
],
[
0.40654,
-0.290139,
0.056327,
-0.095519,
-0.427049,
0.189815
],
[
0.29977,
0.044322,
0.801501,
0.611847,
-0.134929,
0.375322
],
[
0.133505,
-0.050421,
0.146717,
-0.132659,
-0.587076,
-0.148046
],
[
-0.162952,
0.212508,
-0.19398,
-0.355738,
-0.395565,
8e-05
]
],
"network.10.bias": [
-0.409824,
-0.318609,
-0.287025,
-0.087703,
0.034179,
-0.161222
],
"network.12.weight": [
[
-0.311426,
-0.268567,
0.040462,
-0.297144,
-0.049424,
0.096265
]
],
"network.12.bias": [
0.546403
]
}
## Activation Signature
### 0
mean: [0.349251, 2.629498, -0.129757, -1.951815, 2.755446, -1.533005]
std: [2.343874, 2.143537, 2.291334, 1.546924, 2.141741, 2.126932]
### 2
mean: [-0.943267, -0.898157, 3.701360, -2.554107, -2.104748, 2.607172]
std: [0.618040, 0.787871, 2.863079, 1.561476, 1.490687, 2.247766]
### 4
mean: [2.589130, -0.500339, -1.084517, -1.684390, 3.153003, -1.427875]
std: [1.995482, 0.752555, 0.886350, 1.564328, 2.683592, 0.913771]
### 6
mean: [-1.334699, -0.894924, 1.269377, 2.380306, -1.694353, -1.387936]
std: [0.896743, 0.406954, 1.464860, 2.156877, 1.179098, 1.036092]
### 8
mean: [-1.656252, -1.582520, 0.664142, 1.149666, 1.325043, 1.047869]
std: [1.336263, 1.317299, 0.885421, 1.410107, 0.484161, 1.230904]
### 10
mean: [-1.263579, -0.698043, -0.718138, 1.511301, -0.967159, -1.273550]
std: [0.479466, 0.220811, 0.067869, 1.818094, 0.509447, 0.812582]
### 12
mean: [0.080981]
std: [0.525309]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
has_majority
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.3492512106895447, "std": 2.343874216079712}, "1": {"mean": 2.629497766494751, "std": 2.1435365676879883}, "2": {"mean": -0.12975674867630005, "std": 2.2913336753845215}, "3": {"mean": -1.9518154859542847, "std": 1.5469236373901367}, "4": {"mean": 2.755445957183838, "std": 2.1417412757873535}, "5": {"mean": -1.5330051183700562, "std": 2.126932382583618}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.9432669878005981, "std": 0.6180398464202881}, "1": {"mean": -0.898157000541687, "std": 0.787871241569519}, "2": {"mean": 3.701359748840332, "std": 2.863079309463501}, "3": {"mean": -2.554107427597046, "std": 1.5614757537841797}, "4": {"mean": -2.1047475337982178, "std": 1.4906865358352661}, "5": {"mean": 2.6071715354919434, "std": 2.2477662563323975}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 2.589130401611328, "std": 1.9954817295074463}, "1": {"mean": -0.5003387331962585, "std": 0.7525553703308105}, "2": {"mean": -1.0845167636871338, "std": 0.886349618434906}, "3": {"mean": -1.6843899488449097, "std": 1.5643284320831299}, "4": {"mean": 3.153002977371216, "std": 2.683591842651367}, "5": {"mean": -1.4278748035430908, "std": 0.9137706160545349}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -1.3346987962722778, "std": 0.89674311876297}, "1": {"mean": -0.8949238061904907, "std": 0.4069535434246063}, "2": {"mean": 1.2693768739700317, "std": 1.4648597240447998}, "3": {"mean": 2.380305528640747, "std": 2.156876564025879}, "4": {"mean": -1.6943532228469849, "std": 1.1790975332260132}, "5": {"mean": -1.3879364728927612, "std": 1.0360920429229736}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.6562515497207642, "std": 1.3362630605697632}, "1": {"mean": -1.5825200080871582, "std": 1.317299246788025}, "2": {"mean": 0.6641421914100647, "std": 0.8854206800460815}, "3": {"mean": 1.1496657133102417, "std": 1.4101065397262573}, "4": {"mean": 1.3250426054000854, "std": 0.4841613471508026}, "5": {"mean": 1.0478694438934326, "std": 1.2309036254882812}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -1.26357901096344, "std": 0.4794659912586212}, "1": {"mean": -0.6980425715446472, "std": 0.22081081569194794}, "2": {"mean": -0.7181376218795776, "std": 0.06786859780550003}, "3": {"mean": 1.511300802230835, "std": 1.8180937767028809}, "4": {"mean": -0.9671589136123657, "std": 0.509447455406189}, "5": {"mean": -1.2735497951507568, "std": 0.8125824332237244}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": 0.08098069578409195, "std": 0.5253087282180786}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.873055, 0.2277, 0.006831, 0.683613, 0.127053], [-0.16345, 0.402889, 0.186683, 0.836778, -0.086725], [-1.052957, -0.032885, 0.12437, 0.408571, 0.005701], [-0.305762, -0.370699, -0.153204, -0.393847, 0.045501], [-0.217134, 0.22188, 0.446161, 0.842933, -0.004783], [-0.936061, -0.28378, -0.036383, 0.020158, 0.219462]], "network.0.bias": [-0.616137, 0.021327, 0.097528, 0.212677, -0.112363, -0.088916], "network.2.weight": [[-0.036156, -0.249774, 0.312787, -0.050697, -0.109959, -0.079507], [-0.58085, -0.259657, 0.267755, -0.129082, 0.169693, 0.27446], [0.355153, 0.482691, 0.288579, -0.024552, 0.581852, 0.289234], [-0.102095, -0.003659, -0.352073, 0.39378, -0.562468, 0.162847], [-0.470459, -0.46656, -0.228371, 0.08577, 0.134685, 0.130265], [0.542746, 0.364429, -0.03225, -0.232039, 0.394074, 0.330195]], "network.2.bias": [-0.180422, -0.268744, 0.160204, -0.601689, -0.550912, -0.046158], "network.4.weight": [[-0.186463, -0.238222, 0.51911, -0.025518, 0.085426, 0.228173], [0.171915, 0.777169, -0.179746, -0.120801, 0.211964, -0.106255], [-0.07135, -0.218959, -0.069309, -0.298526, -0.009147, -0.308131], [0.326663, 0.153147, -0.234346, -0.397793, -0.386925, -0.400129], [0.243423, -0.482595, 0.594374, 0.285554, -0.085902, 0.439815], [-0.392705, -0.392868, -0.349516, -0.253497, -0.121729, 0.038845]], "network.4.bias": [0.070777, 0.441525, -0.020704, 0.230382, -0.197573, -0.235062], "network.6.weight": [[-0.178819, -0.158792, -0.082261, -0.193444, -0.21084, -0.397684], [-0.008304, -0.390475, -0.350187, 0.35692, -0.156173, -0.059601], [0.091, -0.383484, 0.199269, 0.238265, 0.469047, 0.002542], [0.49888, -0.540233, 0.159517, 0.058835, 0.414729, 0.158765], [0.152961, 0.172936, -0.043045, -0.215173, -0.552308, -0.176479], [-0.065998, -0.24332, 0.161672, -0.214601, -0.350792, 0.094832]], "network.6.bias": [-0.184246, -0.35436, -0.423295, -0.176991, -0.351426, -0.078331], "network.8.weight": [[0.161832, -0.309396, -0.296497, -0.446938, 0.050138, -0.21733], [0.131862, -0.389263, -0.333733, -0.414008, -0.241299, -0.240381], [-0.142757, -0.281696, 0.355424, 0.194096, 0.133108, 0.392347], [-0.369154, -0.384708, 0.29927, 0.480367, 0.197338, 0.326589], [-0.031285, -0.355476, 0.174128, 0.119126, 0.49388, 0.295381], [0.170511, -0.311937, 0.388348, 0.337755, -0.018124, -0.07282]], "network.8.bias": [-0.162755, -0.117822, -0.296121, -0.428857, 0.796202, -0.306676], "network.10.weight": [[-0.219225, 0.042506, 0.441958, 0.084446, -0.42971, -0.63953], [-0.378811, 0.045716, -0.377449, 0.065202, -0.212774, 0.089002], [0.40654, -0.290139, 0.056327, -0.095519, -0.427049, 0.189815], [0.29977, 0.044322, 0.801501, 0.611847, -0.134929, 0.375322], [0.133505, -0.050421, 0.146717, -0.132659, -0.587076, -0.148046], [-0.162952, 0.212508, -0.19398, -0.355738, -0.395565, 8e-05]], "network.10.bias": [-0.409824, -0.318609, -0.287025, -0.087703, 0.034179, -0.161222], "network.12.weight": [[-0.311426, -0.268567, 0.040462, -0.297144, -0.049424, 0.096265]], "network.12.bias": [0.546403]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7145660817623138, "train_acc": 0.44, "val_loss": 0.6976159811019897, "val_acc": 0.46}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6966623067855835, "train_acc": 0.46, "val_loss": 0.6863939166069031, "val_acc": 0.54}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6846117973327637, "train_acc": 0.56, "val_loss": 0.676825761795044, "val_acc": 0.54}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6686347723007202, "train_acc": 0.56, "val_loss": 0.6745055317878723, "val_acc": 0.54}, {"stage": "degraded", "epoch": 4, "global_epoch": 4, "train_loss": 0.6993837356567383, "train_acc": 0.56, "val_loss": 0.6575915813446045, "val_acc": 0.54}, {"stage": "improved", "epoch": 0, "global_epoch": 5, "train_loss": 0.6879318356513977, "train_acc": 0.49, "val_loss": 0.6781542301177979, "val_acc": 0.54}, {"stage": "improved", "epoch": 1, "global_epoch": 6, "train_loss": 0.6799401342868805, "train_acc": 0.545, "val_loss": 0.6513535976409912, "val_acc": 0.7}, {"stage": "improved", "epoch": 2, "global_epoch": 7, "train_loss": 0.6393404304981232, "train_acc": 0.615, "val_loss": 0.6472363471984863, "val_acc": 0.56}, {"stage": "improved", "epoch": 3, "global_epoch": 8, "train_loss": 0.653215229511261, "train_acc": 0.59, "val_loss": 0.6167956590652466, "val_acc": 0.7}, {"stage": "improved", "epoch": 4, "global_epoch": 9, "train_loss": 0.6337046921253204, "train_acc": 0.675, "val_loss": 0.6408571004867554, "val_acc": 0.66}, {"stage": "improved", "epoch": 5, "global_epoch": 10, "train_loss": 0.6401921212673187, "train_acc": 0.67, "val_loss": 0.6193802952766418, "val_acc": 0.66}, {"stage": "improved", "epoch": 6, "global_epoch": 11, "train_loss": 0.6179244816303253, "train_acc": 0.665, "val_loss": 0.6050733923912048, "val_acc": 0.7}, {"stage": "improved", "epoch": 7, "global_epoch": 12, "train_loss": 0.626367449760437, "train_acc": 0.685, "val_loss": 0.5977752804756165, "val_acc": 0.72}, {"stage": "improved", "epoch": 8, "global_epoch": 13, "train_loss": 0.6159002780914307, "train_acc": 0.69, "val_loss": 0.6092081665992737, "val_acc": 0.66}, {"stage": "improved", "epoch": 9, "global_epoch": 14, "train_loss": 0.6135796904563904, "train_acc": 0.655, "val_loss": 0.6080566048622131, "val_acc": 0.66}], "summary": {"total_epochs": 15, "degraded_epochs": 5, "improved_epochs": 10, "patterns": ["has_majority"], "degraded_stage": {"initial_val_loss": 0.6976159811019897, "final_val_loss": 0.6575915813446045, "initial_val_acc": 0.46, "final_val_acc": 0.54, "best_val_acc": 0.54}, "improved_stage": {"initial_val_loss": 0.6781542301177979, "final_val_loss": 0.6080566048622131, "initial_val_acc": 0.54, "final_val_acc": 0.66, "best_val_acc": 0.72, "best_epoch": 12}, "improvement": 0.17999999999999994, "first_improvement_epoch": 4}}
|
57
|
{"target_pattern": "has_majority", "degraded_accuracy": 0.54, "improved_accuracy": 0.82, "improvement": 0.2799999999999999, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 8445, "learning_rate": 0.09004911743914298, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "has_majority", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["has_majority"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.28555,
-0.585842,
-0.75365,
0.223227,
-0.318473
],
[
0.390674,
-0.547687,
-0.289123,
0.483498,
-0.494426
],
[
0.063819,
0.8327,
-0.013864,
0.040511,
0.2056
],
[
-0.125027,
0.864264,
0.673288,
0.105641,
-0.345485
],
[
-0.274279,
0.563073,
0.007496,
-0.860526,
0.190974
],
[
-0.11775,
0.958508,
0.663092,
0.066207,
0.497201
],
[
-0.022728,
0.587926,
0.072356,
0.552019,
0.141062
],
[
0.389843,
-0.175592,
0.188674,
0.31038,
-0.000399
]
],
"network.0.bias": [
-0.033013,
0.330157,
0.789219,
0.659598,
0.292383,
0.447408,
-0.393794,
-0.015652
],
"network.2.weight": [
[
0.155608,
-0.058904,
-0.821688,
-1.022754,
0.257736,
-0.270561,
0.125554,
-0.086035
],
[
-0.11311,
-0.098404,
-0.69032,
-0.139635,
-0.273691,
-0.319631,
-0.457735,
-0.326293
],
[
-0.432644,
0.207913,
-0.306255,
0.21515,
0.088273,
0.365568,
-0.552945,
0.158325
],
[
0.507438,
0.171937,
-0.612509,
-0.101586,
0.213935,
-0.549051,
-0.304922,
0.228154
],
[
-0.405909,
0.062907,
-0.166579,
-0.165713,
-0.715047,
0.016667,
-0.088264,
0.155718
],
[
-0.254419,
-0.4901,
-0.452683,
-0.396663,
-0.211995,
-0.227873,
-0.641117,
-0.912921
],
[
-0.323057,
-0.132185,
-0.476141,
0.365602,
-0.596992,
0.396842,
0.436044,
0.401694
],
[
0.352967,
0.516298,
-0.532285,
-0.311629,
0.027794,
-0.621049,
-0.058326,
0.459311
]
],
"network.2.bias": [
0.055204,
-0.153831,
-0.243659,
0.128319,
-0.398743,
-0.43483,
-0.147643,
0.299864
],
"network.4.weight": [
[
0.416816,
0.038007,
0.116814,
0.694459,
-0.377417,
0.126791,
-0.333282,
0.682632
],
[
-0.032255,
0.34944,
0.319036,
-0.263415,
0.066047,
0.350672,
0.260194,
-0.581387
],
[
0.030874,
0.701526,
0.310916,
0.607645,
0.030159,
-0.404569,
-0.338526,
0.178454
],
[
-0.17792,
0.378919,
0.718673,
0.113415,
0.037629,
-0.422878,
0.023785,
0.354603
],
[
0.080875,
0.170766,
-0.69538,
-0.004985,
0.128942,
0.3107,
0.398145,
-0.598106
],
[
0.221196,
-0.014504,
-0.82754,
-0.216401,
-0.104368,
0.380573,
0.541412,
-0.677297
],
[
-0.806997,
-0.363915,
-0.21169,
0.124288,
-0.015158,
0.315729,
-0.578132,
-0.230187
],
[
0.26194,
0.120044,
-0.244206,
0.201601,
-0.022065,
-0.047956,
0.387993,
-0.214002
]
],
"network.4.bias": [
0.281042,
0.169975,
-0.24208,
0.279367,
-0.000279,
0.134996,
-0.700738,
-0.057568
],
"network.6.weight": [
[
-0.111728,
0.360191,
-0.036835,
-0.108012,
0.256062,
0.56136,
-0.093863,
0.192866
],
[
-0.151732,
0.085939,
0.273134,
-0.297613,
-0.116976,
0.089507,
0.448955,
-0.326165
],
[
-0.341048,
0.074102,
-0.351854,
0.015842,
0.269465,
0.394833,
-0.134617,
0.170126
],
[
0.270858,
0.226148,
0.3568,
0.54347,
-0.043089,
-0.148902,
-0.564998,
-0.427032
],
[
0.706572,
-0.085923,
0.022287,
0.28483,
-0.291831,
-0.368317,
-0.586737,
-0.091869
],
[
0.36832,
0.209231,
0.137012,
0.32513,
-0.222233,
-0.535954,
-0.478629,
-0.161856
],
[
-0.40172,
-0.051807,
0.050888,
-0.204599,
0.305241,
0.397133,
0.598573,
-0.009383
],
[
0.109591,
0.400216,
0.494332,
0.032112,
0.193637,
0.100241,
0.397715,
0.315627
]
],
"network.6.bias": [
-0.128859,
-0.148539,
0.05003,
-0.319375,
0.131063,
0.295449,
0.065541,
0.1174
],
"network.8.weight": [
[
-0.276623,
0.283873,
0.351986,
0.058593,
-0.329713,
-0.249562,
-0.154653,
0.317008
],
[
-0.267166,
0.31809,
-0.034193,
-0.36067,
-0.358363,
-0.434219,
0.483574,
-0.34291
],
[
-0.142642,
-0.038954,
-0.669077,
0.124354,
0.504207,
0.211518,
0.224927,
0.209702
],
[
-0.390305,
-0.363273,
-0.301053,
0.029567,
0.728059,
0.096461,
-0.193203,
0.388256
],
[
0.276055,
-0.013297,
0.449237,
-0.201499,
0.07606,
-0.482161,
0.394831,
0.299453
],
[
0.301715,
0.200831,
0.159707,
-0.048514,
-0.572869,
-0.234166,
0.288475,
-0.202855
],
[
-0.376849,
-0.144325,
0.02382,
0.285778,
0.275034,
0.1251,
-0.410248,
0.456349
],
[
0.196569,
-0.345602,
-0.031006,
-0.14999,
-0.108604,
-0.272771,
0.234914,
-0.190938
]
],
"network.8.bias": [
0.12415,
-0.166943,
0.524987,
0.216289,
0.242227,
-0.135871,
-0.045749,
-0.044547
],
"network.10.weight": [
[
0.001183,
-0.03051,
0.246116,
0.19549,
-0.364823,
-0.222811,
0.274562,
-0.243259
]
],
"network.10.bias": [
0.234619
]
}
## Activation Signature
### 0
mean: [-2.233574, -0.358582, 2.697491, 3.285758, -0.617053, 4.187808, 2.158148, 1.208739]
std: [1.936510, 1.475683, 1.659237, 2.264182, 1.668032, 2.621907, 1.887892, 1.031311]
### 2
mean: [-6.347646, -5.217939, 0.264967, -4.385394, -1.487974, -6.502300, 2.620796, -4.122975]
std: [3.985045, 3.144516, 0.858405, 3.011122, 0.903033, 3.694987, 2.074425, 3.161238]
### 4
mean: [-0.558433, 0.961953, -1.053342, 0.584756, 0.781021, 1.277230, -2.263901, 0.863028]
std: [0.641456, 0.612475, 0.657902, 0.486574, 0.868987, 1.140676, 1.252345, 0.796443]
### 6
mean: [1.154358, -0.494864, 0.998458, -0.448492, -0.553965, -0.322106, 0.604518, 0.885293]
std: [1.188132, 0.201927, 0.843385, 0.574572, 0.862844, 0.930486, 0.760593, 0.764338]
### 8
mean: [0.232425, -0.562215, 0.079860, -0.223540, 1.362658, 0.287911, -0.285981, 0.149719]
std: [0.176557, 0.343386, 0.541066, 0.666955, 1.312506, 0.656444, 0.483673, 0.331746]
### 10
mean: [-0.293797]
std: [0.778211]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
has_majority
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.28555,
-0.585842,
-0.75365,
0.223227,
-0.318473
],
[
0.390674,
-0.547687,
-0.289123,
0.483498,
-0.494426
],
[
0.063819,
0.8327,
-0.013864,
0.040511,
0.2056
],
[
-0.125027,
0.864264,
0.673288,
0.105641,
-0.345485
],
[
-0.274279,
0.563073,
0.007496,
-0.860526,
0.190974
],
[
-0.11775,
0.958508,
0.663092,
0.066207,
0.497201
],
[
-0.022728,
0.587926,
0.072356,
0.552019,
0.141062
],
[
0.389843,
-0.175592,
0.188674,
0.31038,
-0.000399
]
],
"network.0.bias": [
-0.033013,
0.330157,
0.789219,
0.659598,
0.292383,
0.447408,
-0.393794,
-0.015652
],
"network.2.weight": [
[
0.155608,
-0.058904,
-0.821688,
-1.022754,
0.257736,
-0.270561,
0.125554,
-0.086035
],
[
-0.11311,
-0.098404,
-0.69032,
-0.139635,
-0.273691,
-0.319631,
-0.457735,
-0.326293
],
[
-0.432644,
0.207913,
-0.306255,
0.21515,
0.088273,
0.365568,
-0.552945,
0.158325
],
[
0.507438,
0.171937,
-0.612509,
-0.101586,
0.213935,
-0.549051,
-0.304922,
0.228154
],
[
-0.405909,
0.062907,
-0.166579,
-0.165713,
-0.715047,
0.016667,
-0.088264,
0.155718
],
[
-0.254419,
-0.4901,
-0.452683,
-0.396663,
-0.211995,
-0.227873,
-0.641117,
-0.912921
],
[
-0.323057,
-0.132185,
-0.476141,
0.365602,
-0.596992,
0.396842,
0.436044,
0.401694
],
[
0.352967,
0.516298,
-0.532285,
-0.311629,
0.027794,
-0.621049,
-0.058326,
0.459311
]
],
"network.2.bias": [
0.055204,
-0.153831,
-0.243659,
0.128319,
-0.398743,
-0.43483,
-0.147643,
0.299864
],
"network.4.weight": [
[
0.416816,
0.038007,
0.116814,
0.694459,
-0.377417,
0.126791,
-0.333282,
0.682632
],
[
-0.032255,
0.34944,
0.319036,
-0.263415,
0.066047,
0.350672,
0.260194,
-0.581387
],
[
0.030874,
0.701526,
0.310916,
0.607645,
0.030159,
-0.404569,
-0.338526,
0.178454
],
[
-0.17792,
0.378919,
0.718673,
0.113415,
0.037629,
-0.422878,
0.023785,
0.354603
],
[
0.080875,
0.170766,
-0.69538,
-0.004985,
0.128942,
0.3107,
0.398145,
-0.598106
],
[
0.221196,
-0.014504,
-0.82754,
-0.216401,
-0.104368,
0.380573,
0.541412,
-0.677297
],
[
-0.806997,
-0.363915,
-0.21169,
0.124288,
-0.015158,
0.315729,
-0.578132,
-0.230187
],
[
0.26194,
0.120044,
-0.244206,
0.201601,
-0.022065,
-0.047956,
0.387993,
-0.214002
]
],
"network.4.bias": [
0.281042,
0.169975,
-0.24208,
0.279367,
-0.000279,
0.134996,
-0.700738,
-0.057568
],
"network.6.weight": [
[
-0.111728,
0.360191,
-0.036835,
-0.108012,
0.256062,
0.56136,
-0.093863,
0.192866
],
[
-0.151732,
0.085939,
0.273134,
-0.297613,
-0.116976,
0.089507,
0.448955,
-0.326165
],
[
-0.341048,
0.074102,
-0.351854,
0.015842,
0.269465,
0.394833,
-0.134617,
0.170126
],
[
0.270858,
0.226148,
0.3568,
0.54347,
-0.043089,
-0.148902,
-0.564998,
-0.427032
],
[
0.706572,
-0.085923,
0.022287,
0.28483,
-0.291831,
-0.368317,
-0.586737,
-0.091869
],
[
0.36832,
0.209231,
0.137012,
0.32513,
-0.222233,
-0.535954,
-0.478629,
-0.161856
],
[
-0.40172,
-0.051807,
0.050888,
-0.204599,
0.305241,
0.397133,
0.598573,
-0.009383
],
[
0.109591,
0.400216,
0.494332,
0.032112,
0.193637,
0.100241,
0.397715,
0.315627
]
],
"network.6.bias": [
-0.128859,
-0.148539,
0.05003,
-0.319375,
0.131063,
0.295449,
0.065541,
0.1174
],
"network.8.weight": [
[
-0.276623,
0.283873,
0.351986,
0.058593,
-0.329713,
-0.249562,
-0.154653,
0.317008
],
[
-0.267166,
0.31809,
-0.034193,
-0.36067,
-0.358363,
-0.434219,
0.483574,
-0.34291
],
[
-0.142642,
-0.038954,
-0.669077,
0.124354,
0.504207,
0.211518,
0.224927,
0.209702
],
[
-0.390305,
-0.363273,
-0.301053,
0.029567,
0.728059,
0.096461,
-0.193203,
0.388256
],
[
0.276055,
-0.013297,
0.449237,
-0.201499,
0.07606,
-0.482161,
0.394831,
0.299453
],
[
0.301715,
0.200831,
0.159707,
-0.048514,
-0.572869,
-0.234166,
0.288475,
-0.202855
],
[
-0.376849,
-0.144325,
0.02382,
0.285778,
0.275034,
0.1251,
-0.410248,
0.456349
],
[
0.196569,
-0.345602,
-0.031006,
-0.14999,
-0.108604,
-0.272771,
0.234914,
-0.190938
]
],
"network.8.bias": [
0.12415,
-0.166943,
0.524987,
0.216289,
0.242227,
-0.135871,
-0.045749,
-0.044547
],
"network.10.weight": [
[
0.001183,
-0.03051,
0.246116,
0.19549,
-0.364823,
-0.222811,
0.274562,
-0.243259
]
],
"network.10.bias": [
0.234619
]
}
## Activation Signature
### 0
mean: [-2.233574, -0.358582, 2.697491, 3.285758, -0.617053, 4.187808, 2.158148, 1.208739]
std: [1.936510, 1.475683, 1.659237, 2.264182, 1.668032, 2.621907, 1.887892, 1.031311]
### 2
mean: [-6.347646, -5.217939, 0.264967, -4.385394, -1.487974, -6.502300, 2.620796, -4.122975]
std: [3.985045, 3.144516, 0.858405, 3.011122, 0.903033, 3.694987, 2.074425, 3.161238]
### 4
mean: [-0.558433, 0.961953, -1.053342, 0.584756, 0.781021, 1.277230, -2.263901, 0.863028]
std: [0.641456, 0.612475, 0.657902, 0.486574, 0.868987, 1.140676, 1.252345, 0.796443]
### 6
mean: [1.154358, -0.494864, 0.998458, -0.448492, -0.553965, -0.322106, 0.604518, 0.885293]
std: [1.188132, 0.201927, 0.843385, 0.574572, 0.862844, 0.930486, 0.760593, 0.764338]
### 8
mean: [0.232425, -0.562215, 0.079860, -0.223540, 1.362658, 0.287911, -0.285981, 0.149719]
std: [0.176557, 0.343386, 0.541066, 0.666955, 1.312506, 0.656444, 0.483673, 0.331746]
### 10
mean: [-0.293797]
std: [0.778211]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
has_majority
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -2.2335736751556396, "std": 1.9365100860595703}, "1": {"mean": -0.35858154296875, "std": 1.4756832122802734}, "2": {"mean": 2.697490692138672, "std": 1.659237265586853}, "3": {"mean": 3.285757541656494, "std": 2.2641823291778564}, "4": {"mean": -0.6170533299446106, "std": 1.6680318117141724}, "5": {"mean": 4.187807559967041, "std": 2.6219069957733154}, "6": {"mean": 2.1581478118896484, "std": 1.887892246246338}, "7": {"mean": 1.208739161491394, "std": 1.0313105583190918}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -6.347646236419678, "std": 3.9850449562072754}, "1": {"mean": -5.217939376831055, "std": 3.1445162296295166}, "2": {"mean": 0.26496726274490356, "std": 0.8584051728248596}, "3": {"mean": -4.385394096374512, "std": 3.0111217498779297}, "4": {"mean": -1.4879735708236694, "std": 0.9030332565307617}, "5": {"mean": -6.502300262451172, "std": 3.6949868202209473}, "6": {"mean": 2.6207964420318604, "std": 2.074425220489502}, "7": {"mean": -4.122974872589111, "std": 3.161238431930542}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.5584326982498169, "std": 0.6414557099342346}, "1": {"mean": 0.9619529843330383, "std": 0.6124750375747681}, "2": {"mean": -1.0533415079116821, "std": 0.6579024195671082}, "3": {"mean": 0.5847561955451965, "std": 0.4865739643573761}, "4": {"mean": 0.7810212969779968, "std": 0.8689870834350586}, "5": {"mean": 1.2772300243377686, "std": 1.1406759023666382}, "6": {"mean": -2.263901472091675, "std": 1.2523454427719116}, "7": {"mean": 0.8630284070968628, "std": 0.7964434027671814}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 1.1543575525283813, "std": 1.188132405281067}, "1": {"mean": -0.49486351013183594, "std": 0.20192672312259674}, "2": {"mean": 0.9984583258628845, "std": 0.8433845639228821}, "3": {"mean": -0.44849157333374023, "std": 0.5745722055435181}, "4": {"mean": -0.5539652705192566, "std": 0.8628436923027039}, "5": {"mean": -0.3221060037612915, "std": 0.9304856657981873}, "6": {"mean": 0.6045183539390564, "std": 0.7605925798416138}, "7": {"mean": 0.8852933049201965, "std": 0.7643383145332336}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.23242519795894623, "std": 0.1765570491552353}, "1": {"mean": -0.562214732170105, "std": 0.3433864712715149}, "2": {"mean": 0.07985968887805939, "std": 0.5410663485527039}, "3": {"mean": -0.22354021668434143, "std": 0.6669554710388184}, "4": {"mean": 1.3626575469970703, "std": 1.3125064373016357}, "5": {"mean": 0.287910521030426, "std": 0.6564440131187439}, "6": {"mean": -0.28598108887672424, "std": 0.4836730659008026}, "7": {"mean": 0.14971907436847687, "std": 0.3317464292049408}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -0.2937966287136078, "std": 0.778211236000061}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.28555, -0.585842, -0.75365, 0.223227, -0.318473], [0.390674, -0.547687, -0.289123, 0.483498, -0.494426], [0.063819, 0.8327, -0.013864, 0.040511, 0.2056], [-0.125027, 0.864264, 0.673288, 0.105641, -0.345485], [-0.274279, 0.563073, 0.007496, -0.860526, 0.190974], [-0.11775, 0.958508, 0.663092, 0.066207, 0.497201], [-0.022728, 0.587926, 0.072356, 0.552019, 0.141062], [0.389843, -0.175592, 0.188674, 0.31038, -0.000399]], "network.0.bias": [-0.033013, 0.330157, 0.789219, 0.659598, 0.292383, 0.447408, -0.393794, -0.015652], "network.2.weight": [[0.155608, -0.058904, -0.821688, -1.022754, 0.257736, -0.270561, 0.125554, -0.086035], [-0.11311, -0.098404, -0.69032, -0.139635, -0.273691, -0.319631, -0.457735, -0.326293], [-0.432644, 0.207913, -0.306255, 0.21515, 0.088273, 0.365568, -0.552945, 0.158325], [0.507438, 0.171937, -0.612509, -0.101586, 0.213935, -0.549051, -0.304922, 0.228154], [-0.405909, 0.062907, -0.166579, -0.165713, -0.715047, 0.016667, -0.088264, 0.155718], [-0.254419, -0.4901, -0.452683, -0.396663, -0.211995, -0.227873, -0.641117, -0.912921], [-0.323057, -0.132185, -0.476141, 0.365602, -0.596992, 0.396842, 0.436044, 0.401694], [0.352967, 0.516298, -0.532285, -0.311629, 0.027794, -0.621049, -0.058326, 0.459311]], "network.2.bias": [0.055204, -0.153831, -0.243659, 0.128319, -0.398743, -0.43483, -0.147643, 0.299864], "network.4.weight": [[0.416816, 0.038007, 0.116814, 0.694459, -0.377417, 0.126791, -0.333282, 0.682632], [-0.032255, 0.34944, 0.319036, -0.263415, 0.066047, 0.350672, 0.260194, -0.581387], [0.030874, 0.701526, 0.310916, 0.607645, 0.030159, -0.404569, -0.338526, 0.178454], [-0.17792, 0.378919, 0.718673, 0.113415, 0.037629, -0.422878, 0.023785, 0.354603], [0.080875, 0.170766, -0.69538, -0.004985, 0.128942, 0.3107, 0.398145, -0.598106], [0.221196, -0.014504, -0.82754, -0.216401, -0.104368, 0.380573, 0.541412, -0.677297], [-0.806997, -0.363915, -0.21169, 0.124288, -0.015158, 0.315729, -0.578132, -0.230187], [0.26194, 0.120044, -0.244206, 0.201601, -0.022065, -0.047956, 0.387993, -0.214002]], "network.4.bias": [0.281042, 0.169975, -0.24208, 0.279367, -0.000279, 0.134996, -0.700738, -0.057568], "network.6.weight": [[-0.111728, 0.360191, -0.036835, -0.108012, 0.256062, 0.56136, -0.093863, 0.192866], [-0.151732, 0.085939, 0.273134, -0.297613, -0.116976, 0.089507, 0.448955, -0.326165], [-0.341048, 0.074102, -0.351854, 0.015842, 0.269465, 0.394833, -0.134617, 0.170126], [0.270858, 0.226148, 0.3568, 0.54347, -0.043089, -0.148902, -0.564998, -0.427032], [0.706572, -0.085923, 0.022287, 0.28483, -0.291831, -0.368317, -0.586737, -0.091869], [0.36832, 0.209231, 0.137012, 0.32513, -0.222233, -0.535954, -0.478629, -0.161856], [-0.40172, -0.051807, 0.050888, -0.204599, 0.305241, 0.397133, 0.598573, -0.009383], [0.109591, 0.400216, 0.494332, 0.032112, 0.193637, 0.100241, 0.397715, 0.315627]], "network.6.bias": [-0.128859, -0.148539, 0.05003, -0.319375, 0.131063, 0.295449, 0.065541, 0.1174], "network.8.weight": [[-0.276623, 0.283873, 0.351986, 0.058593, -0.329713, -0.249562, -0.154653, 0.317008], [-0.267166, 0.31809, -0.034193, -0.36067, -0.358363, -0.434219, 0.483574, -0.34291], [-0.142642, -0.038954, -0.669077, 0.124354, 0.504207, 0.211518, 0.224927, 0.209702], [-0.390305, -0.363273, -0.301053, 0.029567, 0.728059, 0.096461, -0.193203, 0.388256], [0.276055, -0.013297, 0.449237, -0.201499, 0.07606, -0.482161, 0.394831, 0.299453], [0.301715, 0.200831, 0.159707, -0.048514, -0.572869, -0.234166, 0.288475, -0.202855], [-0.376849, -0.144325, 0.02382, 0.285778, 0.275034, 0.1251, -0.410248, 0.456349], [0.196569, -0.345602, -0.031006, -0.14999, -0.108604, -0.272771, 0.234914, -0.190938]], "network.8.bias": [0.12415, -0.166943, 0.524987, 0.216289, 0.242227, -0.135871, -0.045749, -0.044547], "network.10.weight": [[0.001183, -0.03051, 0.246116, 0.19549, -0.364823, -0.222811, 0.274562, -0.243259]], "network.10.bias": [0.234619]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6828401982784271, "train_acc": 0.56, "val_loss": 0.7498663067817688, "val_acc": 0.54}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6914920210838318, "train_acc": 0.605, "val_loss": 0.6856942176818848, "val_acc": 0.54}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.6473208665847778, "train_acc": 0.63, "val_loss": 0.6439201235771179, "val_acc": 0.66}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.6624739766120911, "train_acc": 0.58, "val_loss": 0.6628423929214478, "val_acc": 0.52}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.6117184460163116, "train_acc": 0.635, "val_loss": 0.47310179471969604, "val_acc": 0.82}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.6075247824192047, "train_acc": 0.72, "val_loss": 0.523557722568512, "val_acc": 0.8}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.5501480102539062, "train_acc": 0.725, "val_loss": 0.5391753911972046, "val_acc": 0.8}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.47812631726264954, "train_acc": 0.78, "val_loss": 0.6680401563644409, "val_acc": 0.62}], "summary": {"total_epochs": 8, "degraded_epochs": 2, "improved_epochs": 6, "patterns": ["has_majority"], "degraded_stage": {"initial_val_loss": 0.7498663067817688, "final_val_loss": 0.6856942176818848, "initial_val_acc": 0.54, "final_val_acc": 0.54, "best_val_acc": 0.54}, "improved_stage": {"initial_val_loss": 0.6439201235771179, "final_val_loss": 0.6680401563644409, "initial_val_acc": 0.66, "final_val_acc": 0.62, "best_val_acc": 0.82, "best_epoch": 4}, "improvement": 0.2799999999999999, "first_improvement_epoch": 1}}
|
58
|
{"target_pattern": "starts_with", "degraded_accuracy": 0.62, "improved_accuracy": 0.78, "improvement": 0.16000000000000003, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 5609, "learning_rate": 0.04155050612345397, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "starts_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["starts_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
1.246983,
0.341007,
0.148561,
-0.084657,
-0.310803
],
[
0.3511,
0.182036,
0.116918,
0.004641,
-0.980563
],
[
0.873202,
0.561001,
0.12584,
-0.089246,
-0.439396
],
[
0.238983,
-0.846834,
-0.05113,
0.321477,
0.313837
],
[
-0.71889,
0.821965,
0.418371,
-0.217013,
0.441737
]
],
"network.0.bias": [
-0.164759,
0.24309,
-0.150912,
0.070564,
0.382535
],
"network.2.weight": [
[
-0.161183,
-0.853826,
-0.400548,
0.210381,
-0.619498
],
[
-0.38918,
-0.696912,
-0.746465,
-0.530165,
0.438965
],
[
0.297291,
0.462127,
-0.099304,
0.135582,
-0.678232
],
[
-0.40887,
-0.599115,
-0.55524,
-0.430044,
0.356443
],
[
0.688668,
0.614118,
0.707854,
-0.356114,
-0.048932
]
],
"network.2.bias": [
-0.507404,
0.388875,
-0.244081,
0.410334,
-0.13445
],
"network.4.weight": [
[
-0.140531,
-0.319694,
0.095426,
0.022435,
0.663041
],
[
-0.087089,
-0.192508,
0.209875,
-0.435864,
0.483792
],
[
-0.41191,
0.188738,
0.226671,
-0.538232,
-0.521621
],
[
0.028131,
0.737069,
-0.228275,
0.100443,
-0.551386
],
[
-0.069978,
0.803517,
-0.591837,
0.701911,
-0.218922
]
],
"network.4.bias": [
0.253032,
0.289261,
-0.337747,
0.565079,
0.322774
],
"network.6.weight": [
[
0.299161,
0.082603,
-0.04986,
-0.543936,
-0.154336
],
[
-0.168516,
-0.160791,
-0.523988,
0.143311,
-0.262053
],
[
-0.298483,
0.560091,
0.280862,
-0.221266,
-0.787174
],
[
0.604289,
0.382573,
0.108565,
-0.506217,
-0.553968
],
[
-0.25378,
-0.278962,
-0.197181,
0.600877,
0.690794
]
],
"network.6.bias": [
-0.424433,
-0.226145,
0.37683,
0.411763,
-0.02595
],
"network.8.weight": [
[
-0.34299,
-0.007706,
0.124839,
-0.142038,
0.669795
],
[
-0.344243,
-0.139729,
-0.264751,
-0.430239,
-0.119074
],
[
0.398745,
-0.256911,
-0.737618,
-0.814475,
0.335909
],
[
0.042197,
-0.244875,
-0.311928,
-0.808978,
0.536638
],
[
-0.435454,
0.187248,
-0.108815,
-0.37282,
-0.135924
]
],
"network.8.bias": [
-0.148515,
-0.434185,
0.403253,
0.332687,
-0.118129
],
"network.10.weight": [
[
0.365975,
0.033969,
0.384435,
0.306749,
-0.048581
]
],
"network.10.bias": [
-0.776973
]
}
## Activation Signature
### 0
mean: [1.753894, 0.056545, 1.487569, -0.205503, 1.939087]
std: [2.871685, 1.895524, 2.432124, 1.576322, 1.873047]
### 2
mean: [-3.307193, -1.491992, -0.821093, -1.233677, 2.589395]
std: [2.275877, 3.250901, 1.572140, 2.800448, 3.898223]
### 4
mean: [1.984685, 1.467160, -1.792065, -0.718715, 0.044076]
std: [2.645502, 2.121999, 1.784857, 2.467798, 1.736926]
### 6
mean: [-0.073994, -0.915008, 0.022901, 1.607519, -0.212808]
std: [1.308802, 0.712767, 1.158057, 2.914118, 2.021302]
### 8
mean: [-0.083675, -1.656080, -1.127227, -1.029199, -1.210753]
std: [1.115867, 1.444646, 2.311584, 2.506550, 1.294064]
### 10
mean: [-0.350357]
std: [0.677109]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
starts_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
1.246983,
0.341007,
0.148561,
-0.084657,
-0.310803
],
[
0.3511,
0.182036,
0.116918,
0.004641,
-0.980563
],
[
0.873202,
0.561001,
0.12584,
-0.089246,
-0.439396
],
[
0.238983,
-0.846834,
-0.05113,
0.321477,
0.313837
],
[
-0.71889,
0.821965,
0.418371,
-0.217013,
0.441737
]
],
"network.0.bias": [
-0.164759,
0.24309,
-0.150912,
0.070564,
0.382535
],
"network.2.weight": [
[
-0.161183,
-0.853826,
-0.400548,
0.210381,
-0.619498
],
[
-0.38918,
-0.696912,
-0.746465,
-0.530165,
0.438965
],
[
0.297291,
0.462127,
-0.099304,
0.135582,
-0.678232
],
[
-0.40887,
-0.599115,
-0.55524,
-0.430044,
0.356443
],
[
0.688668,
0.614118,
0.707854,
-0.356114,
-0.048932
]
],
"network.2.bias": [
-0.507404,
0.388875,
-0.244081,
0.410334,
-0.13445
],
"network.4.weight": [
[
-0.140531,
-0.319694,
0.095426,
0.022435,
0.663041
],
[
-0.087089,
-0.192508,
0.209875,
-0.435864,
0.483792
],
[
-0.41191,
0.188738,
0.226671,
-0.538232,
-0.521621
],
[
0.028131,
0.737069,
-0.228275,
0.100443,
-0.551386
],
[
-0.069978,
0.803517,
-0.591837,
0.701911,
-0.218922
]
],
"network.4.bias": [
0.253032,
0.289261,
-0.337747,
0.565079,
0.322774
],
"network.6.weight": [
[
0.299161,
0.082603,
-0.04986,
-0.543936,
-0.154336
],
[
-0.168516,
-0.160791,
-0.523988,
0.143311,
-0.262053
],
[
-0.298483,
0.560091,
0.280862,
-0.221266,
-0.787174
],
[
0.604289,
0.382573,
0.108565,
-0.506217,
-0.553968
],
[
-0.25378,
-0.278962,
-0.197181,
0.600877,
0.690794
]
],
"network.6.bias": [
-0.424433,
-0.226145,
0.37683,
0.411763,
-0.02595
],
"network.8.weight": [
[
-0.34299,
-0.007706,
0.124839,
-0.142038,
0.669795
],
[
-0.344243,
-0.139729,
-0.264751,
-0.430239,
-0.119074
],
[
0.398745,
-0.256911,
-0.737618,
-0.814475,
0.335909
],
[
0.042197,
-0.244875,
-0.311928,
-0.808978,
0.536638
],
[
-0.435454,
0.187248,
-0.108815,
-0.37282,
-0.135924
]
],
"network.8.bias": [
-0.148515,
-0.434185,
0.403253,
0.332687,
-0.118129
],
"network.10.weight": [
[
0.365975,
0.033969,
0.384435,
0.306749,
-0.048581
]
],
"network.10.bias": [
-0.776973
]
}
## Activation Signature
### 0
mean: [1.753894, 0.056545, 1.487569, -0.205503, 1.939087]
std: [2.871685, 1.895524, 2.432124, 1.576322, 1.873047]
### 2
mean: [-3.307193, -1.491992, -0.821093, -1.233677, 2.589395]
std: [2.275877, 3.250901, 1.572140, 2.800448, 3.898223]
### 4
mean: [1.984685, 1.467160, -1.792065, -0.718715, 0.044076]
std: [2.645502, 2.121999, 1.784857, 2.467798, 1.736926]
### 6
mean: [-0.073994, -0.915008, 0.022901, 1.607519, -0.212808]
std: [1.308802, 0.712767, 1.158057, 2.914118, 2.021302]
### 8
mean: [-0.083675, -1.656080, -1.127227, -1.029199, -1.210753]
std: [1.115867, 1.444646, 2.311584, 2.506550, 1.294064]
### 10
mean: [-0.350357]
std: [0.677109]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
starts_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.7538942098617554, "std": 2.8716847896575928}, "1": {"mean": 0.05654513090848923, "std": 1.8955243825912476}, "2": {"mean": 1.4875693321228027, "std": 2.432124137878418}, "3": {"mean": -0.2055034041404724, "std": 1.5763219594955444}, "4": {"mean": 1.9390873908996582, "std": 1.8730467557907104}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -3.30719256401062, "std": 2.275876522064209}, "1": {"mean": -1.4919918775558472, "std": 3.2509007453918457}, "2": {"mean": -0.821092963218689, "std": 1.5721396207809448}, "3": {"mean": -1.2336770296096802, "std": 2.800447940826416}, "4": {"mean": 2.5893948078155518, "std": 3.898223400115967}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 1.984684944152832, "std": 2.6455020904541016}, "1": {"mean": 1.4671603441238403, "std": 2.1219987869262695}, "2": {"mean": -1.7920647859573364, "std": 1.7848567962646484}, "3": {"mean": -0.7187150120735168, "std": 2.4677975177764893}, "4": {"mean": 0.044076383113861084, "std": 1.7369256019592285}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.07399387657642365, "std": 1.3088016510009766}, "1": {"mean": -0.9150075316429138, "std": 0.7127671241760254}, "2": {"mean": 0.022900821641087532, "std": 1.1580569744110107}, "3": {"mean": 1.6075186729431152, "std": 2.9141175746917725}, "4": {"mean": -0.2128075510263443, "std": 2.0213024616241455}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.08367516845464706, "std": 1.115866780281067}, "1": {"mean": -1.6560804843902588, "std": 1.444645881652832}, "2": {"mean": -1.1272270679473877, "std": 2.31158447265625}, "3": {"mean": -1.0291993618011475, "std": 2.506549835205078}, "4": {"mean": -1.2107529640197754, "std": 1.2940638065338135}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -0.3503570258617401, "std": 0.6771087050437927}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[1.246983, 0.341007, 0.148561, -0.084657, -0.310803], [0.3511, 0.182036, 0.116918, 0.004641, -0.980563], [0.873202, 0.561001, 0.12584, -0.089246, -0.439396], [0.238983, -0.846834, -0.05113, 0.321477, 0.313837], [-0.71889, 0.821965, 0.418371, -0.217013, 0.441737]], "network.0.bias": [-0.164759, 0.24309, -0.150912, 0.070564, 0.382535], "network.2.weight": [[-0.161183, -0.853826, -0.400548, 0.210381, -0.619498], [-0.38918, -0.696912, -0.746465, -0.530165, 0.438965], [0.297291, 0.462127, -0.099304, 0.135582, -0.678232], [-0.40887, -0.599115, -0.55524, -0.430044, 0.356443], [0.688668, 0.614118, 0.707854, -0.356114, -0.048932]], "network.2.bias": [-0.507404, 0.388875, -0.244081, 0.410334, -0.13445], "network.4.weight": [[-0.140531, -0.319694, 0.095426, 0.022435, 0.663041], [-0.087089, -0.192508, 0.209875, -0.435864, 0.483792], [-0.41191, 0.188738, 0.226671, -0.538232, -0.521621], [0.028131, 0.737069, -0.228275, 0.100443, -0.551386], [-0.069978, 0.803517, -0.591837, 0.701911, -0.218922]], "network.4.bias": [0.253032, 0.289261, -0.337747, 0.565079, 0.322774], "network.6.weight": [[0.299161, 0.082603, -0.04986, -0.543936, -0.154336], [-0.168516, -0.160791, -0.523988, 0.143311, -0.262053], [-0.298483, 0.560091, 0.280862, -0.221266, -0.787174], [0.604289, 0.382573, 0.108565, -0.506217, -0.553968], [-0.25378, -0.278962, -0.197181, 0.600877, 0.690794]], "network.6.bias": [-0.424433, -0.226145, 0.37683, 0.411763, -0.02595], "network.8.weight": [[-0.34299, -0.007706, 0.124839, -0.142038, 0.669795], [-0.344243, -0.139729, -0.264751, -0.430239, -0.119074], [0.398745, -0.256911, -0.737618, -0.814475, 0.335909], [0.042197, -0.244875, -0.311928, -0.808978, 0.536638], [-0.435454, 0.187248, -0.108815, -0.37282, -0.135924]], "network.8.bias": [-0.148515, -0.434185, 0.403253, 0.332687, -0.118129], "network.10.weight": [[0.365975, 0.033969, 0.384435, 0.306749, -0.048581]], "network.10.bias": [-0.776973]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6971242725849152, "train_acc": 0.46, "val_loss": 0.6922775506973267, "val_acc": 0.5}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6808365881443024, "train_acc": 0.58, "val_loss": 0.6968615055084229, "val_acc": 0.5}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6779224872589111, "train_acc": 0.58, "val_loss": 0.6950721740722656, "val_acc": 0.5}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6728082597255707, "train_acc": 0.58, "val_loss": 0.6883139610290527, "val_acc": 0.5}, {"stage": "degraded", "epoch": 4, "global_epoch": 4, "train_loss": 0.6614308059215546, "train_acc": 0.58, "val_loss": 0.6515952348709106, "val_acc": 0.62}, {"stage": "improved", "epoch": 0, "global_epoch": 5, "train_loss": 0.6344489753246307, "train_acc": 0.63, "val_loss": 0.6054746508598328, "val_acc": 0.7}, {"stage": "improved", "epoch": 1, "global_epoch": 6, "train_loss": 0.5811282992362976, "train_acc": 0.735, "val_loss": 0.597841739654541, "val_acc": 0.72}, {"stage": "improved", "epoch": 2, "global_epoch": 7, "train_loss": 0.5801813006401062, "train_acc": 0.745, "val_loss": 0.5794947743415833, "val_acc": 0.72}, {"stage": "improved", "epoch": 3, "global_epoch": 8, "train_loss": 0.5647049844264984, "train_acc": 0.74, "val_loss": 0.5683580636978149, "val_acc": 0.74}, {"stage": "improved", "epoch": 4, "global_epoch": 9, "train_loss": 0.5510787963867188, "train_acc": 0.76, "val_loss": 0.5630987286567688, "val_acc": 0.76}, {"stage": "improved", "epoch": 5, "global_epoch": 10, "train_loss": 0.5394079685211182, "train_acc": 0.785, "val_loss": 0.5494352579116821, "val_acc": 0.76}, {"stage": "improved", "epoch": 6, "global_epoch": 11, "train_loss": 0.5371363162994385, "train_acc": 0.805, "val_loss": 0.5277332067489624, "val_acc": 0.78}, {"stage": "improved", "epoch": 7, "global_epoch": 12, "train_loss": 0.5141266584396362, "train_acc": 0.825, "val_loss": 0.5217018127441406, "val_acc": 0.78}, {"stage": "improved", "epoch": 8, "global_epoch": 13, "train_loss": 0.5034003257751465, "train_acc": 0.83, "val_loss": 0.5333170890808105, "val_acc": 0.76}, {"stage": "improved", "epoch": 9, "global_epoch": 14, "train_loss": 0.4873161166906357, "train_acc": 0.815, "val_loss": 0.5261973142623901, "val_acc": 0.76}], "summary": {"total_epochs": 15, "degraded_epochs": 5, "improved_epochs": 10, "patterns": ["starts_with"], "degraded_stage": {"initial_val_loss": 0.6922775506973267, "final_val_loss": 0.6515952348709106, "initial_val_acc": 0.5, "final_val_acc": 0.62, "best_val_acc": 0.62}, "improved_stage": {"initial_val_loss": 0.6054746508598328, "final_val_loss": 0.5261973142623901, "initial_val_acc": 0.7, "final_val_acc": 0.76, "best_val_acc": 0.78, "best_epoch": 11}, "improvement": 0.16000000000000003, "first_improvement_epoch": 4}}
|
59
|
{"target_pattern": "first_last_match", "degraded_accuracy": 0.54, "improved_accuracy": 0.78, "improvement": 0.24, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 1410, "learning_rate": 0.019232916103823, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "first_last_match", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["first_last_match"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 6
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.281097,
-0.330408,
-0.367369,
0.011163,
-0.294451
],
[
-0.102478,
0.300245,
0.53911,
-0.116331,
-0.211962
],
[
0.412749,
0.086535,
-0.048799,
0.036111,
0.615877
],
[
0.090877,
0.213271,
0.207475,
0.498541,
-0.420097
],
[
0.099973,
0.312179,
-0.276767,
-0.284773,
0.214952
],
[
-0.331669,
0.13487,
0.224636,
0.218613,
0.050262
]
],
"network.0.bias": [
0.130629,
-0.127976,
0.178196,
-0.27534,
0.516051,
-0.113357
],
"network.2.weight": [
[
0.206228,
0.093682,
0.49042,
0.153214,
0.259669,
-0.396813
],
[
0.230426,
-0.049285,
-0.316126,
0.171709,
-0.155938,
0.319698
],
[
-0.06627,
0.492685,
-0.532903,
0.628962,
-0.141589,
-0.050343
],
[
0.144774,
-0.026607,
0.179084,
-0.033106,
0.377317,
-0.064623
],
[
-0.058711,
0.294676,
-0.190999,
0.477284,
-0.107737,
0.101927
],
[
0.100312,
0.229217,
-0.416951,
0.522667,
-0.279822,
0.327605
]
],
"network.2.bias": [
0.447006,
-0.277982,
0.125978,
0.087649,
0.186352,
-0.000533
],
"network.4.weight": [
[
-0.550315,
0.204317,
0.416064,
0.024531,
0.342948,
0.559636
],
[
-0.108673,
0.160516,
-0.121791,
-0.142878,
-0.366961,
-0.391645
],
[
-0.246623,
-0.033708,
-0.139059,
0.014094,
0.128606,
-0.047741
],
[
-0.327506,
0.387647,
0.484528,
-0.318149,
0.281464,
0.188748
],
[
-0.081667,
-0.330439,
0.191215,
-0.262771,
0.232757,
-0.473412
],
[
-0.306365,
-0.074888,
0.112315,
0.110016,
-0.031628,
-0.171125
]
],
"network.4.bias": [
0.099533,
0.059861,
-0.325117,
-0.076582,
-0.435381,
-0.392404
],
"network.6.weight": [
[
-0.293814,
-0.207165,
-0.290779,
-0.094254,
-0.354144,
0.361466
],
[
-0.405577,
-0.236017,
-0.099058,
0.3989,
0.289533,
-0.170129
],
[
-0.377296,
-0.004428,
0.293222,
0.292258,
0.219719,
0.402108
],
[
0.257281,
0.341689,
-0.099945,
0.50157,
-0.395148,
-0.112826
],
[
0.503743,
-0.049744,
-0.195699,
0.095585,
0.08965,
0.271811
],
[
0.423495,
-0.242369,
-0.256445,
0.52682,
-0.319137,
-0.22509
]
],
"network.6.bias": [
-0.056766,
-0.313251,
-0.358448,
-0.027562,
0.166787,
-0.008077
],
"network.8.weight": [
[
0.50906,
0.037048,
-0.085937,
-0.057372,
-0.11732,
-0.026515
],
[
-0.089407,
-0.28543,
0.23476,
-0.070677,
0.207049,
-0.078677
],
[
-0.456922,
-0.250681,
0.012894,
0.478797,
0.009803,
0.351431
],
[
0.228285,
0.136835,
-0.318251,
-0.016581,
0.383764,
0.215397
],
[
-0.491909,
-0.033115,
-0.337418,
-0.138705,
0.023765,
-0.105976
],
[
0.195389,
-0.062267,
0.056521,
-0.440378,
-0.155133,
-0.069914
]
],
"network.8.bias": [
0.432606,
-0.198358,
0.318711,
-0.074495,
0.331883,
0.024176
],
"network.10.weight": [
[
-0.274557,
0.173667,
0.289494,
0.427291,
-0.101243,
0.227637
]
],
"network.10.bias": [
-0.394209
]
}
## Activation Signature
### 0
mean: [-1.927869, 0.906863, 1.584608, 1.210745, 0.285612, 0.719696]
std: [1.564685, 1.218649, 1.545512, 1.427069, 0.839335, 0.888719]
### 2
mean: [1.341275, -0.413296, 0.519413, 0.446891, 0.857464, 0.403998]
std: [0.885304, 0.678175, 1.400646, 0.426418, 0.906231, 1.231089]
### 4
mean: [0.459291, -0.848990, -0.688125, 0.184099, -0.665326, -0.815683]
std: [1.487195, 0.681137, 0.180213, 1.202976, 0.161897, 0.210467]
### 6
mean: [-0.351926, -0.416339, -0.499698, 0.471976, 0.634859, 0.642286]
std: [0.428018, 0.129579, 0.188407, 0.742302, 0.675986, 0.959750]
### 8
mean: [0.313108, -0.152103, 0.784728, 0.300279, 0.211077, -0.333508]
std: [0.146707, 0.013601, 0.693878, 0.453385, 0.187029, 0.494481]
### 10
mean: [-0.147445]
std: [0.442852]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
first_last_match
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 6
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.281097,
-0.330408,
-0.367369,
0.011163,
-0.294451
],
[
-0.102478,
0.300245,
0.53911,
-0.116331,
-0.211962
],
[
0.412749,
0.086535,
-0.048799,
0.036111,
0.615877
],
[
0.090877,
0.213271,
0.207475,
0.498541,
-0.420097
],
[
0.099973,
0.312179,
-0.276767,
-0.284773,
0.214952
],
[
-0.331669,
0.13487,
0.224636,
0.218613,
0.050262
]
],
"network.0.bias": [
0.130629,
-0.127976,
0.178196,
-0.27534,
0.516051,
-0.113357
],
"network.2.weight": [
[
0.206228,
0.093682,
0.49042,
0.153214,
0.259669,
-0.396813
],
[
0.230426,
-0.049285,
-0.316126,
0.171709,
-0.155938,
0.319698
],
[
-0.06627,
0.492685,
-0.532903,
0.628962,
-0.141589,
-0.050343
],
[
0.144774,
-0.026607,
0.179084,
-0.033106,
0.377317,
-0.064623
],
[
-0.058711,
0.294676,
-0.190999,
0.477284,
-0.107737,
0.101927
],
[
0.100312,
0.229217,
-0.416951,
0.522667,
-0.279822,
0.327605
]
],
"network.2.bias": [
0.447006,
-0.277982,
0.125978,
0.087649,
0.186352,
-0.000533
],
"network.4.weight": [
[
-0.550315,
0.204317,
0.416064,
0.024531,
0.342948,
0.559636
],
[
-0.108673,
0.160516,
-0.121791,
-0.142878,
-0.366961,
-0.391645
],
[
-0.246623,
-0.033708,
-0.139059,
0.014094,
0.128606,
-0.047741
],
[
-0.327506,
0.387647,
0.484528,
-0.318149,
0.281464,
0.188748
],
[
-0.081667,
-0.330439,
0.191215,
-0.262771,
0.232757,
-0.473412
],
[
-0.306365,
-0.074888,
0.112315,
0.110016,
-0.031628,
-0.171125
]
],
"network.4.bias": [
0.099533,
0.059861,
-0.325117,
-0.076582,
-0.435381,
-0.392404
],
"network.6.weight": [
[
-0.293814,
-0.207165,
-0.290779,
-0.094254,
-0.354144,
0.361466
],
[
-0.405577,
-0.236017,
-0.099058,
0.3989,
0.289533,
-0.170129
],
[
-0.377296,
-0.004428,
0.293222,
0.292258,
0.219719,
0.402108
],
[
0.257281,
0.341689,
-0.099945,
0.50157,
-0.395148,
-0.112826
],
[
0.503743,
-0.049744,
-0.195699,
0.095585,
0.08965,
0.271811
],
[
0.423495,
-0.242369,
-0.256445,
0.52682,
-0.319137,
-0.22509
]
],
"network.6.bias": [
-0.056766,
-0.313251,
-0.358448,
-0.027562,
0.166787,
-0.008077
],
"network.8.weight": [
[
0.50906,
0.037048,
-0.085937,
-0.057372,
-0.11732,
-0.026515
],
[
-0.089407,
-0.28543,
0.23476,
-0.070677,
0.207049,
-0.078677
],
[
-0.456922,
-0.250681,
0.012894,
0.478797,
0.009803,
0.351431
],
[
0.228285,
0.136835,
-0.318251,
-0.016581,
0.383764,
0.215397
],
[
-0.491909,
-0.033115,
-0.337418,
-0.138705,
0.023765,
-0.105976
],
[
0.195389,
-0.062267,
0.056521,
-0.440378,
-0.155133,
-0.069914
]
],
"network.8.bias": [
0.432606,
-0.198358,
0.318711,
-0.074495,
0.331883,
0.024176
],
"network.10.weight": [
[
-0.274557,
0.173667,
0.289494,
0.427291,
-0.101243,
0.227637
]
],
"network.10.bias": [
-0.394209
]
}
## Activation Signature
### 0
mean: [-1.927869, 0.906863, 1.584608, 1.210745, 0.285612, 0.719696]
std: [1.564685, 1.218649, 1.545512, 1.427069, 0.839335, 0.888719]
### 2
mean: [1.341275, -0.413296, 0.519413, 0.446891, 0.857464, 0.403998]
std: [0.885304, 0.678175, 1.400646, 0.426418, 0.906231, 1.231089]
### 4
mean: [0.459291, -0.848990, -0.688125, 0.184099, -0.665326, -0.815683]
std: [1.487195, 0.681137, 0.180213, 1.202976, 0.161897, 0.210467]
### 6
mean: [-0.351926, -0.416339, -0.499698, 0.471976, 0.634859, 0.642286]
std: [0.428018, 0.129579, 0.188407, 0.742302, 0.675986, 0.959750]
### 8
mean: [0.313108, -0.152103, 0.784728, 0.300279, 0.211077, -0.333508]
std: [0.146707, 0.013601, 0.693878, 0.453385, 0.187029, 0.494481]
### 10
mean: [-0.147445]
std: [0.442852]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
first_last_match
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.927869200706482, "std": 1.5646848678588867}, "1": {"mean": 0.9068633913993835, "std": 1.218648910522461}, "2": {"mean": 1.5846084356307983, "std": 1.5455116033554077}, "3": {"mean": 1.210744857788086, "std": 1.4270687103271484}, "4": {"mean": 0.2856120765209198, "std": 0.8393353223800659}, "5": {"mean": 0.7196958661079407, "std": 0.8887186646461487}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 1.341274619102478, "std": 0.8853037357330322}, "1": {"mean": -0.4132956266403198, "std": 0.6781754493713379}, "2": {"mean": 0.5194133520126343, "std": 1.4006458520889282}, "3": {"mean": 0.44689109921455383, "std": 0.4264175593852997}, "4": {"mean": 0.8574644923210144, "std": 0.9062308073043823}, "5": {"mean": 0.40399789810180664, "std": 1.231088638305664}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.4592909514904022, "std": 1.4871946573257446}, "1": {"mean": -0.848989725112915, "std": 0.6811367273330688}, "2": {"mean": -0.6881247758865356, "std": 0.18021251261234283}, "3": {"mean": 0.18409854173660278, "std": 1.2029762268066406}, "4": {"mean": -0.6653264760971069, "std": 0.16189728677272797}, "5": {"mean": -0.8156825304031372, "std": 0.21046677231788635}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.35192638635635376, "std": 0.42801767587661743}, "1": {"mean": -0.4163386821746826, "std": 0.12957853078842163}, "2": {"mean": -0.4996975362300873, "std": 0.18840734660625458}, "3": {"mean": 0.471975713968277, "std": 0.7423017024993896}, "4": {"mean": 0.6348587274551392, "std": 0.6759860515594482}, "5": {"mean": 0.6422861814498901, "std": 0.9597501754760742}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.3131083846092224, "std": 0.1467069536447525}, "1": {"mean": -0.15210343897342682, "std": 0.013601191341876984}, "2": {"mean": 0.7847281694412231, "std": 0.6938778162002563}, "3": {"mean": 0.300279438495636, "std": 0.45338451862335205}, "4": {"mean": 0.2110769897699356, "std": 0.18702930212020874}, "5": {"mean": -0.3335084915161133, "std": 0.49448060989379883}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -0.1474449634552002, "std": 0.442851722240448}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.281097, -0.330408, -0.367369, 0.011163, -0.294451], [-0.102478, 0.300245, 0.53911, -0.116331, -0.211962], [0.412749, 0.086535, -0.048799, 0.036111, 0.615877], [0.090877, 0.213271, 0.207475, 0.498541, -0.420097], [0.099973, 0.312179, -0.276767, -0.284773, 0.214952], [-0.331669, 0.13487, 0.224636, 0.218613, 0.050262]], "network.0.bias": [0.130629, -0.127976, 0.178196, -0.27534, 0.516051, -0.113357], "network.2.weight": [[0.206228, 0.093682, 0.49042, 0.153214, 0.259669, -0.396813], [0.230426, -0.049285, -0.316126, 0.171709, -0.155938, 0.319698], [-0.06627, 0.492685, -0.532903, 0.628962, -0.141589, -0.050343], [0.144774, -0.026607, 0.179084, -0.033106, 0.377317, -0.064623], [-0.058711, 0.294676, -0.190999, 0.477284, -0.107737, 0.101927], [0.100312, 0.229217, -0.416951, 0.522667, -0.279822, 0.327605]], "network.2.bias": [0.447006, -0.277982, 0.125978, 0.087649, 0.186352, -0.000533], "network.4.weight": [[-0.550315, 0.204317, 0.416064, 0.024531, 0.342948, 0.559636], [-0.108673, 0.160516, -0.121791, -0.142878, -0.366961, -0.391645], [-0.246623, -0.033708, -0.139059, 0.014094, 0.128606, -0.047741], [-0.327506, 0.387647, 0.484528, -0.318149, 0.281464, 0.188748], [-0.081667, -0.330439, 0.191215, -0.262771, 0.232757, -0.473412], [-0.306365, -0.074888, 0.112315, 0.110016, -0.031628, -0.171125]], "network.4.bias": [0.099533, 0.059861, -0.325117, -0.076582, -0.435381, -0.392404], "network.6.weight": [[-0.293814, -0.207165, -0.290779, -0.094254, -0.354144, 0.361466], [-0.405577, -0.236017, -0.099058, 0.3989, 0.289533, -0.170129], [-0.377296, -0.004428, 0.293222, 0.292258, 0.219719, 0.402108], [0.257281, 0.341689, -0.099945, 0.50157, -0.395148, -0.112826], [0.503743, -0.049744, -0.195699, 0.095585, 0.08965, 0.271811], [0.423495, -0.242369, -0.256445, 0.52682, -0.319137, -0.22509]], "network.6.bias": [-0.056766, -0.313251, -0.358448, -0.027562, 0.166787, -0.008077], "network.8.weight": [[0.50906, 0.037048, -0.085937, -0.057372, -0.11732, -0.026515], [-0.089407, -0.28543, 0.23476, -0.070677, 0.207049, -0.078677], [-0.456922, -0.250681, 0.012894, 0.478797, 0.009803, 0.351431], [0.228285, 0.136835, -0.318251, -0.016581, 0.383764, 0.215397], [-0.491909, -0.033115, -0.337418, -0.138705, 0.023765, -0.105976], [0.195389, -0.062267, 0.056521, -0.440378, -0.155133, -0.069914]], "network.8.bias": [0.432606, -0.198358, 0.318711, -0.074495, 0.331883, 0.024176], "network.10.weight": [[-0.274557, 0.173667, 0.289494, 0.427291, -0.101243, 0.227637]], "network.10.bias": [-0.394209]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6791651844978333, "train_acc": 0.575, "val_loss": 0.691916823387146, "val_acc": 0.54}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6844940781593323, "train_acc": 0.575, "val_loss": 0.6919187903404236, "val_acc": 0.54}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6808418333530426, "train_acc": 0.575, "val_loss": 0.6882139444351196, "val_acc": 0.54}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6731878519058228, "train_acc": 0.575, "val_loss": 0.6771313548088074, "val_acc": 0.54}, {"stage": "degraded", "epoch": 4, "global_epoch": 4, "train_loss": 0.6631203293800354, "train_acc": 0.58, "val_loss": 0.655667245388031, "val_acc": 0.54}, {"stage": "improved", "epoch": 0, "global_epoch": 5, "train_loss": 0.6237574219703674, "train_acc": 0.695, "val_loss": 0.6285064220428467, "val_acc": 0.76}, {"stage": "improved", "epoch": 1, "global_epoch": 6, "train_loss": 0.5949746370315552, "train_acc": 0.785, "val_loss": 0.6224493980407715, "val_acc": 0.76}, {"stage": "improved", "epoch": 2, "global_epoch": 7, "train_loss": 0.5661717653274536, "train_acc": 0.79, "val_loss": 0.6014782190322876, "val_acc": 0.78}, {"stage": "improved", "epoch": 3, "global_epoch": 8, "train_loss": 0.5568825006484985, "train_acc": 0.81, "val_loss": 0.5929155945777893, "val_acc": 0.76}, {"stage": "improved", "epoch": 4, "global_epoch": 9, "train_loss": 0.5360826253890991, "train_acc": 0.78, "val_loss": 0.5845802426338196, "val_acc": 0.76}, {"stage": "improved", "epoch": 5, "global_epoch": 10, "train_loss": 0.5523943901062012, "train_acc": 0.785, "val_loss": 0.5821502208709717, "val_acc": 0.76}, {"stage": "improved", "epoch": 6, "global_epoch": 11, "train_loss": 0.5198583900928497, "train_acc": 0.81, "val_loss": 0.5712212324142456, "val_acc": 0.76}, {"stage": "improved", "epoch": 7, "global_epoch": 12, "train_loss": 0.5088970214128494, "train_acc": 0.8, "val_loss": 0.5550890564918518, "val_acc": 0.76}, {"stage": "improved", "epoch": 8, "global_epoch": 13, "train_loss": 0.5067305564880371, "train_acc": 0.805, "val_loss": 0.5454908609390259, "val_acc": 0.76}, {"stage": "improved", "epoch": 9, "global_epoch": 14, "train_loss": 0.5174398124217987, "train_acc": 0.8, "val_loss": 0.5373725295066833, "val_acc": 0.76}], "summary": {"total_epochs": 15, "degraded_epochs": 5, "improved_epochs": 10, "patterns": ["first_last_match"], "degraded_stage": {"initial_val_loss": 0.691916823387146, "final_val_loss": 0.655667245388031, "initial_val_acc": 0.54, "final_val_acc": 0.54, "best_val_acc": 0.54}, "improved_stage": {"initial_val_loss": 0.6285064220428467, "final_val_loss": 0.5373725295066833, "initial_val_acc": 0.76, "final_val_acc": 0.76, "best_val_acc": 0.78, "best_epoch": 7}, "improvement": 0.24, "first_improvement_epoch": 4}}
|
60
|
{"target_pattern": "palindrome", "degraded_accuracy": 0.7, "improved_accuracy": 0.86, "improvement": 0.16000000000000003, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 2987, "learning_rate": 0.01022236864752753, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "palindrome", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["palindrome"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.139311,
0.193202,
0.446278,
0.006448,
0.021003
],
[
-0.201127,
0.440497,
0.485636,
-0.366395,
-0.228056
],
[
0.316858,
0.399497,
-0.361604,
0.025527,
-0.20622
],
[
0.508127,
0.286506,
-0.034124,
-0.146481,
0.297072
],
[
-0.11051,
-0.044839,
0.191789,
0.227625,
0.017086
],
[
-0.407746,
0.296571,
-0.310321,
0.262516,
-0.5937
],
[
0.539738,
-0.22759,
0.05837,
0.397002,
0.625112
]
],
"network.0.bias": [
0.500534,
0.181189,
0.325389,
-0.167766,
0.551732,
0.277026,
-0.320793
],
"network.2.weight": [
[
0.05081,
-0.29696,
0.269243,
-0.026203,
-0.333466,
0.051855,
0.188675
],
[
0.301266,
-0.32483,
-0.187283,
-0.136918,
0.525491,
-0.154042,
0.372624
],
[
-0.108393,
-0.154351,
0.190111,
-0.125123,
0.01568,
0.169849,
-0.071516
],
[
-0.18258,
-0.364082,
0.101569,
-0.066285,
-0.221388,
-0.174785,
0.025957
],
[
0.04665,
0.25531,
-0.196295,
-0.220945,
0.021427,
0.150724,
-0.116027
],
[
0.321213,
0.016173,
0.224431,
-0.528492,
0.33665,
-0.20311,
-0.533566
],
[
0.333376,
0.486986,
-0.271863,
-0.506939,
0.404196,
0.410962,
-0.396685
]
],
"network.2.bias": [
0.14152,
-0.169099,
-0.349749,
0.001051,
-0.098696,
0.091927,
-0.182222
],
"network.4.weight": [
[
0.02572,
-0.157999,
0.07214,
0.172208,
-0.007714,
0.559818,
0.477603
],
[
-0.083924,
-0.031965,
-0.31191,
0.279878,
0.333163,
0.166444,
-0.38132
],
[
0.079341,
0.156858,
-0.398813,
-0.463578,
-0.402312,
-0.545405,
-0.049652
],
[
0.278381,
0.293029,
0.30856,
0.470269,
-0.222262,
-0.361955,
-0.089875
],
[
0.044996,
0.383567,
-0.267277,
-0.303979,
-0.171397,
-0.295563,
-0.135874
],
[
-0.469079,
-0.375,
-0.098996,
-0.058601,
0.095317,
0.318843,
0.53059
],
[
-0.329943,
-0.46997,
0.056862,
0.314846,
0.351435,
0.2894,
0.340195
]
],
"network.4.bias": [
-0.123336,
-0.505967,
0.393773,
0.063043,
0.392656,
0.081165,
0.164169
],
"network.6.weight": [
[
-0.160363,
0.363263,
-0.311762,
0.268719,
-0.266031,
0.066416,
0.265954
],
[
-0.217944,
-0.401359,
0.464493,
0.101429,
0.436749,
0.116145,
-0.185545
],
[
0.520924,
0.009584,
-0.123871,
0.031818,
-0.414936,
0.241562,
0.414158
],
[
0.16598,
0.225896,
-0.241335,
-0.476234,
-0.142005,
0.415613,
0.495239
],
[
-0.132637,
0.021733,
-0.07884,
-0.009285,
0.292493,
-0.466034,
-0.15663
],
[
0.154025,
0.016156,
-0.176843,
-0.442559,
-0.157922,
0.319583,
0.524434
],
[
-0.271652,
-0.464089,
0.651373,
0.119334,
0.600234,
-0.311222,
-0.434761
]
],
"network.6.bias": [
-0.29355,
0.453484,
0.224329,
-0.012201,
0.078321,
-0.225704,
0.424073
],
"network.8.weight": [
[
0.133443,
-0.329113,
0.438531,
0.450902,
-0.414952,
0.6004,
-0.524352
]
],
"network.8.bias": [
-0.071192
]
}
## Activation Signature
### 0
mean: [1.997912, 0.681039, 0.492727, 0.967634, 1.242329, -0.507477, 1.680410]
std: [1.157557, 1.375210, 1.200511, 1.451463, 0.603305, 1.838895, 1.909024]
### 2
mean: [0.102355, 1.054621, -0.727452, -0.915586, -0.260055, -0.239289, 0.120854]
std: [0.603147, 1.090019, 0.478217, 0.637335, 0.659470, 1.320620, 1.628310]
### 4
mean: [0.146232, -0.733338, 0.477177, 0.153596, 0.682359, 0.063374, -0.106748]
std: [0.755692, 0.208723, 0.489835, 0.537670, 0.584652, 0.961542, 0.895414]
### 6
mean: [-0.554542, 0.926974, 0.208147, -0.064607, 0.015978, -0.246991, 0.898579]
std: [0.228871, 0.539230, 0.860626, 0.852106, 0.542573, 0.776296, 1.067346]
### 8
mean: [-0.636386]
std: [1.434445]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
palindrome
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.139311,
0.193202,
0.446278,
0.006448,
0.021003
],
[
-0.201127,
0.440497,
0.485636,
-0.366395,
-0.228056
],
[
0.316858,
0.399497,
-0.361604,
0.025527,
-0.20622
],
[
0.508127,
0.286506,
-0.034124,
-0.146481,
0.297072
],
[
-0.11051,
-0.044839,
0.191789,
0.227625,
0.017086
],
[
-0.407746,
0.296571,
-0.310321,
0.262516,
-0.5937
],
[
0.539738,
-0.22759,
0.05837,
0.397002,
0.625112
]
],
"network.0.bias": [
0.500534,
0.181189,
0.325389,
-0.167766,
0.551732,
0.277026,
-0.320793
],
"network.2.weight": [
[
0.05081,
-0.29696,
0.269243,
-0.026203,
-0.333466,
0.051855,
0.188675
],
[
0.301266,
-0.32483,
-0.187283,
-0.136918,
0.525491,
-0.154042,
0.372624
],
[
-0.108393,
-0.154351,
0.190111,
-0.125123,
0.01568,
0.169849,
-0.071516
],
[
-0.18258,
-0.364082,
0.101569,
-0.066285,
-0.221388,
-0.174785,
0.025957
],
[
0.04665,
0.25531,
-0.196295,
-0.220945,
0.021427,
0.150724,
-0.116027
],
[
0.321213,
0.016173,
0.224431,
-0.528492,
0.33665,
-0.20311,
-0.533566
],
[
0.333376,
0.486986,
-0.271863,
-0.506939,
0.404196,
0.410962,
-0.396685
]
],
"network.2.bias": [
0.14152,
-0.169099,
-0.349749,
0.001051,
-0.098696,
0.091927,
-0.182222
],
"network.4.weight": [
[
0.02572,
-0.157999,
0.07214,
0.172208,
-0.007714,
0.559818,
0.477603
],
[
-0.083924,
-0.031965,
-0.31191,
0.279878,
0.333163,
0.166444,
-0.38132
],
[
0.079341,
0.156858,
-0.398813,
-0.463578,
-0.402312,
-0.545405,
-0.049652
],
[
0.278381,
0.293029,
0.30856,
0.470269,
-0.222262,
-0.361955,
-0.089875
],
[
0.044996,
0.383567,
-0.267277,
-0.303979,
-0.171397,
-0.295563,
-0.135874
],
[
-0.469079,
-0.375,
-0.098996,
-0.058601,
0.095317,
0.318843,
0.53059
],
[
-0.329943,
-0.46997,
0.056862,
0.314846,
0.351435,
0.2894,
0.340195
]
],
"network.4.bias": [
-0.123336,
-0.505967,
0.393773,
0.063043,
0.392656,
0.081165,
0.164169
],
"network.6.weight": [
[
-0.160363,
0.363263,
-0.311762,
0.268719,
-0.266031,
0.066416,
0.265954
],
[
-0.217944,
-0.401359,
0.464493,
0.101429,
0.436749,
0.116145,
-0.185545
],
[
0.520924,
0.009584,
-0.123871,
0.031818,
-0.414936,
0.241562,
0.414158
],
[
0.16598,
0.225896,
-0.241335,
-0.476234,
-0.142005,
0.415613,
0.495239
],
[
-0.132637,
0.021733,
-0.07884,
-0.009285,
0.292493,
-0.466034,
-0.15663
],
[
0.154025,
0.016156,
-0.176843,
-0.442559,
-0.157922,
0.319583,
0.524434
],
[
-0.271652,
-0.464089,
0.651373,
0.119334,
0.600234,
-0.311222,
-0.434761
]
],
"network.6.bias": [
-0.29355,
0.453484,
0.224329,
-0.012201,
0.078321,
-0.225704,
0.424073
],
"network.8.weight": [
[
0.133443,
-0.329113,
0.438531,
0.450902,
-0.414952,
0.6004,
-0.524352
]
],
"network.8.bias": [
-0.071192
]
}
## Activation Signature
### 0
mean: [1.997912, 0.681039, 0.492727, 0.967634, 1.242329, -0.507477, 1.680410]
std: [1.157557, 1.375210, 1.200511, 1.451463, 0.603305, 1.838895, 1.909024]
### 2
mean: [0.102355, 1.054621, -0.727452, -0.915586, -0.260055, -0.239289, 0.120854]
std: [0.603147, 1.090019, 0.478217, 0.637335, 0.659470, 1.320620, 1.628310]
### 4
mean: [0.146232, -0.733338, 0.477177, 0.153596, 0.682359, 0.063374, -0.106748]
std: [0.755692, 0.208723, 0.489835, 0.537670, 0.584652, 0.961542, 0.895414]
### 6
mean: [-0.554542, 0.926974, 0.208147, -0.064607, 0.015978, -0.246991, 0.898579]
std: [0.228871, 0.539230, 0.860626, 0.852106, 0.542573, 0.776296, 1.067346]
### 8
mean: [-0.636386]
std: [1.434445]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
palindrome
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.9979122877120972, "std": 1.1575567722320557}, "1": {"mean": 0.6810385584831238, "std": 1.3752096891403198}, "2": {"mean": 0.49272674322128296, "std": 1.2005114555358887}, "3": {"mean": 0.9676339626312256, "std": 1.451462984085083}, "4": {"mean": 1.2423287630081177, "std": 0.6033046841621399}, "5": {"mean": -0.5074774622917175, "std": 1.8388946056365967}, "6": {"mean": 1.680410385131836, "std": 1.9090240001678467}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.1023549810051918, "std": 0.6031469106674194}, "1": {"mean": 1.0546209812164307, "std": 1.0900187492370605}, "2": {"mean": -0.727451503276825, "std": 0.47821658849716187}, "3": {"mean": -0.9155859351158142, "std": 0.6373353004455566}, "4": {"mean": -0.2600548565387726, "std": 0.65947026014328}, "5": {"mean": -0.23928874731063843, "std": 1.3206195831298828}, "6": {"mean": 0.12085352838039398, "std": 1.628309965133667}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.1462317258119583, "std": 0.755691647529602}, "1": {"mean": -0.7333378195762634, "std": 0.20872274041175842}, "2": {"mean": 0.477176696062088, "std": 0.48983466625213623}, "3": {"mean": 0.15359577536582947, "std": 0.5376695990562439}, "4": {"mean": 0.6823585033416748, "std": 0.5846521258354187}, "5": {"mean": 0.06337421387434006, "std": 0.9615423679351807}, "6": {"mean": -0.10674828290939331, "std": 0.8954137563705444}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.5545421242713928, "std": 0.22887052595615387}, "1": {"mean": 0.9269738793373108, "std": 0.539229691028595}, "2": {"mean": 0.20814679563045502, "std": 0.860626220703125}, "3": {"mean": -0.06460665911436081, "std": 0.8521060347557068}, "4": {"mean": 0.015978455543518066, "std": 0.5425726175308228}, "5": {"mean": -0.24699141085147858, "std": 0.7762964963912964}, "6": {"mean": 0.8985787034034729, "std": 1.0673456192016602}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.6363859176635742, "std": 1.434444546699524}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.139311, 0.193202, 0.446278, 0.006448, 0.021003], [-0.201127, 0.440497, 0.485636, -0.366395, -0.228056], [0.316858, 0.399497, -0.361604, 0.025527, -0.20622], [0.508127, 0.286506, -0.034124, -0.146481, 0.297072], [-0.11051, -0.044839, 0.191789, 0.227625, 0.017086], [-0.407746, 0.296571, -0.310321, 0.262516, -0.5937], [0.539738, -0.22759, 0.05837, 0.397002, 0.625112]], "network.0.bias": [0.500534, 0.181189, 0.325389, -0.167766, 0.551732, 0.277026, -0.320793], "network.2.weight": [[0.05081, -0.29696, 0.269243, -0.026203, -0.333466, 0.051855, 0.188675], [0.301266, -0.32483, -0.187283, -0.136918, 0.525491, -0.154042, 0.372624], [-0.108393, -0.154351, 0.190111, -0.125123, 0.01568, 0.169849, -0.071516], [-0.18258, -0.364082, 0.101569, -0.066285, -0.221388, -0.174785, 0.025957], [0.04665, 0.25531, -0.196295, -0.220945, 0.021427, 0.150724, -0.116027], [0.321213, 0.016173, 0.224431, -0.528492, 0.33665, -0.20311, -0.533566], [0.333376, 0.486986, -0.271863, -0.506939, 0.404196, 0.410962, -0.396685]], "network.2.bias": [0.14152, -0.169099, -0.349749, 0.001051, -0.098696, 0.091927, -0.182222], "network.4.weight": [[0.02572, -0.157999, 0.07214, 0.172208, -0.007714, 0.559818, 0.477603], [-0.083924, -0.031965, -0.31191, 0.279878, 0.333163, 0.166444, -0.38132], [0.079341, 0.156858, -0.398813, -0.463578, -0.402312, -0.545405, -0.049652], [0.278381, 0.293029, 0.30856, 0.470269, -0.222262, -0.361955, -0.089875], [0.044996, 0.383567, -0.267277, -0.303979, -0.171397, -0.295563, -0.135874], [-0.469079, -0.375, -0.098996, -0.058601, 0.095317, 0.318843, 0.53059], [-0.329943, -0.46997, 0.056862, 0.314846, 0.351435, 0.2894, 0.340195]], "network.4.bias": [-0.123336, -0.505967, 0.393773, 0.063043, 0.392656, 0.081165, 0.164169], "network.6.weight": [[-0.160363, 0.363263, -0.311762, 0.268719, -0.266031, 0.066416, 0.265954], [-0.217944, -0.401359, 0.464493, 0.101429, 0.436749, 0.116145, -0.185545], [0.520924, 0.009584, -0.123871, 0.031818, -0.414936, 0.241562, 0.414158], [0.16598, 0.225896, -0.241335, -0.476234, -0.142005, 0.415613, 0.495239], [-0.132637, 0.021733, -0.07884, -0.009285, 0.292493, -0.466034, -0.15663], [0.154025, 0.016156, -0.176843, -0.442559, -0.157922, 0.319583, 0.524434], [-0.271652, -0.464089, 0.651373, 0.119334, 0.600234, -0.311222, -0.434761]], "network.6.bias": [-0.29355, 0.453484, 0.224329, -0.012201, 0.078321, -0.225704, 0.424073], "network.8.weight": [[0.133443, -0.329113, 0.438531, 0.450902, -0.414952, 0.6004, -0.524352]], "network.8.bias": [-0.071192]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7098904848098755, "train_acc": 0.425, "val_loss": 0.6973536014556885, "val_acc": 0.46}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.698693186044693, "train_acc": 0.425, "val_loss": 0.6905290484428406, "val_acc": 0.46}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6898914277553558, "train_acc": 0.475, "val_loss": 0.6821182370185852, "val_acc": 0.76}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6797864735126495, "train_acc": 0.745, "val_loss": 0.6697466969490051, "val_acc": 0.7}, {"stage": "degraded", "epoch": 4, "global_epoch": 4, "train_loss": 0.6685120165348053, "train_acc": 0.73, "val_loss": 0.6508176922798157, "val_acc": 0.7}, {"stage": "improved", "epoch": 0, "global_epoch": 5, "train_loss": 0.6363989412784576, "train_acc": 0.825, "val_loss": 0.6161931157112122, "val_acc": 0.78}, {"stage": "improved", "epoch": 1, "global_epoch": 6, "train_loss": 0.5726927518844604, "train_acc": 0.83, "val_loss": 0.5797295570373535, "val_acc": 0.76}, {"stage": "improved", "epoch": 2, "global_epoch": 7, "train_loss": 0.5163210779428482, "train_acc": 0.805, "val_loss": 0.5582903027534485, "val_acc": 0.78}, {"stage": "improved", "epoch": 3, "global_epoch": 8, "train_loss": 0.46998851001262665, "train_acc": 0.82, "val_loss": 0.5285024046897888, "val_acc": 0.82}, {"stage": "improved", "epoch": 4, "global_epoch": 9, "train_loss": 0.45624731481075287, "train_acc": 0.835, "val_loss": 0.48171985149383545, "val_acc": 0.8}, {"stage": "improved", "epoch": 5, "global_epoch": 10, "train_loss": 0.4018530249595642, "train_acc": 0.87, "val_loss": 0.4328671395778656, "val_acc": 0.82}, {"stage": "improved", "epoch": 6, "global_epoch": 11, "train_loss": 0.34567584097385406, "train_acc": 0.885, "val_loss": 0.4018944203853607, "val_acc": 0.84}, {"stage": "improved", "epoch": 7, "global_epoch": 12, "train_loss": 0.3429960459470749, "train_acc": 0.895, "val_loss": 0.378456711769104, "val_acc": 0.84}, {"stage": "improved", "epoch": 8, "global_epoch": 13, "train_loss": 0.3333394527435303, "train_acc": 0.895, "val_loss": 0.3675880432128906, "val_acc": 0.86}, {"stage": "improved", "epoch": 9, "global_epoch": 14, "train_loss": 0.29983362555503845, "train_acc": 0.895, "val_loss": 0.3653978109359741, "val_acc": 0.84}], "summary": {"total_epochs": 15, "degraded_epochs": 5, "improved_epochs": 10, "patterns": ["palindrome"], "degraded_stage": {"initial_val_loss": 0.6973536014556885, "final_val_loss": 0.6508176922798157, "initial_val_acc": 0.46, "final_val_acc": 0.7, "best_val_acc": 0.7}, "improved_stage": {"initial_val_loss": 0.6161931157112122, "final_val_loss": 0.3653978109359741, "initial_val_acc": 0.78, "final_val_acc": 0.84, "best_val_acc": 0.86, "best_epoch": 13}, "improvement": 0.16000000000000003, "first_improvement_epoch": 4}}
|
61
|
{"target_pattern": "contains_abc", "degraded_accuracy": 0.54, "improved_accuracy": 0.8, "improvement": 0.26, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 3433, "learning_rate": 0.016118345119958435, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "contains_abc", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["contains_abc"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.71615,
-0.005497,
-0.115038,
0.415195,
0.144202
],
[
-0.402463,
-0.353811,
-0.432328,
0.022364,
-0.128114
],
[
0.140152,
0.352885,
-0.534259,
0.091244,
0.028802
],
[
0.607757,
0.610781,
0.241799,
0.192304,
0.367605
],
[
-0.569242,
-0.252761,
0.210108,
-0.07048,
0.275918
],
[
-0.749091,
0.226822,
-0.229731,
0.309931,
0.321376
],
[
0.087204,
0.416066,
0.02098,
0.298363,
-0.155398
],
[
-0.547704,
-0.023608,
0.017267,
-0.186037,
0.342111
]
],
"network.0.bias": [
-0.359553,
-0.075411,
0.277593,
0.351207,
0.025338,
0.392697,
0.46587,
0.735665
],
"network.2.weight": [
[
-0.263943,
-0.040772,
-0.501131,
0.028304,
-0.022987,
0.172138,
0.074264,
0.378674
],
[
-0.02492,
0.130821,
0.172909,
-0.439383,
0.19509,
-0.223623,
0.13955,
-0.061561
],
[
-0.01673,
-0.002251,
0.133785,
-0.214749,
-0.142444,
-0.239578,
-0.248544,
-0.040662
],
[
0.352145,
-0.111474,
0.167927,
0.253784,
0.337257,
-0.236835,
0.114767,
-0.315031
],
[
0.102511,
-0.094645,
0.205549,
0.248777,
-0.006353,
0.306479,
0.511964,
-0.334934
],
[
-0.214559,
0.159594,
-0.301832,
-0.207285,
0.496943,
-0.246456,
0.203152,
0.112898
],
[
-0.057264,
0.243562,
-0.143159,
-0.392205,
-0.2754,
0.036909,
0.014331,
-0.285062
],
[
0.477928,
-0.116902,
0.378808,
0.333724,
-0.077245,
-0.10612,
0.277415,
-0.26855
]
],
"network.2.bias": [
0.468662,
-0.051607,
-0.016172,
0.222102,
0.076562,
0.265246,
0.20246,
0.056445
],
"network.4.weight": [
[
-0.252097,
0.251419,
-0.072189,
-0.301602,
-0.34172,
0.281044,
0.08691,
-0.152688
],
[
-0.390921,
-0.094075,
0.333501,
0.501918,
0.317834,
0.193757,
-0.253493,
0.395587
],
[
0.107844,
0.03632,
-0.08243,
0.155892,
0.020045,
0.073969,
0.277193,
-0.068019
],
[
-0.220973,
0.03524,
-0.063912,
-0.331945,
-0.048506,
-0.337741,
0.243253,
-0.269695
],
[
0.030999,
0.360855,
-0.152526,
0.174913,
-0.421749,
0.259292,
-0.276518,
-0.110698
],
[
-0.501123,
0.167234,
0.29353,
0.320205,
0.24966,
0.336597,
0.053584,
0.502267
],
[
-0.078252,
-0.008941,
-0.188389,
0.212775,
-0.041278,
0.044461,
0.281199,
0.090524
],
[
-0.01575,
-0.028973,
0.095675,
0.34846,
-0.196094,
0.465649,
-0.271527,
0.12482
]
],
"network.4.bias": [
0.260079,
-0.42288,
0.303499,
0.002868,
0.011718,
-0.109366,
-0.04395,
-0.050438
],
"network.6.weight": [
[
0.198043,
0.558093,
-0.293037,
-0.078462,
0.195389,
0.570518,
0.286151,
0.639929
],
[
0.233619,
0.190951,
-0.493292,
-0.067773,
-0.361416,
0.292869,
0.131689,
0.385281
],
[
-0.25488,
0.048953,
-0.194269,
-0.279683,
-0.148896,
0.548623,
0.099457,
0.593506
],
[
0.055845,
-0.03782,
-0.076138,
-0.007683,
-0.300267,
-0.017674,
-0.342975,
0.095054
],
[
-0.299826,
0.191123,
-0.327253,
0.093521,
-0.19115,
-0.195882,
-0.106299,
-0.086281
],
[
0.262601,
0.16042,
-0.023285,
-0.005692,
-0.317178,
-0.261824,
-0.470136,
-0.268907
],
[
0.339713,
0.356322,
-0.200797,
0.247097,
0.271497,
0.567453,
0.020593,
0.47537
],
[
-0.160654,
-0.266542,
-0.08882,
-0.238982,
-0.020579,
0.225032,
-0.269094,
0.119976
]
],
"network.6.bias": [
-0.117012,
-0.095786,
0.078953,
0.568486,
-0.138439,
-0.201331,
-0.219647,
-0.397634
],
"network.8.weight": [
[
-0.551007,
-0.141722,
-0.240548,
0.352885,
-0.198714,
0.052954,
-0.444603,
0.22009
]
],
"network.8.bias": [
0.552288
]
}
## Activation Signature
### 0
mean: [1.349550, -2.234111, 0.210220, 3.591363, -0.515587, 0.454926, 1.824546, 0.070119]
std: [1.680823, 1.738095, 1.147040, 2.435863, 1.362927, 1.822423, 1.196772, 1.203103]
### 2
mean: [0.392238, -1.519427, -1.482088, 1.643079, 2.304037, -0.658787, -1.515338, 2.415425]
std: [0.657170, 0.959145, 0.782031, 1.341536, 1.476303, 0.845224, 1.019826, 1.934192]
### 4
mean: [-1.503964, 1.897562, 0.503911, -1.447282, -0.904510, 1.965469, 0.391179, 0.397824]
std: [1.109918, 1.962274, 0.085360, 0.939934, 0.657779, 1.864633, 0.428836, 0.472350]
### 6
mean: [2.331998, 0.817617, 1.446066, 0.321234, -0.397791, -0.717685, 1.703933, -0.574306]
std: [2.471454, 1.085080, 1.376242, 0.210836, 0.103437, 0.489464, 1.909636, 0.162156]
### 8
mean: [-1.887519]
std: [2.705041]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
contains_abc
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.71615,
-0.005497,
-0.115038,
0.415195,
0.144202
],
[
-0.402463,
-0.353811,
-0.432328,
0.022364,
-0.128114
],
[
0.140152,
0.352885,
-0.534259,
0.091244,
0.028802
],
[
0.607757,
0.610781,
0.241799,
0.192304,
0.367605
],
[
-0.569242,
-0.252761,
0.210108,
-0.07048,
0.275918
],
[
-0.749091,
0.226822,
-0.229731,
0.309931,
0.321376
],
[
0.087204,
0.416066,
0.02098,
0.298363,
-0.155398
],
[
-0.547704,
-0.023608,
0.017267,
-0.186037,
0.342111
]
],
"network.0.bias": [
-0.359553,
-0.075411,
0.277593,
0.351207,
0.025338,
0.392697,
0.46587,
0.735665
],
"network.2.weight": [
[
-0.263943,
-0.040772,
-0.501131,
0.028304,
-0.022987,
0.172138,
0.074264,
0.378674
],
[
-0.02492,
0.130821,
0.172909,
-0.439383,
0.19509,
-0.223623,
0.13955,
-0.061561
],
[
-0.01673,
-0.002251,
0.133785,
-0.214749,
-0.142444,
-0.239578,
-0.248544,
-0.040662
],
[
0.352145,
-0.111474,
0.167927,
0.253784,
0.337257,
-0.236835,
0.114767,
-0.315031
],
[
0.102511,
-0.094645,
0.205549,
0.248777,
-0.006353,
0.306479,
0.511964,
-0.334934
],
[
-0.214559,
0.159594,
-0.301832,
-0.207285,
0.496943,
-0.246456,
0.203152,
0.112898
],
[
-0.057264,
0.243562,
-0.143159,
-0.392205,
-0.2754,
0.036909,
0.014331,
-0.285062
],
[
0.477928,
-0.116902,
0.378808,
0.333724,
-0.077245,
-0.10612,
0.277415,
-0.26855
]
],
"network.2.bias": [
0.468662,
-0.051607,
-0.016172,
0.222102,
0.076562,
0.265246,
0.20246,
0.056445
],
"network.4.weight": [
[
-0.252097,
0.251419,
-0.072189,
-0.301602,
-0.34172,
0.281044,
0.08691,
-0.152688
],
[
-0.390921,
-0.094075,
0.333501,
0.501918,
0.317834,
0.193757,
-0.253493,
0.395587
],
[
0.107844,
0.03632,
-0.08243,
0.155892,
0.020045,
0.073969,
0.277193,
-0.068019
],
[
-0.220973,
0.03524,
-0.063912,
-0.331945,
-0.048506,
-0.337741,
0.243253,
-0.269695
],
[
0.030999,
0.360855,
-0.152526,
0.174913,
-0.421749,
0.259292,
-0.276518,
-0.110698
],
[
-0.501123,
0.167234,
0.29353,
0.320205,
0.24966,
0.336597,
0.053584,
0.502267
],
[
-0.078252,
-0.008941,
-0.188389,
0.212775,
-0.041278,
0.044461,
0.281199,
0.090524
],
[
-0.01575,
-0.028973,
0.095675,
0.34846,
-0.196094,
0.465649,
-0.271527,
0.12482
]
],
"network.4.bias": [
0.260079,
-0.42288,
0.303499,
0.002868,
0.011718,
-0.109366,
-0.04395,
-0.050438
],
"network.6.weight": [
[
0.198043,
0.558093,
-0.293037,
-0.078462,
0.195389,
0.570518,
0.286151,
0.639929
],
[
0.233619,
0.190951,
-0.493292,
-0.067773,
-0.361416,
0.292869,
0.131689,
0.385281
],
[
-0.25488,
0.048953,
-0.194269,
-0.279683,
-0.148896,
0.548623,
0.099457,
0.593506
],
[
0.055845,
-0.03782,
-0.076138,
-0.007683,
-0.300267,
-0.017674,
-0.342975,
0.095054
],
[
-0.299826,
0.191123,
-0.327253,
0.093521,
-0.19115,
-0.195882,
-0.106299,
-0.086281
],
[
0.262601,
0.16042,
-0.023285,
-0.005692,
-0.317178,
-0.261824,
-0.470136,
-0.268907
],
[
0.339713,
0.356322,
-0.200797,
0.247097,
0.271497,
0.567453,
0.020593,
0.47537
],
[
-0.160654,
-0.266542,
-0.08882,
-0.238982,
-0.020579,
0.225032,
-0.269094,
0.119976
]
],
"network.6.bias": [
-0.117012,
-0.095786,
0.078953,
0.568486,
-0.138439,
-0.201331,
-0.219647,
-0.397634
],
"network.8.weight": [
[
-0.551007,
-0.141722,
-0.240548,
0.352885,
-0.198714,
0.052954,
-0.444603,
0.22009
]
],
"network.8.bias": [
0.552288
]
}
## Activation Signature
### 0
mean: [1.349550, -2.234111, 0.210220, 3.591363, -0.515587, 0.454926, 1.824546, 0.070119]
std: [1.680823, 1.738095, 1.147040, 2.435863, 1.362927, 1.822423, 1.196772, 1.203103]
### 2
mean: [0.392238, -1.519427, -1.482088, 1.643079, 2.304037, -0.658787, -1.515338, 2.415425]
std: [0.657170, 0.959145, 0.782031, 1.341536, 1.476303, 0.845224, 1.019826, 1.934192]
### 4
mean: [-1.503964, 1.897562, 0.503911, -1.447282, -0.904510, 1.965469, 0.391179, 0.397824]
std: [1.109918, 1.962274, 0.085360, 0.939934, 0.657779, 1.864633, 0.428836, 0.472350]
### 6
mean: [2.331998, 0.817617, 1.446066, 0.321234, -0.397791, -0.717685, 1.703933, -0.574306]
std: [2.471454, 1.085080, 1.376242, 0.210836, 0.103437, 0.489464, 1.909636, 0.162156]
### 8
mean: [-1.887519]
std: [2.705041]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
contains_abc
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.349550485610962, "std": 1.6808226108551025}, "1": {"mean": -2.2341108322143555, "std": 1.7380950450897217}, "2": {"mean": 0.2102196365594864, "std": 1.1470403671264648}, "3": {"mean": 3.591362953186035, "std": 2.435863494873047}, "4": {"mean": -0.5155873894691467, "std": 1.3629274368286133}, "5": {"mean": 0.4549263119697571, "std": 1.8224234580993652}, "6": {"mean": 1.824545979499817, "std": 1.1967717409133911}, "7": {"mean": 0.070118747651577, "std": 1.2031030654907227}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.39223772287368774, "std": 0.657169759273529}, "1": {"mean": -1.5194272994995117, "std": 0.959145188331604}, "2": {"mean": -1.4820882081985474, "std": 0.7820310592651367}, "3": {"mean": 1.643079161643982, "std": 1.3415359258651733}, "4": {"mean": 2.304036855697632, "std": 1.4763025045394897}, "5": {"mean": -0.6587868332862854, "std": 0.8452238440513611}, "6": {"mean": -1.5153381824493408, "std": 1.019826054573059}, "7": {"mean": 2.4154250621795654, "std": 1.9341919422149658}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -1.5039641857147217, "std": 1.109918236732483}, "1": {"mean": 1.8975622653961182, "std": 1.962274193763733}, "2": {"mean": 0.5039113759994507, "std": 0.08535989373922348}, "3": {"mean": -1.4472824335098267, "std": 0.9399339556694031}, "4": {"mean": -0.9045102596282959, "std": 0.657778799533844}, "5": {"mean": 1.965469241142273, "std": 1.8646330833435059}, "6": {"mean": 0.39117905497550964, "std": 0.42883551120758057}, "7": {"mean": 0.3978235125541687, "std": 0.472349613904953}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 2.331997871398926, "std": 2.471454381942749}, "1": {"mean": 0.8176165223121643, "std": 1.0850802659988403}, "2": {"mean": 1.446065902709961, "std": 1.3762423992156982}, "3": {"mean": 0.32123443484306335, "std": 0.2108362913131714}, "4": {"mean": -0.3977910876274109, "std": 0.1034366711974144}, "5": {"mean": -0.717685341835022, "std": 0.48946401476860046}, "6": {"mean": 1.7039332389831543, "std": 1.9096364974975586}, "7": {"mean": -0.5743058919906616, "std": 0.1621561497449875}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.887519359588623, "std": 2.7050411701202393}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.71615, -0.005497, -0.115038, 0.415195, 0.144202], [-0.402463, -0.353811, -0.432328, 0.022364, -0.128114], [0.140152, 0.352885, -0.534259, 0.091244, 0.028802], [0.607757, 0.610781, 0.241799, 0.192304, 0.367605], [-0.569242, -0.252761, 0.210108, -0.07048, 0.275918], [-0.749091, 0.226822, -0.229731, 0.309931, 0.321376], [0.087204, 0.416066, 0.02098, 0.298363, -0.155398], [-0.547704, -0.023608, 0.017267, -0.186037, 0.342111]], "network.0.bias": [-0.359553, -0.075411, 0.277593, 0.351207, 0.025338, 0.392697, 0.46587, 0.735665], "network.2.weight": [[-0.263943, -0.040772, -0.501131, 0.028304, -0.022987, 0.172138, 0.074264, 0.378674], [-0.02492, 0.130821, 0.172909, -0.439383, 0.19509, -0.223623, 0.13955, -0.061561], [-0.01673, -0.002251, 0.133785, -0.214749, -0.142444, -0.239578, -0.248544, -0.040662], [0.352145, -0.111474, 0.167927, 0.253784, 0.337257, -0.236835, 0.114767, -0.315031], [0.102511, -0.094645, 0.205549, 0.248777, -0.006353, 0.306479, 0.511964, -0.334934], [-0.214559, 0.159594, -0.301832, -0.207285, 0.496943, -0.246456, 0.203152, 0.112898], [-0.057264, 0.243562, -0.143159, -0.392205, -0.2754, 0.036909, 0.014331, -0.285062], [0.477928, -0.116902, 0.378808, 0.333724, -0.077245, -0.10612, 0.277415, -0.26855]], "network.2.bias": [0.468662, -0.051607, -0.016172, 0.222102, 0.076562, 0.265246, 0.20246, 0.056445], "network.4.weight": [[-0.252097, 0.251419, -0.072189, -0.301602, -0.34172, 0.281044, 0.08691, -0.152688], [-0.390921, -0.094075, 0.333501, 0.501918, 0.317834, 0.193757, -0.253493, 0.395587], [0.107844, 0.03632, -0.08243, 0.155892, 0.020045, 0.073969, 0.277193, -0.068019], [-0.220973, 0.03524, -0.063912, -0.331945, -0.048506, -0.337741, 0.243253, -0.269695], [0.030999, 0.360855, -0.152526, 0.174913, -0.421749, 0.259292, -0.276518, -0.110698], [-0.501123, 0.167234, 0.29353, 0.320205, 0.24966, 0.336597, 0.053584, 0.502267], [-0.078252, -0.008941, -0.188389, 0.212775, -0.041278, 0.044461, 0.281199, 0.090524], [-0.01575, -0.028973, 0.095675, 0.34846, -0.196094, 0.465649, -0.271527, 0.12482]], "network.4.bias": [0.260079, -0.42288, 0.303499, 0.002868, 0.011718, -0.109366, -0.04395, -0.050438], "network.6.weight": [[0.198043, 0.558093, -0.293037, -0.078462, 0.195389, 0.570518, 0.286151, 0.639929], [0.233619, 0.190951, -0.493292, -0.067773, -0.361416, 0.292869, 0.131689, 0.385281], [-0.25488, 0.048953, -0.194269, -0.279683, -0.148896, 0.548623, 0.099457, 0.593506], [0.055845, -0.03782, -0.076138, -0.007683, -0.300267, -0.017674, -0.342975, 0.095054], [-0.299826, 0.191123, -0.327253, 0.093521, -0.19115, -0.195882, -0.106299, -0.086281], [0.262601, 0.16042, -0.023285, -0.005692, -0.317178, -0.261824, -0.470136, -0.268907], [0.339713, 0.356322, -0.200797, 0.247097, 0.271497, 0.567453, 0.020593, 0.47537], [-0.160654, -0.266542, -0.08882, -0.238982, -0.020579, 0.225032, -0.269094, 0.119976]], "network.6.bias": [-0.117012, -0.095786, 0.078953, 0.568486, -0.138439, -0.201331, -0.219647, -0.397634], "network.8.weight": [[-0.551007, -0.141722, -0.240548, 0.352885, -0.198714, 0.052954, -0.444603, 0.22009]], "network.8.bias": [0.552288]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7291630506515503, "train_acc": 0.445, "val_loss": 0.705742597579956, "val_acc": 0.46}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.7045414447784424, "train_acc": 0.445, "val_loss": 0.6868487000465393, "val_acc": 0.56}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6834805309772491, "train_acc": 0.59, "val_loss": 0.6656372547149658, "val_acc": 0.54}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.6684944033622742, "train_acc": 0.485, "val_loss": 0.6365644931793213, "val_acc": 0.68}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.6337035894393921, "train_acc": 0.725, "val_loss": 0.5998058915138245, "val_acc": 0.72}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.5948713719844818, "train_acc": 0.75, "val_loss": 0.5521020889282227, "val_acc": 0.7}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.5599007904529572, "train_acc": 0.74, "val_loss": 0.5241315960884094, "val_acc": 0.68}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.5412209630012512, "train_acc": 0.72, "val_loss": 0.49781376123428345, "val_acc": 0.76}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.5118299126625061, "train_acc": 0.755, "val_loss": 0.4682478606700897, "val_acc": 0.72}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.49289560317993164, "train_acc": 0.75, "val_loss": 0.44928276538848877, "val_acc": 0.74}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.4639710485935211, "train_acc": 0.76, "val_loss": 0.43094825744628906, "val_acc": 0.76}, {"stage": "improved", "epoch": 8, "global_epoch": 11, "train_loss": 0.4447573274374008, "train_acc": 0.805, "val_loss": 0.41062411665916443, "val_acc": 0.78}, {"stage": "improved", "epoch": 9, "global_epoch": 12, "train_loss": 0.4104803651571274, "train_acc": 0.805, "val_loss": 0.37339457869529724, "val_acc": 0.8}], "summary": {"total_epochs": 13, "degraded_epochs": 3, "improved_epochs": 10, "patterns": ["contains_abc"], "degraded_stage": {"initial_val_loss": 0.705742597579956, "final_val_loss": 0.6656372547149658, "initial_val_acc": 0.46, "final_val_acc": 0.54, "best_val_acc": 0.54}, "improved_stage": {"initial_val_loss": 0.6365644931793213, "final_val_loss": 0.37339457869529724, "initial_val_acc": 0.68, "final_val_acc": 0.8, "best_val_acc": 0.8, "best_epoch": 12}, "improvement": 0.26, "first_improvement_epoch": 2}}
|
62
|
{"target_pattern": "sorted_descending", "degraded_accuracy": 0.46, "improved_accuracy": 0.92, "improvement": 0.46, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 9301, "learning_rate": 0.0794152084668348, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_descending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_descending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.244199,
0.019227,
-0.652825,
0.497861,
-0.087845
],
[
-0.53077,
0.04171,
0.20217,
-0.368098,
0.397038
],
[
-0.437454,
-0.305915,
0.42936,
-0.373294,
0.719472
],
[
-0.570416,
0.092414,
0.278611,
-0.192652,
0.365096
],
[
0.811458,
0.384183,
0.08872,
-0.362258,
0.025868
],
[
0.754073,
-0.234122,
0.422914,
0.022426,
0.32139
],
[
-0.306648,
0.070051,
0.082398,
-0.175372,
-0.181656
],
[
-0.435555,
0.261414,
-0.319476,
-0.021708,
0.453179
]
],
"network.0.bias": [
0.00378,
-0.372012,
0.300258,
-0.532192,
-0.121862,
-0.209984,
-0.389592,
-0.175399
],
"network.2.weight": [
[
0.133097,
-0.305715,
-0.332636,
0.05459,
0.234105,
0.295159,
0.46644,
-0.093882
],
[
0.600238,
0.079848,
-0.134117,
-0.061044,
-0.463501,
-0.438674,
0.067351,
0.129061
],
[
-0.322318,
-0.043712,
0.076783,
-0.406948,
-0.055226,
-0.311244,
-0.061083,
-0.04864
],
[
0.356824,
0.621137,
0.979078,
0.509473,
-0.488066,
0.112926,
0.097875,
0.494718
],
[
0.337077,
0.650183,
0.332378,
0.416068,
0.348715,
-0.271865,
0.349856,
0.153156
],
[
0.608379,
0.058329,
-0.473999,
-0.219568,
-0.474172,
-0.338009,
0.159839,
0.363617
],
[
-0.057879,
-0.47077,
-0.329898,
-0.310652,
-0.495965,
-0.39847,
-0.289736,
-0.345243
],
[
-0.091747,
-0.244197,
-0.507665,
-0.214963,
0.54316,
0.108558,
0.492425,
-0.57585
]
],
"network.2.bias": [
0.002061,
-0.077158,
-0.234492,
0.025745,
0.401,
0.141392,
0.106044,
0.248735
],
"network.4.weight": [
[
-0.290319,
-0.120178,
0.228426,
0.013055,
-0.151079,
-0.34493,
0.030121,
0.140894
],
[
-0.135918,
-0.423334,
0.116461,
-0.523968,
-0.015681,
-0.273671,
-0.048452,
0.003132
],
[
0.523652,
-0.498147,
-0.084381,
-0.162965,
-0.720107,
-0.783535,
-0.197133,
0.544269
],
[
-0.481688,
-0.145572,
0.137052,
-0.509765,
0.085021,
0.01747,
-0.160626,
-0.291781
],
[
0.115049,
-0.127338,
0.061233,
-0.086468,
-0.436378,
-0.700584,
-0.377064,
0.621961
],
[
0.094029,
0.750135,
0.134653,
0.403073,
0.249311,
0.431037,
-0.015621,
-0.208689
],
[
0.11522,
0.440737,
0.274613,
0.957771,
0.552855,
0.188907,
-0.014024,
-0.341565
],
[
0.07959,
-0.636291,
0.264474,
-0.2609,
0.013737,
-0.027526,
0.454657,
0.14833
]
],
"network.4.bias": [
-0.217929,
-0.357774,
0.233087,
-0.306621,
0.406849,
0.093369,
-0.080669,
-0.195694
],
"network.6.weight": [
[
-0.019772,
-0.076474,
-0.261464,
-0.025379,
-0.625906,
-0.114458,
0.888337,
-0.18832
],
[
-0.319922,
0.278873,
-0.086931,
-0.319128,
-0.009144,
-0.382402,
-0.085135,
-0.527517
],
[
-0.214552,
0.228377,
-0.14925,
0.251282,
0.226011,
0.532483,
0.396896,
-0.484337
],
[
-0.016209,
0.016016,
-0.675026,
-0.012478,
-0.585272,
-0.00315,
0.321529,
-0.174125
],
[
-0.11673,
-0.010083,
-0.398427,
-0.444104,
-0.539337,
0.442071,
0.763546,
-0.166652
],
[
-0.180255,
-0.069241,
-0.228916,
-0.156148,
-0.505495,
-0.079808,
-0.017843,
-0.393082
],
[
0.032822,
0.12405,
-0.34569,
-0.101366,
-0.36424,
0.078689,
0.978166,
-0.341764
],
[
0.023588,
-0.515421,
0.118522,
-0.093211,
-0.54101,
-0.331597,
-0.530703,
-0.504356
]
],
"network.6.bias": [
0.154321,
-0.412325,
-0.100548,
0.346117,
0.140108,
-0.289905,
0.172132,
-0.168256
],
"network.8.weight": [
[
0.443465,
0.07383,
0.472389,
0.415031,
0.659422,
-0.113564,
0.965759,
0.172118
],
[
0.499227,
-0.250542,
0.283062,
0.196037,
0.913881,
0.092179,
0.904925,
-0.028534
],
[
0.226816,
-0.272782,
-0.282152,
0.149471,
-0.338137,
-0.195474,
-0.234149,
-0.24088
],
[
0.035817,
-0.082914,
-0.217852,
0.076621,
-0.176774,
-0.324214,
0.127586,
-0.219264
],
[
0.220204,
0.249811,
0.552779,
0.319601,
0.542489,
0.092364,
0.381013,
-0.142704
],
[
0.136869,
0.310308,
0.372211,
0.309767,
0.17925,
0.142805,
0.056584,
0.220102
],
[
-0.225276,
-0.279014,
-0.310845,
0.13778,
-0.242693,
0.047501,
-0.182091,
0.331776
],
[
0.506281,
-0.329217,
0.282821,
0.410439,
0.314271,
0.103154,
0.534801,
-0.022368
]
],
"network.8.bias": [
-0.090465,
-0.145289,
-0.331563,
-0.448993,
-0.190606,
-0.116722,
-0.27673,
0.135337
],
"network.10.weight": [
[
0.017975,
-0.048555,
-0.253439,
-0.344392,
0.226612,
0.152696,
-0.079656,
-0.345177
],
[
-0.162143,
0.014568,
0.112638,
0.238497,
-0.083106,
-0.143778,
-0.102679,
0.097597
],
[
-0.082082,
-0.068271,
-0.331332,
0.180097,
-0.299242,
0.230359,
-0.142893,
0.100011
],
[
-0.331984,
-0.19091,
-0.092273,
-0.32926,
0.11677,
0.153864,
-0.068567,
-0.086159
],
[
-0.090058,
-0.011356,
-0.314621,
0.054928,
0.015215,
-0.066865,
-0.31765,
-0.072712
],
[
0.45623,
0.803329,
0.174296,
-0.015036,
0.440783,
0.217878,
-0.098884,
0.015901
],
[
-0.00519,
-0.609486,
-0.085444,
-0.317783,
0.073377,
-0.192928,
0.213643,
-0.476895
],
[
0.404857,
0.745903,
-0.036989,
0.153605,
0.459327,
0.069012,
0.260294,
0.066289
]
],
"network.10.bias": [
-0.071394,
-0.096644,
-0.416742,
-0.301477,
-0.282989,
-0.19062,
-0.389979,
-0.039447
],
"network.12.weight": [
[
-0.027402,
-0.039165,
0.159587,
-0.182419,
0.171105,
-0.328393,
-0.05612,
-0.245832
]
],
"network.12.bias": [
0.153849
]
}
## Activation Signature
### 0
mean: [-0.669460, -0.833897, 0.182153, -0.454503, 1.028406, 1.629688, -1.071548, -0.396046]
std: [1.854730, 1.297849, 1.943905, 1.220727, 2.147697, 2.068893, 0.777190, 1.229393]
### 2
mean: [0.517045, -1.271639, -1.049378, 0.983784, 1.097028, -1.158911, -1.818789, 0.393625]
std: [1.039399, 1.935617, 0.647216, 2.154942, 0.770020, 1.981942, 1.831852, 1.691164]
### 4
mean: [-0.543963, -1.291011, -0.276529, -1.452035, 0.191077, 1.023232, 1.716448, -0.466263]
std: [0.272027, 0.952857, 1.785949, 0.901772, 1.256296, 1.129818, 2.266163, 0.685061]
### 6
mean: [1.090418, -1.068026, 1.173189, 0.173349, 1.410687, -0.870273, 1.564454, -1.765005]
std: [2.345822, 0.539160, 1.438119, 1.656263, 2.690109, 0.682024, 2.643596, 1.336540]
### 8
mean: [4.500963, 4.494639, -1.290019, -0.686598, 2.767562, 1.204871, -1.683140, 3.134282]
std: [5.509926, 5.595027, 1.129255, 0.278372, 3.522008, 1.557808, 1.703697, 3.610466]
### 10
mean: [-0.464959, -0.869780, -1.342376, -2.419172, -1.009399, 7.075077, -4.702689, 6.759826]
std: [0.396739, 0.966713, 1.158312, 2.554486, 0.869764, 8.900249, 5.181146, 8.317016]
### 12
mean: [-3.851660]
std: [4.950858]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_descending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.244199,
0.019227,
-0.652825,
0.497861,
-0.087845
],
[
-0.53077,
0.04171,
0.20217,
-0.368098,
0.397038
],
[
-0.437454,
-0.305915,
0.42936,
-0.373294,
0.719472
],
[
-0.570416,
0.092414,
0.278611,
-0.192652,
0.365096
],
[
0.811458,
0.384183,
0.08872,
-0.362258,
0.025868
],
[
0.754073,
-0.234122,
0.422914,
0.022426,
0.32139
],
[
-0.306648,
0.070051,
0.082398,
-0.175372,
-0.181656
],
[
-0.435555,
0.261414,
-0.319476,
-0.021708,
0.453179
]
],
"network.0.bias": [
0.00378,
-0.372012,
0.300258,
-0.532192,
-0.121862,
-0.209984,
-0.389592,
-0.175399
],
"network.2.weight": [
[
0.133097,
-0.305715,
-0.332636,
0.05459,
0.234105,
0.295159,
0.46644,
-0.093882
],
[
0.600238,
0.079848,
-0.134117,
-0.061044,
-0.463501,
-0.438674,
0.067351,
0.129061
],
[
-0.322318,
-0.043712,
0.076783,
-0.406948,
-0.055226,
-0.311244,
-0.061083,
-0.04864
],
[
0.356824,
0.621137,
0.979078,
0.509473,
-0.488066,
0.112926,
0.097875,
0.494718
],
[
0.337077,
0.650183,
0.332378,
0.416068,
0.348715,
-0.271865,
0.349856,
0.153156
],
[
0.608379,
0.058329,
-0.473999,
-0.219568,
-0.474172,
-0.338009,
0.159839,
0.363617
],
[
-0.057879,
-0.47077,
-0.329898,
-0.310652,
-0.495965,
-0.39847,
-0.289736,
-0.345243
],
[
-0.091747,
-0.244197,
-0.507665,
-0.214963,
0.54316,
0.108558,
0.492425,
-0.57585
]
],
"network.2.bias": [
0.002061,
-0.077158,
-0.234492,
0.025745,
0.401,
0.141392,
0.106044,
0.248735
],
"network.4.weight": [
[
-0.290319,
-0.120178,
0.228426,
0.013055,
-0.151079,
-0.34493,
0.030121,
0.140894
],
[
-0.135918,
-0.423334,
0.116461,
-0.523968,
-0.015681,
-0.273671,
-0.048452,
0.003132
],
[
0.523652,
-0.498147,
-0.084381,
-0.162965,
-0.720107,
-0.783535,
-0.197133,
0.544269
],
[
-0.481688,
-0.145572,
0.137052,
-0.509765,
0.085021,
0.01747,
-0.160626,
-0.291781
],
[
0.115049,
-0.127338,
0.061233,
-0.086468,
-0.436378,
-0.700584,
-0.377064,
0.621961
],
[
0.094029,
0.750135,
0.134653,
0.403073,
0.249311,
0.431037,
-0.015621,
-0.208689
],
[
0.11522,
0.440737,
0.274613,
0.957771,
0.552855,
0.188907,
-0.014024,
-0.341565
],
[
0.07959,
-0.636291,
0.264474,
-0.2609,
0.013737,
-0.027526,
0.454657,
0.14833
]
],
"network.4.bias": [
-0.217929,
-0.357774,
0.233087,
-0.306621,
0.406849,
0.093369,
-0.080669,
-0.195694
],
"network.6.weight": [
[
-0.019772,
-0.076474,
-0.261464,
-0.025379,
-0.625906,
-0.114458,
0.888337,
-0.18832
],
[
-0.319922,
0.278873,
-0.086931,
-0.319128,
-0.009144,
-0.382402,
-0.085135,
-0.527517
],
[
-0.214552,
0.228377,
-0.14925,
0.251282,
0.226011,
0.532483,
0.396896,
-0.484337
],
[
-0.016209,
0.016016,
-0.675026,
-0.012478,
-0.585272,
-0.00315,
0.321529,
-0.174125
],
[
-0.11673,
-0.010083,
-0.398427,
-0.444104,
-0.539337,
0.442071,
0.763546,
-0.166652
],
[
-0.180255,
-0.069241,
-0.228916,
-0.156148,
-0.505495,
-0.079808,
-0.017843,
-0.393082
],
[
0.032822,
0.12405,
-0.34569,
-0.101366,
-0.36424,
0.078689,
0.978166,
-0.341764
],
[
0.023588,
-0.515421,
0.118522,
-0.093211,
-0.54101,
-0.331597,
-0.530703,
-0.504356
]
],
"network.6.bias": [
0.154321,
-0.412325,
-0.100548,
0.346117,
0.140108,
-0.289905,
0.172132,
-0.168256
],
"network.8.weight": [
[
0.443465,
0.07383,
0.472389,
0.415031,
0.659422,
-0.113564,
0.965759,
0.172118
],
[
0.499227,
-0.250542,
0.283062,
0.196037,
0.913881,
0.092179,
0.904925,
-0.028534
],
[
0.226816,
-0.272782,
-0.282152,
0.149471,
-0.338137,
-0.195474,
-0.234149,
-0.24088
],
[
0.035817,
-0.082914,
-0.217852,
0.076621,
-0.176774,
-0.324214,
0.127586,
-0.219264
],
[
0.220204,
0.249811,
0.552779,
0.319601,
0.542489,
0.092364,
0.381013,
-0.142704
],
[
0.136869,
0.310308,
0.372211,
0.309767,
0.17925,
0.142805,
0.056584,
0.220102
],
[
-0.225276,
-0.279014,
-0.310845,
0.13778,
-0.242693,
0.047501,
-0.182091,
0.331776
],
[
0.506281,
-0.329217,
0.282821,
0.410439,
0.314271,
0.103154,
0.534801,
-0.022368
]
],
"network.8.bias": [
-0.090465,
-0.145289,
-0.331563,
-0.448993,
-0.190606,
-0.116722,
-0.27673,
0.135337
],
"network.10.weight": [
[
0.017975,
-0.048555,
-0.253439,
-0.344392,
0.226612,
0.152696,
-0.079656,
-0.345177
],
[
-0.162143,
0.014568,
0.112638,
0.238497,
-0.083106,
-0.143778,
-0.102679,
0.097597
],
[
-0.082082,
-0.068271,
-0.331332,
0.180097,
-0.299242,
0.230359,
-0.142893,
0.100011
],
[
-0.331984,
-0.19091,
-0.092273,
-0.32926,
0.11677,
0.153864,
-0.068567,
-0.086159
],
[
-0.090058,
-0.011356,
-0.314621,
0.054928,
0.015215,
-0.066865,
-0.31765,
-0.072712
],
[
0.45623,
0.803329,
0.174296,
-0.015036,
0.440783,
0.217878,
-0.098884,
0.015901
],
[
-0.00519,
-0.609486,
-0.085444,
-0.317783,
0.073377,
-0.192928,
0.213643,
-0.476895
],
[
0.404857,
0.745903,
-0.036989,
0.153605,
0.459327,
0.069012,
0.260294,
0.066289
]
],
"network.10.bias": [
-0.071394,
-0.096644,
-0.416742,
-0.301477,
-0.282989,
-0.19062,
-0.389979,
-0.039447
],
"network.12.weight": [
[
-0.027402,
-0.039165,
0.159587,
-0.182419,
0.171105,
-0.328393,
-0.05612,
-0.245832
]
],
"network.12.bias": [
0.153849
]
}
## Activation Signature
### 0
mean: [-0.669460, -0.833897, 0.182153, -0.454503, 1.028406, 1.629688, -1.071548, -0.396046]
std: [1.854730, 1.297849, 1.943905, 1.220727, 2.147697, 2.068893, 0.777190, 1.229393]
### 2
mean: [0.517045, -1.271639, -1.049378, 0.983784, 1.097028, -1.158911, -1.818789, 0.393625]
std: [1.039399, 1.935617, 0.647216, 2.154942, 0.770020, 1.981942, 1.831852, 1.691164]
### 4
mean: [-0.543963, -1.291011, -0.276529, -1.452035, 0.191077, 1.023232, 1.716448, -0.466263]
std: [0.272027, 0.952857, 1.785949, 0.901772, 1.256296, 1.129818, 2.266163, 0.685061]
### 6
mean: [1.090418, -1.068026, 1.173189, 0.173349, 1.410687, -0.870273, 1.564454, -1.765005]
std: [2.345822, 0.539160, 1.438119, 1.656263, 2.690109, 0.682024, 2.643596, 1.336540]
### 8
mean: [4.500963, 4.494639, -1.290019, -0.686598, 2.767562, 1.204871, -1.683140, 3.134282]
std: [5.509926, 5.595027, 1.129255, 0.278372, 3.522008, 1.557808, 1.703697, 3.610466]
### 10
mean: [-0.464959, -0.869780, -1.342376, -2.419172, -1.009399, 7.075077, -4.702689, 6.759826]
std: [0.396739, 0.966713, 1.158312, 2.554486, 0.869764, 8.900249, 5.181146, 8.317016]
### 12
mean: [-3.851660]
std: [4.950858]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_descending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -0.6694602370262146, "std": 1.8547303676605225}, "1": {"mean": -0.8338971734046936, "std": 1.297849178314209}, "2": {"mean": 0.18215273320674896, "std": 1.943905234336853}, "3": {"mean": -0.45450302958488464, "std": 1.2207273244857788}, "4": {"mean": 1.0284063816070557, "std": 2.1476969718933105}, "5": {"mean": 1.6296875476837158, "std": 2.06889271736145}, "6": {"mean": -1.0715481042861938, "std": 0.7771901488304138}, "7": {"mean": -0.3960462808609009, "std": 1.2293933629989624}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.5170445442199707, "std": 1.0393986701965332}, "1": {"mean": -1.2716392278671265, "std": 1.9356166124343872}, "2": {"mean": -1.0493783950805664, "std": 0.6472157835960388}, "3": {"mean": 0.9837835431098938, "std": 2.1549415588378906}, "4": {"mean": 1.0970280170440674, "std": 0.7700195908546448}, "5": {"mean": -1.158911108970642, "std": 1.981942057609558}, "6": {"mean": -1.8187885284423828, "std": 1.8318519592285156}, "7": {"mean": 0.39362525939941406, "std": 1.6911635398864746}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.5439625382423401, "std": 0.27202650904655457}, "1": {"mean": -1.2910113334655762, "std": 0.9528571367263794}, "2": {"mean": -0.2765286862850189, "std": 1.7859487533569336}, "3": {"mean": -1.4520353078842163, "std": 0.9017717838287354}, "4": {"mean": 0.19107718765735626, "std": 1.2562958002090454}, "5": {"mean": 1.0232315063476562, "std": 1.1298178434371948}, "6": {"mean": 1.716448426246643, "std": 2.266162872314453}, "7": {"mean": -0.4662628173828125, "std": 0.6850610971450806}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 1.090417504310608, "std": 2.345822334289551}, "1": {"mean": -1.0680255889892578, "std": 0.5391597747802734}, "2": {"mean": 1.1731892824172974, "std": 1.438118815422058}, "3": {"mean": 0.17334875464439392, "std": 1.6562626361846924}, "4": {"mean": 1.410686731338501, "std": 2.6901094913482666}, "5": {"mean": -0.8702731728553772, "std": 0.6820238828659058}, "6": {"mean": 1.564454436302185, "std": 2.6435956954956055}, "7": {"mean": -1.765005350112915, "std": 1.3365401029586792}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 4.500962734222412, "std": 5.509925842285156}, "1": {"mean": 4.4946393966674805, "std": 5.595026969909668}, "2": {"mean": -1.290018916130066, "std": 1.1292552947998047}, "3": {"mean": -0.6865980625152588, "std": 0.27837175130844116}, "4": {"mean": 2.767561912536621, "std": 3.522008180618286}, "5": {"mean": 1.2048711776733398, "std": 1.5578075647354126}, "6": {"mean": -1.6831399202346802, "std": 1.7036967277526855}, "7": {"mean": 3.134281635284424, "std": 3.6104660034179688}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -0.4649594724178314, "std": 0.39673861861228943}, "1": {"mean": -0.869780421257019, "std": 0.966713011264801}, "2": {"mean": -1.342375636100769, "std": 1.158312201499939}, "3": {"mean": -2.4191715717315674, "std": 2.554486036300659}, "4": {"mean": -1.0093986988067627, "std": 0.8697639107704163}, "5": {"mean": 7.075076580047607, "std": 8.900248527526855}, "6": {"mean": -4.702689170837402, "std": 5.181146144866943}, "7": {"mean": 6.759826183319092, "std": 8.317015647888184}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -3.8516595363616943, "std": 4.950858116149902}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.244199, 0.019227, -0.652825, 0.497861, -0.087845], [-0.53077, 0.04171, 0.20217, -0.368098, 0.397038], [-0.437454, -0.305915, 0.42936, -0.373294, 0.719472], [-0.570416, 0.092414, 0.278611, -0.192652, 0.365096], [0.811458, 0.384183, 0.08872, -0.362258, 0.025868], [0.754073, -0.234122, 0.422914, 0.022426, 0.32139], [-0.306648, 0.070051, 0.082398, -0.175372, -0.181656], [-0.435555, 0.261414, -0.319476, -0.021708, 0.453179]], "network.0.bias": [0.00378, -0.372012, 0.300258, -0.532192, -0.121862, -0.209984, -0.389592, -0.175399], "network.2.weight": [[0.133097, -0.305715, -0.332636, 0.05459, 0.234105, 0.295159, 0.46644, -0.093882], [0.600238, 0.079848, -0.134117, -0.061044, -0.463501, -0.438674, 0.067351, 0.129061], [-0.322318, -0.043712, 0.076783, -0.406948, -0.055226, -0.311244, -0.061083, -0.04864], [0.356824, 0.621137, 0.979078, 0.509473, -0.488066, 0.112926, 0.097875, 0.494718], [0.337077, 0.650183, 0.332378, 0.416068, 0.348715, -0.271865, 0.349856, 0.153156], [0.608379, 0.058329, -0.473999, -0.219568, -0.474172, -0.338009, 0.159839, 0.363617], [-0.057879, -0.47077, -0.329898, -0.310652, -0.495965, -0.39847, -0.289736, -0.345243], [-0.091747, -0.244197, -0.507665, -0.214963, 0.54316, 0.108558, 0.492425, -0.57585]], "network.2.bias": [0.002061, -0.077158, -0.234492, 0.025745, 0.401, 0.141392, 0.106044, 0.248735], "network.4.weight": [[-0.290319, -0.120178, 0.228426, 0.013055, -0.151079, -0.34493, 0.030121, 0.140894], [-0.135918, -0.423334, 0.116461, -0.523968, -0.015681, -0.273671, -0.048452, 0.003132], [0.523652, -0.498147, -0.084381, -0.162965, -0.720107, -0.783535, -0.197133, 0.544269], [-0.481688, -0.145572, 0.137052, -0.509765, 0.085021, 0.01747, -0.160626, -0.291781], [0.115049, -0.127338, 0.061233, -0.086468, -0.436378, -0.700584, -0.377064, 0.621961], [0.094029, 0.750135, 0.134653, 0.403073, 0.249311, 0.431037, -0.015621, -0.208689], [0.11522, 0.440737, 0.274613, 0.957771, 0.552855, 0.188907, -0.014024, -0.341565], [0.07959, -0.636291, 0.264474, -0.2609, 0.013737, -0.027526, 0.454657, 0.14833]], "network.4.bias": [-0.217929, -0.357774, 0.233087, -0.306621, 0.406849, 0.093369, -0.080669, -0.195694], "network.6.weight": [[-0.019772, -0.076474, -0.261464, -0.025379, -0.625906, -0.114458, 0.888337, -0.18832], [-0.319922, 0.278873, -0.086931, -0.319128, -0.009144, -0.382402, -0.085135, -0.527517], [-0.214552, 0.228377, -0.14925, 0.251282, 0.226011, 0.532483, 0.396896, -0.484337], [-0.016209, 0.016016, -0.675026, -0.012478, -0.585272, -0.00315, 0.321529, -0.174125], [-0.11673, -0.010083, -0.398427, -0.444104, -0.539337, 0.442071, 0.763546, -0.166652], [-0.180255, -0.069241, -0.228916, -0.156148, -0.505495, -0.079808, -0.017843, -0.393082], [0.032822, 0.12405, -0.34569, -0.101366, -0.36424, 0.078689, 0.978166, -0.341764], [0.023588, -0.515421, 0.118522, -0.093211, -0.54101, -0.331597, -0.530703, -0.504356]], "network.6.bias": [0.154321, -0.412325, -0.100548, 0.346117, 0.140108, -0.289905, 0.172132, -0.168256], "network.8.weight": [[0.443465, 0.07383, 0.472389, 0.415031, 0.659422, -0.113564, 0.965759, 0.172118], [0.499227, -0.250542, 0.283062, 0.196037, 0.913881, 0.092179, 0.904925, -0.028534], [0.226816, -0.272782, -0.282152, 0.149471, -0.338137, -0.195474, -0.234149, -0.24088], [0.035817, -0.082914, -0.217852, 0.076621, -0.176774, -0.324214, 0.127586, -0.219264], [0.220204, 0.249811, 0.552779, 0.319601, 0.542489, 0.092364, 0.381013, -0.142704], [0.136869, 0.310308, 0.372211, 0.309767, 0.17925, 0.142805, 0.056584, 0.220102], [-0.225276, -0.279014, -0.310845, 0.13778, -0.242693, 0.047501, -0.182091, 0.331776], [0.506281, -0.329217, 0.282821, 0.410439, 0.314271, 0.103154, 0.534801, -0.022368]], "network.8.bias": [-0.090465, -0.145289, -0.331563, -0.448993, -0.190606, -0.116722, -0.27673, 0.135337], "network.10.weight": [[0.017975, -0.048555, -0.253439, -0.344392, 0.226612, 0.152696, -0.079656, -0.345177], [-0.162143, 0.014568, 0.112638, 0.238497, -0.083106, -0.143778, -0.102679, 0.097597], [-0.082082, -0.068271, -0.331332, 0.180097, -0.299242, 0.230359, -0.142893, 0.100011], [-0.331984, -0.19091, -0.092273, -0.32926, 0.11677, 0.153864, -0.068567, -0.086159], [-0.090058, -0.011356, -0.314621, 0.054928, 0.015215, -0.066865, -0.31765, -0.072712], [0.45623, 0.803329, 0.174296, -0.015036, 0.440783, 0.217878, -0.098884, 0.015901], [-0.00519, -0.609486, -0.085444, -0.317783, 0.073377, -0.192928, 0.213643, -0.476895], [0.404857, 0.745903, -0.036989, 0.153605, 0.459327, 0.069012, 0.260294, 0.066289]], "network.10.bias": [-0.071394, -0.096644, -0.416742, -0.301477, -0.282989, -0.19062, -0.389979, -0.039447], "network.12.weight": [[-0.027402, -0.039165, 0.159587, -0.182419, 0.171105, -0.328393, -0.05612, -0.245832]], "network.12.bias": [0.153849]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6899471879005432, "train_acc": 0.555, "val_loss": 0.7204413414001465, "val_acc": 0.46}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6566338241100311, "train_acc": 0.58, "val_loss": 0.6797667145729065, "val_acc": 0.46}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.6603061556816101, "train_acc": 0.51, "val_loss": 0.6930835843086243, "val_acc": 0.46}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.6767034232616425, "train_acc": 0.485, "val_loss": 0.6258223056793213, "val_acc": 0.7}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.5235426276922226, "train_acc": 0.77, "val_loss": 0.41012924909591675, "val_acc": 0.92}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.39728547632694244, "train_acc": 0.925, "val_loss": 0.3755057156085968, "val_acc": 0.92}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.32867497205734253, "train_acc": 0.935, "val_loss": 0.352335661649704, "val_acc": 0.92}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.2909769266843796, "train_acc": 0.95, "val_loss": 0.32986411452293396, "val_acc": 0.92}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.28522033989429474, "train_acc": 0.95, "val_loss": 0.3096386194229126, "val_acc": 0.92}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.252244234085083, "train_acc": 0.955, "val_loss": 0.2921809256076813, "val_acc": 0.92}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.2404918149113655, "train_acc": 0.95, "val_loss": 0.2776013910770416, "val_acc": 0.92}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.20445748418569565, "train_acc": 0.96, "val_loss": 0.2658655643463135, "val_acc": 0.92}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["sorted_descending"], "degraded_stage": {"initial_val_loss": 0.7204413414001465, "final_val_loss": 0.6797667145729065, "initial_val_acc": 0.46, "final_val_acc": 0.46, "best_val_acc": 0.46}, "improved_stage": {"initial_val_loss": 0.6930835843086243, "final_val_loss": 0.2658655643463135, "initial_val_acc": 0.46, "final_val_acc": 0.92, "best_val_acc": 0.92, "best_epoch": 4}, "improvement": 0.46, "first_improvement_epoch": 1}}
|
63
|
{"target_pattern": "first_last_match", "degraded_accuracy": 0.48, "improved_accuracy": 0.76, "improvement": 0.28, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 8264, "learning_rate": 0.06745739206568764, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "first_last_match", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["first_last_match"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-1.061712,
-0.572723,
0.141541,
0.068523,
0.816453
],
[
-0.749255,
-0.135412,
0.028043,
-0.005888,
1.08257
],
[
-0.216142,
-0.395608,
-0.35785,
-0.244309,
-0.305778
],
[
-0.231491,
-0.222324,
0.184767,
-0.199086,
0.698054
],
[
-0.409887,
-0.277577,
-0.01932,
-0.471375,
-0.68898
],
[
0.691359,
0.471846,
0.109713,
0.184706,
0.080333
],
[
-0.289604,
0.207487,
0.13253,
0.103636,
-0.362962
],
[
-0.629345,
0.31124,
-0.418684,
0.124955,
0.072816
]
],
"network.0.bias": [
-0.094394,
0.173618,
0.074195,
-0.491869,
-0.511474,
0.710552,
-0.690601,
-0.556414
],
"network.2.weight": [
[
-0.524995,
-0.376504,
-0.300763,
-0.444088,
-0.400801,
-0.718894,
-0.085947,
-0.411875
],
[
-0.124819,
-0.250065,
-0.0833,
0.244981,
0.055407,
0.52757,
0.486416,
0.221377
],
[
-0.353755,
-0.615504,
-0.0079,
0.211631,
0.137017,
-0.091332,
0.034913,
0.055345
],
[
-0.647675,
-0.377578,
0.029983,
-0.287159,
-0.234683,
0.055383,
-0.166382,
-0.711932
],
[
0.334539,
0.521489,
0.046629,
0.006636,
0.066369,
-0.159105,
-0.33995,
-0.048588
],
[
0.710547,
0.389782,
-0.200069,
0.181999,
-0.34695,
0.101798,
0.241884,
-0.163874
],
[
-0.06297,
-0.070665,
0.079392,
-0.336544,
0.241526,
-0.335032,
-0.074308,
-0.553012
],
[
0.554435,
0.995317,
-0.136961,
0.422027,
-0.250951,
0.089084,
-0.48529,
0.142285
]
],
"network.2.bias": [
-0.436851,
0.196737,
-0.095318,
-0.11298,
-0.253634,
0.614864,
-0.329419,
0.020968
],
"network.4.weight": [
[
0.019648,
0.138752,
-0.258586,
0.203333,
-0.347458,
-0.44649,
0.136728,
0.063014
],
[
-0.160896,
-0.093632,
-0.379638,
-0.28894,
0.125454,
-0.284203,
-0.473974,
0.134472
],
[
-0.027275,
-0.119247,
-0.001152,
-0.499436,
-0.641116,
-0.323974,
0.018429,
-0.225835
],
[
-0.490524,
-0.512795,
-0.104485,
0.494024,
-0.065048,
-0.316327,
0.044433,
-0.337028
],
[
-0.221368,
-0.063171,
0.114581,
-0.604014,
-0.220972,
-0.600161,
0.131276,
-0.036249
],
[
-0.341716,
-0.146992,
0.039076,
-0.667044,
-0.065209,
-0.525217,
0.335515,
0.224203
],
[
0.170236,
-0.331171,
0.086177,
0.136015,
0.071993,
0.578253,
-0.3082,
0.636366
],
[
0.07101,
-0.916106,
-0.418053,
0.024031,
0.425963,
-0.115176,
-0.19379,
0.827815
]
],
"network.4.bias": [
-0.379247,
-0.020925,
-0.009207,
-0.16419,
-0.217383,
0.020478,
0.023328,
0.160562
],
"network.6.weight": [
[
-0.294536,
-0.151575,
0.068544,
0.138185,
-0.315163,
0.125892,
-0.23007,
-0.302583
],
[
-0.148352,
-0.154811,
-0.008297,
0.073114,
-0.265115,
-0.164496,
0.276991,
-0.040182
],
[
0.311241,
-0.043116,
0.099644,
-0.435421,
0.062416,
0.421788,
-0.665621,
0.052163
],
[
-0.310074,
-0.331169,
0.091716,
-0.105178,
0.773336,
0.255292,
0.174688,
-0.593687
],
[
0.039692,
-0.196073,
-0.39601,
0.513855,
0.575915,
0.000679,
-0.242899,
-0.592157
],
[
-0.021989,
-0.103461,
0.518131,
-0.21356,
0.07487,
0.224581,
-0.142127,
0.043358
],
[
-0.011782,
-0.16436,
-0.173533,
-0.227878,
0.007494,
0.253824,
0.745782,
0.923548
],
[
0.21282,
0.141041,
0.09536,
0.160994,
0.028164,
-0.3882,
-0.207591,
0.49239
]
],
"network.6.bias": [
-0.389028,
-0.517186,
0.131226,
0.492981,
-0.292979,
-0.010838,
0.083761,
-0.096793
],
"network.8.weight": [
[
-0.137726,
0.39312,
0.182557,
-0.847809,
-0.243457,
-0.272459,
0.930779,
0.062183
],
[
0.269555,
0.313924,
0.244109,
0.120021,
0.63553,
-0.040697,
-0.013889,
-0.611414
],
[
0.100668,
0.152475,
0.205478,
-0.506289,
0.001647,
0.211453,
0.910036,
-0.061145
],
[
0.076847,
-0.595474,
-0.405433,
0.307867,
0.024958,
-0.038671,
0.165388,
-0.558571
],
[
-0.087536,
0.046048,
0.301124,
-0.080291,
0.131746,
-0.05887,
-0.295475,
-0.438255
],
[
-0.505984,
-0.213786,
0.113037,
0.184483,
-0.240867,
0.123203,
-0.322015,
-0.533184
],
[
-0.136759,
-0.240346,
-0.223569,
-0.154711,
-0.058773,
-0.12493,
-0.317776,
0.29923
],
[
-0.115777,
0.201799,
-0.242451,
-0.512658,
-0.130133,
-0.157092,
0.980573,
-0.184011
]
],
"network.8.bias": [
0.101862,
-0.117607,
-0.209601,
-0.126828,
-0.399779,
-0.25114,
-0.028458,
0.213317
],
"network.10.weight": [
[
-0.409721,
-0.79051,
-0.083073,
0.098727,
-0.208597,
-0.193628,
0.345529,
-0.455203
],
[
0.090074,
-0.453936,
-0.336461,
-0.055889,
-0.007064,
0.091411,
-0.217199,
-0.039967
],
[
-0.185261,
-0.250562,
-0.149713,
-0.310794,
-0.116257,
0.251797,
0.016701,
0.152577
],
[
0.201754,
-0.407591,
0.021336,
-0.105077,
0.100267,
0.123229,
0.146604,
0.325616
],
[
0.319589,
0.193598,
0.520731,
-0.31669,
-0.318919,
0.107876,
-0.283256,
-0.046941
],
[
-0.098393,
-0.29393,
-0.09544,
0.091718,
-0.309536,
0.128559,
-0.307865,
-0.248463
],
[
0.216574,
-0.187546,
0.311602,
0.201148,
-0.201806,
0.07778,
-0.224763,
0.298194
],
[
0.276131,
-0.128138,
0.078387,
0.006956,
-0.116458,
-0.074013,
0.253648,
0.263081
]
],
"network.10.bias": [
-0.211868,
-0.11998,
-0.217628,
-0.28413,
-0.369829,
-0.325552,
-0.157529,
-0.318813
],
"network.12.weight": [
[
0.214864,
-0.015266,
0.027294,
-0.140124,
-0.316361,
-0.069746,
-0.295879,
-0.088
]
],
"network.12.bias": [
0.283847
]
}
## Activation Signature
### 0
mean: [-1.009700, 0.375585, -2.564209, -0.365108, -3.428299, 3.155066, -0.621480, -1.289269]
std: [2.852535, 2.345353, 1.729918, 1.488445, 2.209950, 2.074584, 0.972291, 1.642708]
### 2
mean: [-3.609810, 1.757479, -1.061626, -0.880674, -0.121493, 1.772721, -1.721024, 1.707438]
std: [2.046173, 1.263054, 1.238866, 1.598343, 1.337119, 1.524439, 0.862603, 2.592325]
### 4
mean: [-0.955874, -0.417343, -1.466251, -2.214583, -1.568936, -0.836673, 1.591233, -0.058961]
std: [0.951953, 0.190589, 1.677863, 1.354924, 1.205723, 0.349891, 2.748386, 2.996126]
### 6
mean: [-1.056862, -0.107448, -0.896669, 0.197344, -1.263890, -0.198911, 2.192726, 0.046772]
std: [1.341929, 0.662848, 1.695655, 0.957031, 2.065157, 0.287876, 4.213951, 0.634167]
### 8
mean: [1.847467, -0.143803, 1.576521, 0.130038, -1.153409, -1.018675, -0.793391, 2.119384]
std: [4.305826, 0.266981, 3.964861, 0.122607, 1.456977, 1.798665, 1.265972, 4.228659]
### 10
mean: [-2.097443, -0.621845, -0.552033, 0.817009, 1.006024, -1.195069, 1.457735, 0.907103]
std: [3.993957, 1.095965, 0.738243, 2.317804, 3.184814, 1.839210, 3.404357, 2.596407]
### 12
mean: [-0.797464]
std: [2.507605]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
first_last_match
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-1.061712,
-0.572723,
0.141541,
0.068523,
0.816453
],
[
-0.749255,
-0.135412,
0.028043,
-0.005888,
1.08257
],
[
-0.216142,
-0.395608,
-0.35785,
-0.244309,
-0.305778
],
[
-0.231491,
-0.222324,
0.184767,
-0.199086,
0.698054
],
[
-0.409887,
-0.277577,
-0.01932,
-0.471375,
-0.68898
],
[
0.691359,
0.471846,
0.109713,
0.184706,
0.080333
],
[
-0.289604,
0.207487,
0.13253,
0.103636,
-0.362962
],
[
-0.629345,
0.31124,
-0.418684,
0.124955,
0.072816
]
],
"network.0.bias": [
-0.094394,
0.173618,
0.074195,
-0.491869,
-0.511474,
0.710552,
-0.690601,
-0.556414
],
"network.2.weight": [
[
-0.524995,
-0.376504,
-0.300763,
-0.444088,
-0.400801,
-0.718894,
-0.085947,
-0.411875
],
[
-0.124819,
-0.250065,
-0.0833,
0.244981,
0.055407,
0.52757,
0.486416,
0.221377
],
[
-0.353755,
-0.615504,
-0.0079,
0.211631,
0.137017,
-0.091332,
0.034913,
0.055345
],
[
-0.647675,
-0.377578,
0.029983,
-0.287159,
-0.234683,
0.055383,
-0.166382,
-0.711932
],
[
0.334539,
0.521489,
0.046629,
0.006636,
0.066369,
-0.159105,
-0.33995,
-0.048588
],
[
0.710547,
0.389782,
-0.200069,
0.181999,
-0.34695,
0.101798,
0.241884,
-0.163874
],
[
-0.06297,
-0.070665,
0.079392,
-0.336544,
0.241526,
-0.335032,
-0.074308,
-0.553012
],
[
0.554435,
0.995317,
-0.136961,
0.422027,
-0.250951,
0.089084,
-0.48529,
0.142285
]
],
"network.2.bias": [
-0.436851,
0.196737,
-0.095318,
-0.11298,
-0.253634,
0.614864,
-0.329419,
0.020968
],
"network.4.weight": [
[
0.019648,
0.138752,
-0.258586,
0.203333,
-0.347458,
-0.44649,
0.136728,
0.063014
],
[
-0.160896,
-0.093632,
-0.379638,
-0.28894,
0.125454,
-0.284203,
-0.473974,
0.134472
],
[
-0.027275,
-0.119247,
-0.001152,
-0.499436,
-0.641116,
-0.323974,
0.018429,
-0.225835
],
[
-0.490524,
-0.512795,
-0.104485,
0.494024,
-0.065048,
-0.316327,
0.044433,
-0.337028
],
[
-0.221368,
-0.063171,
0.114581,
-0.604014,
-0.220972,
-0.600161,
0.131276,
-0.036249
],
[
-0.341716,
-0.146992,
0.039076,
-0.667044,
-0.065209,
-0.525217,
0.335515,
0.224203
],
[
0.170236,
-0.331171,
0.086177,
0.136015,
0.071993,
0.578253,
-0.3082,
0.636366
],
[
0.07101,
-0.916106,
-0.418053,
0.024031,
0.425963,
-0.115176,
-0.19379,
0.827815
]
],
"network.4.bias": [
-0.379247,
-0.020925,
-0.009207,
-0.16419,
-0.217383,
0.020478,
0.023328,
0.160562
],
"network.6.weight": [
[
-0.294536,
-0.151575,
0.068544,
0.138185,
-0.315163,
0.125892,
-0.23007,
-0.302583
],
[
-0.148352,
-0.154811,
-0.008297,
0.073114,
-0.265115,
-0.164496,
0.276991,
-0.040182
],
[
0.311241,
-0.043116,
0.099644,
-0.435421,
0.062416,
0.421788,
-0.665621,
0.052163
],
[
-0.310074,
-0.331169,
0.091716,
-0.105178,
0.773336,
0.255292,
0.174688,
-0.593687
],
[
0.039692,
-0.196073,
-0.39601,
0.513855,
0.575915,
0.000679,
-0.242899,
-0.592157
],
[
-0.021989,
-0.103461,
0.518131,
-0.21356,
0.07487,
0.224581,
-0.142127,
0.043358
],
[
-0.011782,
-0.16436,
-0.173533,
-0.227878,
0.007494,
0.253824,
0.745782,
0.923548
],
[
0.21282,
0.141041,
0.09536,
0.160994,
0.028164,
-0.3882,
-0.207591,
0.49239
]
],
"network.6.bias": [
-0.389028,
-0.517186,
0.131226,
0.492981,
-0.292979,
-0.010838,
0.083761,
-0.096793
],
"network.8.weight": [
[
-0.137726,
0.39312,
0.182557,
-0.847809,
-0.243457,
-0.272459,
0.930779,
0.062183
],
[
0.269555,
0.313924,
0.244109,
0.120021,
0.63553,
-0.040697,
-0.013889,
-0.611414
],
[
0.100668,
0.152475,
0.205478,
-0.506289,
0.001647,
0.211453,
0.910036,
-0.061145
],
[
0.076847,
-0.595474,
-0.405433,
0.307867,
0.024958,
-0.038671,
0.165388,
-0.558571
],
[
-0.087536,
0.046048,
0.301124,
-0.080291,
0.131746,
-0.05887,
-0.295475,
-0.438255
],
[
-0.505984,
-0.213786,
0.113037,
0.184483,
-0.240867,
0.123203,
-0.322015,
-0.533184
],
[
-0.136759,
-0.240346,
-0.223569,
-0.154711,
-0.058773,
-0.12493,
-0.317776,
0.29923
],
[
-0.115777,
0.201799,
-0.242451,
-0.512658,
-0.130133,
-0.157092,
0.980573,
-0.184011
]
],
"network.8.bias": [
0.101862,
-0.117607,
-0.209601,
-0.126828,
-0.399779,
-0.25114,
-0.028458,
0.213317
],
"network.10.weight": [
[
-0.409721,
-0.79051,
-0.083073,
0.098727,
-0.208597,
-0.193628,
0.345529,
-0.455203
],
[
0.090074,
-0.453936,
-0.336461,
-0.055889,
-0.007064,
0.091411,
-0.217199,
-0.039967
],
[
-0.185261,
-0.250562,
-0.149713,
-0.310794,
-0.116257,
0.251797,
0.016701,
0.152577
],
[
0.201754,
-0.407591,
0.021336,
-0.105077,
0.100267,
0.123229,
0.146604,
0.325616
],
[
0.319589,
0.193598,
0.520731,
-0.31669,
-0.318919,
0.107876,
-0.283256,
-0.046941
],
[
-0.098393,
-0.29393,
-0.09544,
0.091718,
-0.309536,
0.128559,
-0.307865,
-0.248463
],
[
0.216574,
-0.187546,
0.311602,
0.201148,
-0.201806,
0.07778,
-0.224763,
0.298194
],
[
0.276131,
-0.128138,
0.078387,
0.006956,
-0.116458,
-0.074013,
0.253648,
0.263081
]
],
"network.10.bias": [
-0.211868,
-0.11998,
-0.217628,
-0.28413,
-0.369829,
-0.325552,
-0.157529,
-0.318813
],
"network.12.weight": [
[
0.214864,
-0.015266,
0.027294,
-0.140124,
-0.316361,
-0.069746,
-0.295879,
-0.088
]
],
"network.12.bias": [
0.283847
]
}
## Activation Signature
### 0
mean: [-1.009700, 0.375585, -2.564209, -0.365108, -3.428299, 3.155066, -0.621480, -1.289269]
std: [2.852535, 2.345353, 1.729918, 1.488445, 2.209950, 2.074584, 0.972291, 1.642708]
### 2
mean: [-3.609810, 1.757479, -1.061626, -0.880674, -0.121493, 1.772721, -1.721024, 1.707438]
std: [2.046173, 1.263054, 1.238866, 1.598343, 1.337119, 1.524439, 0.862603, 2.592325]
### 4
mean: [-0.955874, -0.417343, -1.466251, -2.214583, -1.568936, -0.836673, 1.591233, -0.058961]
std: [0.951953, 0.190589, 1.677863, 1.354924, 1.205723, 0.349891, 2.748386, 2.996126]
### 6
mean: [-1.056862, -0.107448, -0.896669, 0.197344, -1.263890, -0.198911, 2.192726, 0.046772]
std: [1.341929, 0.662848, 1.695655, 0.957031, 2.065157, 0.287876, 4.213951, 0.634167]
### 8
mean: [1.847467, -0.143803, 1.576521, 0.130038, -1.153409, -1.018675, -0.793391, 2.119384]
std: [4.305826, 0.266981, 3.964861, 0.122607, 1.456977, 1.798665, 1.265972, 4.228659]
### 10
mean: [-2.097443, -0.621845, -0.552033, 0.817009, 1.006024, -1.195069, 1.457735, 0.907103]
std: [3.993957, 1.095965, 0.738243, 2.317804, 3.184814, 1.839210, 3.404357, 2.596407]
### 12
mean: [-0.797464]
std: [2.507605]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
first_last_match
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.0097001791000366, "std": 2.8525354862213135}, "1": {"mean": 0.37558504939079285, "std": 2.345353126525879}, "2": {"mean": -2.564209222793579, "std": 1.7299184799194336}, "3": {"mean": -0.36510810256004333, "std": 1.4884445667266846}, "4": {"mean": -3.4282991886138916, "std": 2.2099504470825195}, "5": {"mean": 3.1550662517547607, "std": 2.0745842456817627}, "6": {"mean": -0.6214801669120789, "std": 0.9722913503646851}, "7": {"mean": -1.2892686128616333, "std": 1.6427083015441895}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -3.6098101139068604, "std": 2.046173334121704}, "1": {"mean": 1.7574787139892578, "std": 1.2630544900894165}, "2": {"mean": -1.0616257190704346, "std": 1.2388660907745361}, "3": {"mean": -0.880673885345459, "std": 1.5983425378799438}, "4": {"mean": -0.1214931458234787, "std": 1.3371186256408691}, "5": {"mean": 1.7727214097976685, "std": 1.5244392156600952}, "6": {"mean": -1.7210240364074707, "std": 0.8626025915145874}, "7": {"mean": 1.7074376344680786, "std": 2.592325448989868}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.9558736085891724, "std": 0.9519528746604919}, "1": {"mean": -0.4173429608345032, "std": 0.19058862328529358}, "2": {"mean": -1.4662507772445679, "std": 1.677863359451294}, "3": {"mean": -2.214583396911621, "std": 1.3549244403839111}, "4": {"mean": -1.5689356327056885, "std": 1.205722689628601}, "5": {"mean": -0.8366730809211731, "std": 0.3498906195163727}, "6": {"mean": 1.5912331342697144, "std": 2.7483856678009033}, "7": {"mean": -0.058961205184459686, "std": 2.996126413345337}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -1.0568618774414062, "std": 1.3419286012649536}, "1": {"mean": -0.10744761675596237, "std": 0.6628476977348328}, "2": {"mean": -0.8966686725616455, "std": 1.6956554651260376}, "3": {"mean": 0.19734397530555725, "std": 0.9570311903953552}, "4": {"mean": -1.2638903856277466, "std": 2.065156936645508}, "5": {"mean": -0.19891111552715302, "std": 0.28787586092948914}, "6": {"mean": 2.1927263736724854, "std": 4.213951110839844}, "7": {"mean": 0.04677172377705574, "std": 0.6341667175292969}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 1.8474667072296143, "std": 4.305826187133789}, "1": {"mean": -0.1438027024269104, "std": 0.2669810950756073}, "2": {"mean": 1.5765209197998047, "std": 3.9648611545562744}, "3": {"mean": 0.13003778457641602, "std": 0.12260686606168747}, "4": {"mean": -1.1534087657928467, "std": 1.4569768905639648}, "5": {"mean": -1.018674612045288, "std": 1.7986650466918945}, "6": {"mean": -0.7933911085128784, "std": 1.265972375869751}, "7": {"mean": 2.1193840503692627, "std": 4.228659152984619}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -2.097442626953125, "std": 3.993957281112671}, "1": {"mean": -0.6218451857566833, "std": 1.0959653854370117}, "2": {"mean": -0.552033007144928, "std": 0.7382429838180542}, "3": {"mean": 0.8170093894004822, "std": 2.3178043365478516}, "4": {"mean": 1.0060241222381592, "std": 3.184814453125}, "5": {"mean": -1.1950691938400269, "std": 1.8392102718353271}, "6": {"mean": 1.4577350616455078, "std": 3.4043567180633545}, "7": {"mean": 0.9071032404899597, "std": 2.596407175064087}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -0.7974644303321838, "std": 2.5076053142547607}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-1.061712, -0.572723, 0.141541, 0.068523, 0.816453], [-0.749255, -0.135412, 0.028043, -0.005888, 1.08257], [-0.216142, -0.395608, -0.35785, -0.244309, -0.305778], [-0.231491, -0.222324, 0.184767, -0.199086, 0.698054], [-0.409887, -0.277577, -0.01932, -0.471375, -0.68898], [0.691359, 0.471846, 0.109713, 0.184706, 0.080333], [-0.289604, 0.207487, 0.13253, 0.103636, -0.362962], [-0.629345, 0.31124, -0.418684, 0.124955, 0.072816]], "network.0.bias": [-0.094394, 0.173618, 0.074195, -0.491869, -0.511474, 0.710552, -0.690601, -0.556414], "network.2.weight": [[-0.524995, -0.376504, -0.300763, -0.444088, -0.400801, -0.718894, -0.085947, -0.411875], [-0.124819, -0.250065, -0.0833, 0.244981, 0.055407, 0.52757, 0.486416, 0.221377], [-0.353755, -0.615504, -0.0079, 0.211631, 0.137017, -0.091332, 0.034913, 0.055345], [-0.647675, -0.377578, 0.029983, -0.287159, -0.234683, 0.055383, -0.166382, -0.711932], [0.334539, 0.521489, 0.046629, 0.006636, 0.066369, -0.159105, -0.33995, -0.048588], [0.710547, 0.389782, -0.200069, 0.181999, -0.34695, 0.101798, 0.241884, -0.163874], [-0.06297, -0.070665, 0.079392, -0.336544, 0.241526, -0.335032, -0.074308, -0.553012], [0.554435, 0.995317, -0.136961, 0.422027, -0.250951, 0.089084, -0.48529, 0.142285]], "network.2.bias": [-0.436851, 0.196737, -0.095318, -0.11298, -0.253634, 0.614864, -0.329419, 0.020968], "network.4.weight": [[0.019648, 0.138752, -0.258586, 0.203333, -0.347458, -0.44649, 0.136728, 0.063014], [-0.160896, -0.093632, -0.379638, -0.28894, 0.125454, -0.284203, -0.473974, 0.134472], [-0.027275, -0.119247, -0.001152, -0.499436, -0.641116, -0.323974, 0.018429, -0.225835], [-0.490524, -0.512795, -0.104485, 0.494024, -0.065048, -0.316327, 0.044433, -0.337028], [-0.221368, -0.063171, 0.114581, -0.604014, -0.220972, -0.600161, 0.131276, -0.036249], [-0.341716, -0.146992, 0.039076, -0.667044, -0.065209, -0.525217, 0.335515, 0.224203], [0.170236, -0.331171, 0.086177, 0.136015, 0.071993, 0.578253, -0.3082, 0.636366], [0.07101, -0.916106, -0.418053, 0.024031, 0.425963, -0.115176, -0.19379, 0.827815]], "network.4.bias": [-0.379247, -0.020925, -0.009207, -0.16419, -0.217383, 0.020478, 0.023328, 0.160562], "network.6.weight": [[-0.294536, -0.151575, 0.068544, 0.138185, -0.315163, 0.125892, -0.23007, -0.302583], [-0.148352, -0.154811, -0.008297, 0.073114, -0.265115, -0.164496, 0.276991, -0.040182], [0.311241, -0.043116, 0.099644, -0.435421, 0.062416, 0.421788, -0.665621, 0.052163], [-0.310074, -0.331169, 0.091716, -0.105178, 0.773336, 0.255292, 0.174688, -0.593687], [0.039692, -0.196073, -0.39601, 0.513855, 0.575915, 0.000679, -0.242899, -0.592157], [-0.021989, -0.103461, 0.518131, -0.21356, 0.07487, 0.224581, -0.142127, 0.043358], [-0.011782, -0.16436, -0.173533, -0.227878, 0.007494, 0.253824, 0.745782, 0.923548], [0.21282, 0.141041, 0.09536, 0.160994, 0.028164, -0.3882, -0.207591, 0.49239]], "network.6.bias": [-0.389028, -0.517186, 0.131226, 0.492981, -0.292979, -0.010838, 0.083761, -0.096793], "network.8.weight": [[-0.137726, 0.39312, 0.182557, -0.847809, -0.243457, -0.272459, 0.930779, 0.062183], [0.269555, 0.313924, 0.244109, 0.120021, 0.63553, -0.040697, -0.013889, -0.611414], [0.100668, 0.152475, 0.205478, -0.506289, 0.001647, 0.211453, 0.910036, -0.061145], [0.076847, -0.595474, -0.405433, 0.307867, 0.024958, -0.038671, 0.165388, -0.558571], [-0.087536, 0.046048, 0.301124, -0.080291, 0.131746, -0.05887, -0.295475, -0.438255], [-0.505984, -0.213786, 0.113037, 0.184483, -0.240867, 0.123203, -0.322015, -0.533184], [-0.136759, -0.240346, -0.223569, -0.154711, -0.058773, -0.12493, -0.317776, 0.29923], [-0.115777, 0.201799, -0.242451, -0.512658, -0.130133, -0.157092, 0.980573, -0.184011]], "network.8.bias": [0.101862, -0.117607, -0.209601, -0.126828, -0.399779, -0.25114, -0.028458, 0.213317], "network.10.weight": [[-0.409721, -0.79051, -0.083073, 0.098727, -0.208597, -0.193628, 0.345529, -0.455203], [0.090074, -0.453936, -0.336461, -0.055889, -0.007064, 0.091411, -0.217199, -0.039967], [-0.185261, -0.250562, -0.149713, -0.310794, -0.116257, 0.251797, 0.016701, 0.152577], [0.201754, -0.407591, 0.021336, -0.105077, 0.100267, 0.123229, 0.146604, 0.325616], [0.319589, 0.193598, 0.520731, -0.31669, -0.318919, 0.107876, -0.283256, -0.046941], [-0.098393, -0.29393, -0.09544, 0.091718, -0.309536, 0.128559, -0.307865, -0.248463], [0.216574, -0.187546, 0.311602, 0.201148, -0.201806, 0.07778, -0.224763, 0.298194], [0.276131, -0.128138, 0.078387, 0.006956, -0.116458, -0.074013, 0.253648, 0.263081]], "network.10.bias": [-0.211868, -0.11998, -0.217628, -0.28413, -0.369829, -0.325552, -0.157529, -0.318813], "network.12.weight": [[0.214864, -0.015266, 0.027294, -0.140124, -0.316361, -0.069746, -0.295879, -0.088]], "network.12.bias": [0.283847]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6807601451873779, "train_acc": 0.575, "val_loss": 0.7482194304466248, "val_acc": 0.48}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6847920417785645, "train_acc": 0.575, "val_loss": 0.6960969567298889, "val_acc": 0.48}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.7809063196182251, "train_acc": 0.505, "val_loss": 0.7308866381645203, "val_acc": 0.48}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.7009685337543488, "train_acc": 0.505, "val_loss": 0.6917961239814758, "val_acc": 0.56}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.6966330409049988, "train_acc": 0.525, "val_loss": 0.6912455558776855, "val_acc": 0.52}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.6914148330688477, "train_acc": 0.495, "val_loss": 0.6832075715065002, "val_acc": 0.52}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.6799335479736328, "train_acc": 0.545, "val_loss": 0.6388636231422424, "val_acc": 0.7}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.600528746843338, "train_acc": 0.675, "val_loss": 0.5641230940818787, "val_acc": 0.56}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.5287051200866699, "train_acc": 0.65, "val_loss": 0.5082626342773438, "val_acc": 0.76}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.4900132864713669, "train_acc": 0.74, "val_loss": 0.4962032437324524, "val_acc": 0.76}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.4921348989009857, "train_acc": 0.75, "val_loss": 0.48662295937538147, "val_acc": 0.76}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.4785608798265457, "train_acc": 0.745, "val_loss": 0.4887620508670807, "val_acc": 0.74}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["first_last_match"], "degraded_stage": {"initial_val_loss": 0.7482194304466248, "final_val_loss": 0.6960969567298889, "initial_val_acc": 0.48, "final_val_acc": 0.48, "best_val_acc": 0.48}, "improved_stage": {"initial_val_loss": 0.7308866381645203, "final_val_loss": 0.4887620508670807, "initial_val_acc": 0.48, "final_val_acc": 0.74, "best_val_acc": 0.76, "best_epoch": 8}, "improvement": 0.28, "first_improvement_epoch": 1}}
|
64
|
{"target_pattern": "increasing_pairs", "degraded_accuracy": 0.46, "improved_accuracy": 0.86, "improvement": 0.39999999999999997, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 7048, "learning_rate": 0.04198946419763812, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "increasing_pairs", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["increasing_pairs"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.010575,
-0.416981,
-0.255672,
-0.167333,
-0.042616
],
[
0.805166,
0.461446,
0.229836,
0.276951,
-0.628012
],
[
-0.280247,
-0.05447,
-0.218709,
-0.288266,
-0.117191
],
[
-0.459905,
0.086551,
-0.166768,
0.381912,
0.216749
],
[
0.069496,
-0.718745,
-0.521559,
-0.402187,
-0.133374
]
],
"network.0.bias": [
0.010853,
0.457098,
-0.149541,
0.258659,
-0.061634
],
"network.2.weight": [
[
-0.004313,
-0.090201,
0.186035,
-0.410127,
-0.441787
],
[
-0.311564,
0.968204,
-0.270729,
-0.121164,
-0.153235
],
[
0.277229,
0.421218,
-0.031427,
0.280467,
0.082875
],
[
0.121802,
0.733962,
0.257957,
-0.057738,
-0.186837
],
[
0.195656,
0.076274,
-0.263849,
-0.042594,
-0.221525
]
],
"network.2.bias": [
-0.416259,
0.085647,
0.160351,
0.199855,
-0.562219
],
"network.4.weight": [
[
0.21244,
0.795885,
-0.198362,
0.611678,
0.134051
],
[
-0.080652,
-0.393139,
0.348763,
-0.329608,
0.083121
],
[
-0.131014,
0.668089,
0.651098,
0.493089,
-0.685178
],
[
0.249221,
-0.314572,
0.205188,
-0.372485,
0.384112
],
[
-0.002303,
0.609935,
0.327403,
0.868491,
0.268291
]
],
"network.4.bias": [
0.127325,
0.765499,
0.397185,
0.376161,
-0.220594
],
"network.6.weight": [
[
-0.198957,
0.938174,
0.472179,
0.459111,
-0.686163
],
[
0.772152,
-0.260848,
0.529413,
-0.050405,
0.519004
],
[
-0.121302,
-0.193719,
0.030025,
-0.290803,
-0.269826
],
[
0.025052,
-0.133217,
-0.048296,
-0.090815,
-0.314992
],
[
0.796041,
-0.319647,
0.572764,
-0.111982,
0.804242
]
],
"network.6.bias": [
0.462224,
-0.269454,
-0.444863,
-0.533491,
-0.420745
],
"network.8.weight": [
[
0.738528,
-0.857584,
0.191088,
-0.150785,
-0.77504
]
],
"network.8.bias": [
0.028642
]
}
## Activation Signature
### 0
mean: [-1.707604, 2.599398, -1.817113, 0.582001, -3.401305]
std: [1.166765, 2.441354, 1.126255, 1.386354, 2.207991]
### 2
mean: [-1.027231, 2.577725, 1.544440, 2.120063, -0.395362]
std: [0.408822, 2.273441, 0.951098, 1.714066, 0.190483]
### 4
mean: [3.184530, -0.414977, 4.179945, -0.911814, 3.710847]
std: [2.663046, 1.139147, 2.936452, 1.159122, 3.161377]
### 6
mean: [-0.493861, 6.266849, -1.773848, -1.861925, 7.413200]
std: [1.598565, 5.309855, 1.022009, 1.033993, 6.421037]
### 8
mean: [-10.800026]
std: [9.844454]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
increasing_pairs
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.010575,
-0.416981,
-0.255672,
-0.167333,
-0.042616
],
[
0.805166,
0.461446,
0.229836,
0.276951,
-0.628012
],
[
-0.280247,
-0.05447,
-0.218709,
-0.288266,
-0.117191
],
[
-0.459905,
0.086551,
-0.166768,
0.381912,
0.216749
],
[
0.069496,
-0.718745,
-0.521559,
-0.402187,
-0.133374
]
],
"network.0.bias": [
0.010853,
0.457098,
-0.149541,
0.258659,
-0.061634
],
"network.2.weight": [
[
-0.004313,
-0.090201,
0.186035,
-0.410127,
-0.441787
],
[
-0.311564,
0.968204,
-0.270729,
-0.121164,
-0.153235
],
[
0.277229,
0.421218,
-0.031427,
0.280467,
0.082875
],
[
0.121802,
0.733962,
0.257957,
-0.057738,
-0.186837
],
[
0.195656,
0.076274,
-0.263849,
-0.042594,
-0.221525
]
],
"network.2.bias": [
-0.416259,
0.085647,
0.160351,
0.199855,
-0.562219
],
"network.4.weight": [
[
0.21244,
0.795885,
-0.198362,
0.611678,
0.134051
],
[
-0.080652,
-0.393139,
0.348763,
-0.329608,
0.083121
],
[
-0.131014,
0.668089,
0.651098,
0.493089,
-0.685178
],
[
0.249221,
-0.314572,
0.205188,
-0.372485,
0.384112
],
[
-0.002303,
0.609935,
0.327403,
0.868491,
0.268291
]
],
"network.4.bias": [
0.127325,
0.765499,
0.397185,
0.376161,
-0.220594
],
"network.6.weight": [
[
-0.198957,
0.938174,
0.472179,
0.459111,
-0.686163
],
[
0.772152,
-0.260848,
0.529413,
-0.050405,
0.519004
],
[
-0.121302,
-0.193719,
0.030025,
-0.290803,
-0.269826
],
[
0.025052,
-0.133217,
-0.048296,
-0.090815,
-0.314992
],
[
0.796041,
-0.319647,
0.572764,
-0.111982,
0.804242
]
],
"network.6.bias": [
0.462224,
-0.269454,
-0.444863,
-0.533491,
-0.420745
],
"network.8.weight": [
[
0.738528,
-0.857584,
0.191088,
-0.150785,
-0.77504
]
],
"network.8.bias": [
0.028642
]
}
## Activation Signature
### 0
mean: [-1.707604, 2.599398, -1.817113, 0.582001, -3.401305]
std: [1.166765, 2.441354, 1.126255, 1.386354, 2.207991]
### 2
mean: [-1.027231, 2.577725, 1.544440, 2.120063, -0.395362]
std: [0.408822, 2.273441, 0.951098, 1.714066, 0.190483]
### 4
mean: [3.184530, -0.414977, 4.179945, -0.911814, 3.710847]
std: [2.663046, 1.139147, 2.936452, 1.159122, 3.161377]
### 6
mean: [-0.493861, 6.266849, -1.773848, -1.861925, 7.413200]
std: [1.598565, 5.309855, 1.022009, 1.033993, 6.421037]
### 8
mean: [-10.800026]
std: [9.844454]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
increasing_pairs
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.707604169845581, "std": 1.1667653322219849}, "1": {"mean": 2.599398374557495, "std": 2.441354274749756}, "2": {"mean": -1.817112922668457, "std": 1.1262545585632324}, "3": {"mean": 0.5820006728172302, "std": 1.3863544464111328}, "4": {"mean": -3.4013054370880127, "std": 2.207991123199463}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -1.0272308588027954, "std": 0.4088220000267029}, "1": {"mean": 2.5777249336242676, "std": 2.2734410762786865}, "2": {"mean": 1.544439673423767, "std": 0.951097846031189}, "3": {"mean": 2.120062828063965, "std": 1.7140663862228394}, "4": {"mean": -0.39536213874816895, "std": 0.19048278033733368}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 3.184530019760132, "std": 2.66304612159729}, "1": {"mean": -0.41497689485549927, "std": 1.1391470432281494}, "2": {"mean": 4.179945468902588, "std": 2.9364516735076904}, "3": {"mean": -0.9118142127990723, "std": 1.1591219902038574}, "4": {"mean": 3.7108473777770996, "std": 3.161377429962158}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.4938609302043915, "std": 1.5985649824142456}, "1": {"mean": 6.266849040985107, "std": 5.3098554611206055}, "2": {"mean": -1.773848295211792, "std": 1.022009015083313}, "3": {"mean": -1.8619248867034912, "std": 1.0339930057525635}, "4": {"mean": 7.413200378417969, "std": 6.421037197113037}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -10.800025939941406, "std": 9.844453811645508}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.010575, -0.416981, -0.255672, -0.167333, -0.042616], [0.805166, 0.461446, 0.229836, 0.276951, -0.628012], [-0.280247, -0.05447, -0.218709, -0.288266, -0.117191], [-0.459905, 0.086551, -0.166768, 0.381912, 0.216749], [0.069496, -0.718745, -0.521559, -0.402187, -0.133374]], "network.0.bias": [0.010853, 0.457098, -0.149541, 0.258659, -0.061634], "network.2.weight": [[-0.004313, -0.090201, 0.186035, -0.410127, -0.441787], [-0.311564, 0.968204, -0.270729, -0.121164, -0.153235], [0.277229, 0.421218, -0.031427, 0.280467, 0.082875], [0.121802, 0.733962, 0.257957, -0.057738, -0.186837], [0.195656, 0.076274, -0.263849, -0.042594, -0.221525]], "network.2.bias": [-0.416259, 0.085647, 0.160351, 0.199855, -0.562219], "network.4.weight": [[0.21244, 0.795885, -0.198362, 0.611678, 0.134051], [-0.080652, -0.393139, 0.348763, -0.329608, 0.083121], [-0.131014, 0.668089, 0.651098, 0.493089, -0.685178], [0.249221, -0.314572, 0.205188, -0.372485, 0.384112], [-0.002303, 0.609935, 0.327403, 0.868491, 0.268291]], "network.4.bias": [0.127325, 0.765499, 0.397185, 0.376161, -0.220594], "network.6.weight": [[-0.198957, 0.938174, 0.472179, 0.459111, -0.686163], [0.772152, -0.260848, 0.529413, -0.050405, 0.519004], [-0.121302, -0.193719, 0.030025, -0.290803, -0.269826], [0.025052, -0.133217, -0.048296, -0.090815, -0.314992], [0.796041, -0.319647, 0.572764, -0.111982, 0.804242]], "network.6.bias": [0.462224, -0.269454, -0.444863, -0.533491, -0.420745], "network.8.weight": [[0.738528, -0.857584, 0.191088, -0.150785, -0.77504]], "network.8.bias": [0.028642]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6801446974277496, "train_acc": 0.59, "val_loss": 0.7243406772613525, "val_acc": 0.46}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6557194292545319, "train_acc": 0.59, "val_loss": 0.6747613549232483, "val_acc": 0.46}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.6176320016384125, "train_acc": 0.51, "val_loss": 0.5618075728416443, "val_acc": 0.46}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.5129021108150482, "train_acc": 0.595, "val_loss": 0.4646202027797699, "val_acc": 0.82}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.40901148319244385, "train_acc": 0.825, "val_loss": 0.4184323251247406, "val_acc": 0.82}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.37675556540489197, "train_acc": 0.83, "val_loss": 0.4127207100391388, "val_acc": 0.8}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.36833859980106354, "train_acc": 0.835, "val_loss": 0.3954853415489197, "val_acc": 0.82}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.3648236095905304, "train_acc": 0.83, "val_loss": 0.375579833984375, "val_acc": 0.82}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.40909792482852936, "train_acc": 0.83, "val_loss": 0.35653603076934814, "val_acc": 0.86}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.3834828734397888, "train_acc": 0.835, "val_loss": 0.42664775252342224, "val_acc": 0.8}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.3773651272058487, "train_acc": 0.83, "val_loss": 0.4472007751464844, "val_acc": 0.78}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.4005550742149353, "train_acc": 0.825, "val_loss": 0.4416261315345764, "val_acc": 0.78}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["increasing_pairs"], "degraded_stage": {"initial_val_loss": 0.7243406772613525, "final_val_loss": 0.6747613549232483, "initial_val_acc": 0.46, "final_val_acc": 0.46, "best_val_acc": 0.46}, "improved_stage": {"initial_val_loss": 0.5618075728416443, "final_val_loss": 0.4416261315345764, "initial_val_acc": 0.46, "final_val_acc": 0.78, "best_val_acc": 0.86, "best_epoch": 8}, "improvement": 0.39999999999999997, "first_improvement_epoch": 1}}
|
65
|
{"target_pattern": "increasing_pairs", "degraded_accuracy": 0.48, "improved_accuracy": 0.82, "improvement": 0.33999999999999997, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 1250, "learning_rate": 0.054650206537940234, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "increasing_pairs", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["increasing_pairs"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.934934,
-0.260922,
0.026911,
0.384486,
0.771448
],
[
0.550521,
0.639938,
-0.188503,
-0.282363,
0.070847
],
[
0.444901,
0.6927,
0.553121,
0.08242,
-0.368801
],
[
0.002388,
-0.222874,
-0.502278,
-0.059963,
-0.005599
],
[
0.836298,
0.600861,
0.29088,
0.256074,
0.169388
]
],
"network.0.bias": [
0.377179,
-0.151119,
-0.067405,
-0.024371,
0.126129
],
"network.2.weight": [
[
-0.131963,
-0.092834,
0.027991,
0.050634,
-0.410087
],
[
-0.300634,
-0.424816,
0.085312,
-0.14586,
-0.501722
],
[
-0.255985,
0.22719,
-0.497382,
0.230305,
-0.294834
],
[
0.544478,
-0.44224,
-0.258321,
-0.285701,
-0.113275
],
[
-0.593262,
0.46944,
0.564041,
0.103199,
0.353456
]
],
"network.2.bias": [
-0.396273,
-0.508675,
-0.427524,
0.07497,
0.393321
],
"network.4.weight": [
[
-0.065375,
-0.123496,
0.17113,
0.0535,
-0.233584
],
[
-0.089483,
0.311284,
-0.059541,
-0.290253,
-0.092015
],
[
0.297861,
0.228305,
0.229019,
-0.551359,
0.682265
],
[
0.264929,
-0.106503,
0.044911,
0.532812,
-0.326466
],
[
-0.152481,
-0.562398,
-0.184258,
-0.157155,
0.355652
]
],
"network.4.bias": [
-0.33845,
-0.259165,
0.4781,
0.386322,
0.292067
],
"network.6.weight": [
[
0.151877,
-0.128109,
-0.289059,
-0.56339,
-0.188439
],
[
-0.097168,
-0.050539,
0.142074,
0.03896,
-0.378356
],
[
-0.162262,
0.260598,
-0.09125,
0.236173,
-0.189133
],
[
0.035641,
-0.352103,
0.85442,
-0.595641,
0.382086
],
[
-0.049795,
-0.069921,
-0.173712,
-0.416044,
0.129959
]
],
"network.6.bias": [
-0.007426,
-0.43058,
0.333244,
0.393767,
-0.200279
],
"network.8.weight": [
[
0.441881,
-0.035055,
-0.362642,
-0.367282,
-0.148325
],
[
0.069726,
-0.246139,
0.367934,
-0.100207,
0.149073
],
[
0.525184,
-0.034271,
-0.635587,
0.771559,
-0.441328
],
[
-0.390995,
0.429394,
0.20892,
-0.231339,
-0.234511
],
[
-0.0283,
0.078757,
-0.51208,
0.823528,
-0.41718
]
],
"network.8.bias": [
-0.241859,
-0.229344,
0.037384,
-0.160797,
0.12548
],
"network.10.weight": [
[
-0.038997,
-0.060836,
-0.499819,
-0.078753,
-0.575119
]
],
"network.10.bias": [
0.136829
]
}
## Activation Signature
### 0
mean: [0.570422, 0.788188, 2.625802, -1.612223, 3.627424]
std: [2.515685, 1.785628, 2.377141, 1.162354, 2.752368]
### 2
mean: [-2.070659, -2.926377, -2.899792, -0.837457, 2.966235]
std: [1.171620, 1.744303, 1.546511, 1.981546, 3.280228]
### 4
mean: [-1.041718, -0.643708, 2.395805, -0.438423, 1.336714]
std: [0.747903, 0.270581, 2.334277, 1.237491, 1.165877]
### 6
mean: [-1.131595, -0.578266, -0.080676, 2.850813, -0.570136]
std: [0.744557, 0.131134, 0.498703, 2.533797, 0.212411]
### 8
mean: [-1.365324, -0.461862, 2.175990, -0.799112, 2.434494]
std: [0.850425, 0.311932, 2.007819, 0.605099, 2.115275]
### 10
mean: [-2.386423]
std: [2.176884]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
increasing_pairs
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.934934,
-0.260922,
0.026911,
0.384486,
0.771448
],
[
0.550521,
0.639938,
-0.188503,
-0.282363,
0.070847
],
[
0.444901,
0.6927,
0.553121,
0.08242,
-0.368801
],
[
0.002388,
-0.222874,
-0.502278,
-0.059963,
-0.005599
],
[
0.836298,
0.600861,
0.29088,
0.256074,
0.169388
]
],
"network.0.bias": [
0.377179,
-0.151119,
-0.067405,
-0.024371,
0.126129
],
"network.2.weight": [
[
-0.131963,
-0.092834,
0.027991,
0.050634,
-0.410087
],
[
-0.300634,
-0.424816,
0.085312,
-0.14586,
-0.501722
],
[
-0.255985,
0.22719,
-0.497382,
0.230305,
-0.294834
],
[
0.544478,
-0.44224,
-0.258321,
-0.285701,
-0.113275
],
[
-0.593262,
0.46944,
0.564041,
0.103199,
0.353456
]
],
"network.2.bias": [
-0.396273,
-0.508675,
-0.427524,
0.07497,
0.393321
],
"network.4.weight": [
[
-0.065375,
-0.123496,
0.17113,
0.0535,
-0.233584
],
[
-0.089483,
0.311284,
-0.059541,
-0.290253,
-0.092015
],
[
0.297861,
0.228305,
0.229019,
-0.551359,
0.682265
],
[
0.264929,
-0.106503,
0.044911,
0.532812,
-0.326466
],
[
-0.152481,
-0.562398,
-0.184258,
-0.157155,
0.355652
]
],
"network.4.bias": [
-0.33845,
-0.259165,
0.4781,
0.386322,
0.292067
],
"network.6.weight": [
[
0.151877,
-0.128109,
-0.289059,
-0.56339,
-0.188439
],
[
-0.097168,
-0.050539,
0.142074,
0.03896,
-0.378356
],
[
-0.162262,
0.260598,
-0.09125,
0.236173,
-0.189133
],
[
0.035641,
-0.352103,
0.85442,
-0.595641,
0.382086
],
[
-0.049795,
-0.069921,
-0.173712,
-0.416044,
0.129959
]
],
"network.6.bias": [
-0.007426,
-0.43058,
0.333244,
0.393767,
-0.200279
],
"network.8.weight": [
[
0.441881,
-0.035055,
-0.362642,
-0.367282,
-0.148325
],
[
0.069726,
-0.246139,
0.367934,
-0.100207,
0.149073
],
[
0.525184,
-0.034271,
-0.635587,
0.771559,
-0.441328
],
[
-0.390995,
0.429394,
0.20892,
-0.231339,
-0.234511
],
[
-0.0283,
0.078757,
-0.51208,
0.823528,
-0.41718
]
],
"network.8.bias": [
-0.241859,
-0.229344,
0.037384,
-0.160797,
0.12548
],
"network.10.weight": [
[
-0.038997,
-0.060836,
-0.499819,
-0.078753,
-0.575119
]
],
"network.10.bias": [
0.136829
]
}
## Activation Signature
### 0
mean: [0.570422, 0.788188, 2.625802, -1.612223, 3.627424]
std: [2.515685, 1.785628, 2.377141, 1.162354, 2.752368]
### 2
mean: [-2.070659, -2.926377, -2.899792, -0.837457, 2.966235]
std: [1.171620, 1.744303, 1.546511, 1.981546, 3.280228]
### 4
mean: [-1.041718, -0.643708, 2.395805, -0.438423, 1.336714]
std: [0.747903, 0.270581, 2.334277, 1.237491, 1.165877]
### 6
mean: [-1.131595, -0.578266, -0.080676, 2.850813, -0.570136]
std: [0.744557, 0.131134, 0.498703, 2.533797, 0.212411]
### 8
mean: [-1.365324, -0.461862, 2.175990, -0.799112, 2.434494]
std: [0.850425, 0.311932, 2.007819, 0.605099, 2.115275]
### 10
mean: [-2.386423]
std: [2.176884]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
increasing_pairs
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.5704216957092285, "std": 2.5156853199005127}, "1": {"mean": 0.7881875038146973, "std": 1.785628080368042}, "2": {"mean": 2.6258020401000977, "std": 2.377140998840332}, "3": {"mean": -1.6122227907180786, "std": 1.1623536348342896}, "4": {"mean": 3.6274235248565674, "std": 2.752368450164795}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -2.0706591606140137, "std": 1.1716197729110718}, "1": {"mean": -2.9263765811920166, "std": 1.74430251121521}, "2": {"mean": -2.899792432785034, "std": 1.546511173248291}, "3": {"mean": -0.8374572992324829, "std": 1.9815458059310913}, "4": {"mean": 2.966235399246216, "std": 3.2802281379699707}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -1.0417180061340332, "std": 0.7479032874107361}, "1": {"mean": -0.6437076926231384, "std": 0.2705812454223633}, "2": {"mean": 2.3958053588867188, "std": 2.3342766761779785}, "3": {"mean": -0.43842318654060364, "std": 1.2374910116195679}, "4": {"mean": 1.3367140293121338, "std": 1.1658765077590942}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -1.1315948963165283, "std": 0.7445568442344666}, "1": {"mean": -0.578265905380249, "std": 0.13113422691822052}, "2": {"mean": -0.08067597448825836, "std": 0.4987033009529114}, "3": {"mean": 2.8508126735687256, "std": 2.533797025680542}, "4": {"mean": -0.5701364874839783, "std": 0.21241146326065063}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.3653240203857422, "std": 0.8504253625869751}, "1": {"mean": -0.4618624448776245, "std": 0.3119317293167114}, "2": {"mean": 2.175990343093872, "std": 2.0078189373016357}, "3": {"mean": -0.79911208152771, "std": 0.6050994992256165}, "4": {"mean": 2.4344937801361084, "std": 2.1152749061584473}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -2.386423349380493, "std": 2.176884174346924}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.934934, -0.260922, 0.026911, 0.384486, 0.771448], [0.550521, 0.639938, -0.188503, -0.282363, 0.070847], [0.444901, 0.6927, 0.553121, 0.08242, -0.368801], [0.002388, -0.222874, -0.502278, -0.059963, -0.005599], [0.836298, 0.600861, 0.29088, 0.256074, 0.169388]], "network.0.bias": [0.377179, -0.151119, -0.067405, -0.024371, 0.126129], "network.2.weight": [[-0.131963, -0.092834, 0.027991, 0.050634, -0.410087], [-0.300634, -0.424816, 0.085312, -0.14586, -0.501722], [-0.255985, 0.22719, -0.497382, 0.230305, -0.294834], [0.544478, -0.44224, -0.258321, -0.285701, -0.113275], [-0.593262, 0.46944, 0.564041, 0.103199, 0.353456]], "network.2.bias": [-0.396273, -0.508675, -0.427524, 0.07497, 0.393321], "network.4.weight": [[-0.065375, -0.123496, 0.17113, 0.0535, -0.233584], [-0.089483, 0.311284, -0.059541, -0.290253, -0.092015], [0.297861, 0.228305, 0.229019, -0.551359, 0.682265], [0.264929, -0.106503, 0.044911, 0.532812, -0.326466], [-0.152481, -0.562398, -0.184258, -0.157155, 0.355652]], "network.4.bias": [-0.33845, -0.259165, 0.4781, 0.386322, 0.292067], "network.6.weight": [[0.151877, -0.128109, -0.289059, -0.56339, -0.188439], [-0.097168, -0.050539, 0.142074, 0.03896, -0.378356], [-0.162262, 0.260598, -0.09125, 0.236173, -0.189133], [0.035641, -0.352103, 0.85442, -0.595641, 0.382086], [-0.049795, -0.069921, -0.173712, -0.416044, 0.129959]], "network.6.bias": [-0.007426, -0.43058, 0.333244, 0.393767, -0.200279], "network.8.weight": [[0.441881, -0.035055, -0.362642, -0.367282, -0.148325], [0.069726, -0.246139, 0.367934, -0.100207, 0.149073], [0.525184, -0.034271, -0.635587, 0.771559, -0.441328], [-0.390995, 0.429394, 0.20892, -0.231339, -0.234511], [-0.0283, 0.078757, -0.51208, 0.823528, -0.41718]], "network.8.bias": [-0.241859, -0.229344, 0.037384, -0.160797, 0.12548], "network.10.weight": [[-0.038997, -0.060836, -0.499819, -0.078753, -0.575119]], "network.10.bias": [0.136829]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7019431293010712, "train_acc": 0.42, "val_loss": 0.6993151903152466, "val_acc": 0.48}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6928939819335938, "train_acc": 0.575, "val_loss": 0.7065917253494263, "val_acc": 0.48}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6655651032924652, "train_acc": 0.575, "val_loss": 0.6767085194587708, "val_acc": 0.48}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.649448573589325, "train_acc": 0.575, "val_loss": 0.6455482244491577, "val_acc": 0.48}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6085567474365234, "train_acc": 0.505, "val_loss": 0.5565104484558105, "val_acc": 0.58}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.4966808706521988, "train_acc": 0.725, "val_loss": 0.4961193799972534, "val_acc": 0.82}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.44247572124004364, "train_acc": 0.845, "val_loss": 0.4579848349094391, "val_acc": 0.82}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.42275112867355347, "train_acc": 0.845, "val_loss": 0.4436541795730591, "val_acc": 0.82}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.41060496866703033, "train_acc": 0.845, "val_loss": 0.45697566866874695, "val_acc": 0.8}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.4132508486509323, "train_acc": 0.84, "val_loss": 0.4210212826728821, "val_acc": 0.82}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.39685556292533875, "train_acc": 0.845, "val_loss": 0.4416176676750183, "val_acc": 0.8}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.38593992590904236, "train_acc": 0.83, "val_loss": 0.43867743015289307, "val_acc": 0.78}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.3667055815458298, "train_acc": 0.845, "val_loss": 0.4282516539096832, "val_acc": 0.8}], "summary": {"total_epochs": 13, "degraded_epochs": 4, "improved_epochs": 9, "patterns": ["increasing_pairs"], "degraded_stage": {"initial_val_loss": 0.6993151903152466, "final_val_loss": 0.6455482244491577, "initial_val_acc": 0.48, "final_val_acc": 0.48, "best_val_acc": 0.48}, "improved_stage": {"initial_val_loss": 0.5565104484558105, "final_val_loss": 0.4282516539096832, "initial_val_acc": 0.58, "final_val_acc": 0.8, "best_val_acc": 0.82, "best_epoch": 5}, "improvement": 0.33999999999999997, "first_improvement_epoch": 3}}
|
66
|
{"target_pattern": "decreasing_pairs", "degraded_accuracy": 0.78, "improved_accuracy": 0.98, "improvement": 0.19999999999999996, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 6942, "learning_rate": 0.039815314088510245, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "decreasing_pairs", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["decreasing_pairs"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.227079,
0.557429,
0.459927,
0.619559,
0.328868
],
[
0.251918,
0.567443,
0.003999,
-0.130545,
-0.0728
],
[
-0.013836,
-0.395478,
0.435463,
-0.004783,
0.251563
],
[
-0.66849,
0.060497,
-0.613641,
-0.278684,
-0.113068
],
[
-0.214232,
0.296091,
-0.082227,
-0.554973,
0.482054
],
[
0.747576,
0.104405,
-0.176079,
0.046555,
0.07986
],
[
0.745304,
0.495255,
0.098443,
-0.254022,
-0.148871
]
],
"network.0.bias": [
0.175678,
-0.305302,
0.456649,
0.210862,
0.327502,
-0.396519,
-0.26923
],
"network.2.weight": [
[
0.002902,
0.352366,
-0.13024,
0.269893,
-0.257469,
0.537997,
0.387234
],
[
-0.171713,
0.269443,
0.098579,
0.259447,
-0.106689,
-0.460497,
-0.041612
],
[
0.313692,
-0.216112,
0.325031,
-0.110411,
0.316096,
0.07553,
-0.379888
],
[
-0.057523,
0.176522,
0.053177,
0.418469,
0.063654,
-0.121238,
-0.324123
],
[
-0.230757,
0.217832,
0.081936,
-0.19183,
-0.05248,
-0.189832,
0.067506
],
[
0.000925,
0.093627,
-0.013219,
0.526605,
0.081738,
-0.250024,
-0.251646
],
[
0.486127,
-0.044783,
0.179021,
0.222453,
-0.33921,
0.01647,
0.487362
]
],
"network.2.bias": [
-0.133546,
0.124135,
0.537916,
0.247606,
-0.114213,
-0.274197,
0.262121
],
"network.4.weight": [
[
-0.157782,
-0.416613,
0.058463,
-0.0054,
-0.397707,
-0.108901,
0.360425
],
[
-0.092422,
-0.121135,
-0.208155,
0.219268,
0.65827,
0.089386,
-0.005727
],
[
-0.043614,
-0.573353,
0.2829,
0.183651,
-0.412439,
-0.122446,
-0.135007
],
[
0.443846,
0.452174,
-0.563017,
0.253408,
-0.062788,
0.403539,
0.462646
],
[
0.525771,
0.031293,
-0.026374,
0.461457,
-0.641263,
-0.418189,
0.336416
],
[
-0.47508,
-0.064274,
-0.162915,
-0.61751,
-0.009595,
-0.355449,
0.120376
],
[
-0.234354,
-0.544429,
0.507238,
-0.443175,
-0.143335,
-0.578929,
0.189727
]
],
"network.4.bias": [
0.378614,
-0.241372,
0.214703,
-0.324607,
0.273866,
-0.141985,
0.595128
],
"network.6.weight": [
[
0.258041,
-0.113237,
-0.022881,
0.1001,
0.158739,
-0.384071,
0.368342
],
[
0.116184,
0.113253,
0.160378,
-0.384128,
0.202009,
0.015379,
0.37673
],
[
0.23054,
-0.373386,
-0.311114,
0.397758,
0.412934,
0.408562,
-0.173394
],
[
0.572271,
-0.093633,
0.371864,
-0.094295,
-0.219991,
0.017366,
0.52888
],
[
0.209994,
-0.264606,
-0.208136,
-0.274102,
-0.003631,
0.020816,
0.003267
],
[
0.323184,
-0.457224,
-0.383527,
-0.44369,
0.167916,
-0.692301,
0.496306
],
[
-0.129863,
-0.479706,
0.329227,
-0.496933,
0.363664,
-0.026145,
0.322177
]
],
"network.6.bias": [
-0.18436,
-0.036902,
-0.117634,
0.515092,
0.253492,
0.132924,
0.084228
],
"network.8.weight": [
[
0.182531,
-0.186984,
0.54148,
0.298407,
-0.551389,
-0.410223,
-0.434515
],
[
0.294797,
0.287417,
0.06193,
0.179031,
0.396167,
0.534031,
0.076609
],
[
0.046817,
0.161556,
-0.173526,
-0.012845,
0.024732,
0.036908,
-0.025962
],
[
0.109256,
-0.031018,
0.607769,
-0.216239,
-0.525682,
-0.329546,
-0.112971
],
[
-0.163059,
0.129435,
-0.054356,
0.43096,
0.286194,
0.416721,
0.458263
],
[
0.145036,
0.20238,
-0.355236,
0.424101,
0.402418,
0.326193,
0.451333
],
[
0.423968,
0.228516,
-0.42569,
0.140628,
-0.123891,
0.578659,
0.168807
]
],
"network.8.bias": [
-0.086348,
-0.192533,
0.03542,
-0.049355,
0.259854,
0.397353,
0.630856
],
"network.10.weight": [
[
0.286402,
-0.125689,
0.053193,
0.527414,
-0.280203,
-0.622101,
-0.41049
]
],
"network.10.bias": [
-0.097602
]
}
## Activation Signature
### 0
mean: [4.168976, 0.680825, 0.928420, -2.529698, -0.166891, 0.554565, 1.038017]
std: [2.551262, 1.274918, 1.165092, 2.158403, 1.234872, 1.544292, 2.079095]
### 2
mean: [0.783606, -0.710804, 1.674313, -0.277742, -0.896056, -0.683485, 2.902005]
std: [1.875650, 0.762436, 1.015240, 0.691242, 0.509217, 0.728648, 1.939938]
### 4
mean: [1.457834, -0.761150, 0.364543, 0.426526, 1.804627, -0.463112, 1.861303]
std: [0.560574, 0.215640, 0.475520, 1.800739, 1.478587, 0.562603, 0.797951]
### 6
mean: [1.211343, 0.892971, 0.802829, 1.925710, 0.315977, 1.437920, 0.927893]
std: [0.579089, 0.637619, 1.456361, 0.898871, 0.358862, 0.751137, 0.597107]
### 8
mean: [-0.104737, 1.638331, 0.089537, -0.546307, 1.971664, 2.152023, 2.104776]
std: [1.177183, 0.853391, 0.304008, 1.372430, 1.106499, 1.425482, 1.204931]
### 10
mean: [-2.837162]
std: [2.274658]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
decreasing_pairs
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.227079,
0.557429,
0.459927,
0.619559,
0.328868
],
[
0.251918,
0.567443,
0.003999,
-0.130545,
-0.0728
],
[
-0.013836,
-0.395478,
0.435463,
-0.004783,
0.251563
],
[
-0.66849,
0.060497,
-0.613641,
-0.278684,
-0.113068
],
[
-0.214232,
0.296091,
-0.082227,
-0.554973,
0.482054
],
[
0.747576,
0.104405,
-0.176079,
0.046555,
0.07986
],
[
0.745304,
0.495255,
0.098443,
-0.254022,
-0.148871
]
],
"network.0.bias": [
0.175678,
-0.305302,
0.456649,
0.210862,
0.327502,
-0.396519,
-0.26923
],
"network.2.weight": [
[
0.002902,
0.352366,
-0.13024,
0.269893,
-0.257469,
0.537997,
0.387234
],
[
-0.171713,
0.269443,
0.098579,
0.259447,
-0.106689,
-0.460497,
-0.041612
],
[
0.313692,
-0.216112,
0.325031,
-0.110411,
0.316096,
0.07553,
-0.379888
],
[
-0.057523,
0.176522,
0.053177,
0.418469,
0.063654,
-0.121238,
-0.324123
],
[
-0.230757,
0.217832,
0.081936,
-0.19183,
-0.05248,
-0.189832,
0.067506
],
[
0.000925,
0.093627,
-0.013219,
0.526605,
0.081738,
-0.250024,
-0.251646
],
[
0.486127,
-0.044783,
0.179021,
0.222453,
-0.33921,
0.01647,
0.487362
]
],
"network.2.bias": [
-0.133546,
0.124135,
0.537916,
0.247606,
-0.114213,
-0.274197,
0.262121
],
"network.4.weight": [
[
-0.157782,
-0.416613,
0.058463,
-0.0054,
-0.397707,
-0.108901,
0.360425
],
[
-0.092422,
-0.121135,
-0.208155,
0.219268,
0.65827,
0.089386,
-0.005727
],
[
-0.043614,
-0.573353,
0.2829,
0.183651,
-0.412439,
-0.122446,
-0.135007
],
[
0.443846,
0.452174,
-0.563017,
0.253408,
-0.062788,
0.403539,
0.462646
],
[
0.525771,
0.031293,
-0.026374,
0.461457,
-0.641263,
-0.418189,
0.336416
],
[
-0.47508,
-0.064274,
-0.162915,
-0.61751,
-0.009595,
-0.355449,
0.120376
],
[
-0.234354,
-0.544429,
0.507238,
-0.443175,
-0.143335,
-0.578929,
0.189727
]
],
"network.4.bias": [
0.378614,
-0.241372,
0.214703,
-0.324607,
0.273866,
-0.141985,
0.595128
],
"network.6.weight": [
[
0.258041,
-0.113237,
-0.022881,
0.1001,
0.158739,
-0.384071,
0.368342
],
[
0.116184,
0.113253,
0.160378,
-0.384128,
0.202009,
0.015379,
0.37673
],
[
0.23054,
-0.373386,
-0.311114,
0.397758,
0.412934,
0.408562,
-0.173394
],
[
0.572271,
-0.093633,
0.371864,
-0.094295,
-0.219991,
0.017366,
0.52888
],
[
0.209994,
-0.264606,
-0.208136,
-0.274102,
-0.003631,
0.020816,
0.003267
],
[
0.323184,
-0.457224,
-0.383527,
-0.44369,
0.167916,
-0.692301,
0.496306
],
[
-0.129863,
-0.479706,
0.329227,
-0.496933,
0.363664,
-0.026145,
0.322177
]
],
"network.6.bias": [
-0.18436,
-0.036902,
-0.117634,
0.515092,
0.253492,
0.132924,
0.084228
],
"network.8.weight": [
[
0.182531,
-0.186984,
0.54148,
0.298407,
-0.551389,
-0.410223,
-0.434515
],
[
0.294797,
0.287417,
0.06193,
0.179031,
0.396167,
0.534031,
0.076609
],
[
0.046817,
0.161556,
-0.173526,
-0.012845,
0.024732,
0.036908,
-0.025962
],
[
0.109256,
-0.031018,
0.607769,
-0.216239,
-0.525682,
-0.329546,
-0.112971
],
[
-0.163059,
0.129435,
-0.054356,
0.43096,
0.286194,
0.416721,
0.458263
],
[
0.145036,
0.20238,
-0.355236,
0.424101,
0.402418,
0.326193,
0.451333
],
[
0.423968,
0.228516,
-0.42569,
0.140628,
-0.123891,
0.578659,
0.168807
]
],
"network.8.bias": [
-0.086348,
-0.192533,
0.03542,
-0.049355,
0.259854,
0.397353,
0.630856
],
"network.10.weight": [
[
0.286402,
-0.125689,
0.053193,
0.527414,
-0.280203,
-0.622101,
-0.41049
]
],
"network.10.bias": [
-0.097602
]
}
## Activation Signature
### 0
mean: [4.168976, 0.680825, 0.928420, -2.529698, -0.166891, 0.554565, 1.038017]
std: [2.551262, 1.274918, 1.165092, 2.158403, 1.234872, 1.544292, 2.079095]
### 2
mean: [0.783606, -0.710804, 1.674313, -0.277742, -0.896056, -0.683485, 2.902005]
std: [1.875650, 0.762436, 1.015240, 0.691242, 0.509217, 0.728648, 1.939938]
### 4
mean: [1.457834, -0.761150, 0.364543, 0.426526, 1.804627, -0.463112, 1.861303]
std: [0.560574, 0.215640, 0.475520, 1.800739, 1.478587, 0.562603, 0.797951]
### 6
mean: [1.211343, 0.892971, 0.802829, 1.925710, 0.315977, 1.437920, 0.927893]
std: [0.579089, 0.637619, 1.456361, 0.898871, 0.358862, 0.751137, 0.597107]
### 8
mean: [-0.104737, 1.638331, 0.089537, -0.546307, 1.971664, 2.152023, 2.104776]
std: [1.177183, 0.853391, 0.304008, 1.372430, 1.106499, 1.425482, 1.204931]
### 10
mean: [-2.837162]
std: [2.274658]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
decreasing_pairs
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 4.168975830078125, "std": 2.551262378692627}, "1": {"mean": 0.6808246374130249, "std": 1.2749183177947998}, "2": {"mean": 0.9284204244613647, "std": 1.1650919914245605}, "3": {"mean": -2.529697895050049, "std": 2.158403158187866}, "4": {"mean": -0.166890949010849, "std": 1.234871745109558}, "5": {"mean": 0.5545652508735657, "std": 1.5442919731140137}, "6": {"mean": 1.0380167961120605, "std": 2.0790953636169434}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.7836060523986816, "std": 1.8756498098373413}, "1": {"mean": -0.7108038663864136, "std": 0.7624362111091614}, "2": {"mean": 1.674312710762024, "std": 1.015239953994751}, "3": {"mean": -0.27774158120155334, "std": 0.6912415027618408}, "4": {"mean": -0.8960556983947754, "std": 0.5092165470123291}, "5": {"mean": -0.6834850311279297, "std": 0.7286484241485596}, "6": {"mean": 2.9020049571990967, "std": 1.939937710762024}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 1.457834005355835, "std": 0.5605736970901489}, "1": {"mean": -0.7611498236656189, "std": 0.2156396359205246}, "2": {"mean": 0.36454349756240845, "std": 0.47551989555358887}, "3": {"mean": 0.426525741815567, "std": 1.8007385730743408}, "4": {"mean": 1.8046271800994873, "std": 1.47858726978302}, "5": {"mean": -0.46311241388320923, "std": 0.5626025199890137}, "6": {"mean": 1.8613032102584839, "std": 0.7979506254196167}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 1.2113431692123413, "std": 0.5790887475013733}, "1": {"mean": 0.8929712772369385, "std": 0.6376193165779114}, "2": {"mean": 0.8028290867805481, "std": 1.4563614130020142}, "3": {"mean": 1.9257102012634277, "std": 0.8988707661628723}, "4": {"mean": 0.31597667932510376, "std": 0.3588617444038391}, "5": {"mean": 1.4379204511642456, "std": 0.7511366009712219}, "6": {"mean": 0.9278931617736816, "std": 0.5971073508262634}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.10473745316267014, "std": 1.1771833896636963}, "1": {"mean": 1.6383310556411743, "std": 0.8533912897109985}, "2": {"mean": 0.08953733742237091, "std": 0.30400776863098145}, "3": {"mean": -0.546306848526001, "std": 1.372430443763733}, "4": {"mean": 1.971664309501648, "std": 1.1064989566802979}, "5": {"mean": 2.1520230770111084, "std": 1.4254822731018066}, "6": {"mean": 2.10477614402771, "std": 1.2049307823181152}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -2.837162494659424, "std": 2.274658203125}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.227079, 0.557429, 0.459927, 0.619559, 0.328868], [0.251918, 0.567443, 0.003999, -0.130545, -0.0728], [-0.013836, -0.395478, 0.435463, -0.004783, 0.251563], [-0.66849, 0.060497, -0.613641, -0.278684, -0.113068], [-0.214232, 0.296091, -0.082227, -0.554973, 0.482054], [0.747576, 0.104405, -0.176079, 0.046555, 0.07986], [0.745304, 0.495255, 0.098443, -0.254022, -0.148871]], "network.0.bias": [0.175678, -0.305302, 0.456649, 0.210862, 0.327502, -0.396519, -0.26923], "network.2.weight": [[0.002902, 0.352366, -0.13024, 0.269893, -0.257469, 0.537997, 0.387234], [-0.171713, 0.269443, 0.098579, 0.259447, -0.106689, -0.460497, -0.041612], [0.313692, -0.216112, 0.325031, -0.110411, 0.316096, 0.07553, -0.379888], [-0.057523, 0.176522, 0.053177, 0.418469, 0.063654, -0.121238, -0.324123], [-0.230757, 0.217832, 0.081936, -0.19183, -0.05248, -0.189832, 0.067506], [0.000925, 0.093627, -0.013219, 0.526605, 0.081738, -0.250024, -0.251646], [0.486127, -0.044783, 0.179021, 0.222453, -0.33921, 0.01647, 0.487362]], "network.2.bias": [-0.133546, 0.124135, 0.537916, 0.247606, -0.114213, -0.274197, 0.262121], "network.4.weight": [[-0.157782, -0.416613, 0.058463, -0.0054, -0.397707, -0.108901, 0.360425], [-0.092422, -0.121135, -0.208155, 0.219268, 0.65827, 0.089386, -0.005727], [-0.043614, -0.573353, 0.2829, 0.183651, -0.412439, -0.122446, -0.135007], [0.443846, 0.452174, -0.563017, 0.253408, -0.062788, 0.403539, 0.462646], [0.525771, 0.031293, -0.026374, 0.461457, -0.641263, -0.418189, 0.336416], [-0.47508, -0.064274, -0.162915, -0.61751, -0.009595, -0.355449, 0.120376], [-0.234354, -0.544429, 0.507238, -0.443175, -0.143335, -0.578929, 0.189727]], "network.4.bias": [0.378614, -0.241372, 0.214703, -0.324607, 0.273866, -0.141985, 0.595128], "network.6.weight": [[0.258041, -0.113237, -0.022881, 0.1001, 0.158739, -0.384071, 0.368342], [0.116184, 0.113253, 0.160378, -0.384128, 0.202009, 0.015379, 0.37673], [0.23054, -0.373386, -0.311114, 0.397758, 0.412934, 0.408562, -0.173394], [0.572271, -0.093633, 0.371864, -0.094295, -0.219991, 0.017366, 0.52888], [0.209994, -0.264606, -0.208136, -0.274102, -0.003631, 0.020816, 0.003267], [0.323184, -0.457224, -0.383527, -0.44369, 0.167916, -0.692301, 0.496306], [-0.129863, -0.479706, 0.329227, -0.496933, 0.363664, -0.026145, 0.322177]], "network.6.bias": [-0.18436, -0.036902, -0.117634, 0.515092, 0.253492, 0.132924, 0.084228], "network.8.weight": [[0.182531, -0.186984, 0.54148, 0.298407, -0.551389, -0.410223, -0.434515], [0.294797, 0.287417, 0.06193, 0.179031, 0.396167, 0.534031, 0.076609], [0.046817, 0.161556, -0.173526, -0.012845, 0.024732, 0.036908, -0.025962], [0.109256, -0.031018, 0.607769, -0.216239, -0.525682, -0.329546, -0.112971], [-0.163059, 0.129435, -0.054356, 0.43096, 0.286194, 0.416721, 0.458263], [0.145036, 0.20238, -0.355236, 0.424101, 0.402418, 0.326193, 0.451333], [0.423968, 0.228516, -0.42569, 0.140628, -0.123891, 0.578659, 0.168807]], "network.8.bias": [-0.086348, -0.192533, 0.03542, -0.049355, 0.259854, 0.397353, 0.630856], "network.10.weight": [[0.286402, -0.125689, 0.053193, 0.527414, -0.280203, -0.622101, -0.41049]], "network.10.bias": [-0.097602]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6840048432350159, "train_acc": 0.59, "val_loss": 0.7063709497451782, "val_acc": 0.44}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.654735803604126, "train_acc": 0.59, "val_loss": 0.5990546941757202, "val_acc": 0.78}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.5230997651815414, "train_acc": 0.795, "val_loss": 0.3946112394332886, "val_acc": 0.82}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.3484022915363312, "train_acc": 0.83, "val_loss": 0.16685524582862854, "val_acc": 0.94}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.31337377429008484, "train_acc": 0.865, "val_loss": 0.2188466191291809, "val_acc": 0.92}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.2570873647928238, "train_acc": 0.885, "val_loss": 0.09470602124929428, "val_acc": 0.98}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.19996201992034912, "train_acc": 0.91, "val_loss": 0.07708209753036499, "val_acc": 0.98}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.20194578170776367, "train_acc": 0.91, "val_loss": 0.07248193770647049, "val_acc": 0.98}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.20723528414964676, "train_acc": 0.91, "val_loss": 0.07788137346506119, "val_acc": 0.98}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.19362574070692062, "train_acc": 0.92, "val_loss": 0.08922076225280762, "val_acc": 0.98}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.20131225883960724, "train_acc": 0.925, "val_loss": 0.07576259225606918, "val_acc": 0.98}], "summary": {"total_epochs": 11, "degraded_epochs": 2, "improved_epochs": 9, "patterns": ["decreasing_pairs"], "degraded_stage": {"initial_val_loss": 0.7063709497451782, "final_val_loss": 0.5990546941757202, "initial_val_acc": 0.44, "final_val_acc": 0.78, "best_val_acc": 0.78}, "improved_stage": {"initial_val_loss": 0.3946112394332886, "final_val_loss": 0.07576259225606918, "initial_val_acc": 0.82, "final_val_acc": 0.98, "best_val_acc": 0.98, "best_epoch": 5}, "improvement": 0.19999999999999996, "first_improvement_epoch": 1}}
|
67
|
{"target_pattern": "decreasing_pairs", "degraded_accuracy": 0.44, "improved_accuracy": 0.98, "improvement": 0.54, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 9339, "learning_rate": 0.028907412749536947, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "decreasing_pairs", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["decreasing_pairs"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.598627,
-0.168119,
0.269685,
0.461029,
-0.231523
],
[
-0.624035,
0.196546,
-0.172766,
0.323372,
0.19696
],
[
-0.154035,
-0.006033,
-0.454197,
0.379071,
0.104111
],
[
-0.484359,
0.054547,
0.439236,
-0.164646,
0.584382
],
[
-0.713141,
0.254444,
0.155965,
0.591237,
0.210261
],
[
-0.226956,
-0.234816,
-0.136635,
0.075597,
0.093668
]
],
"network.0.bias": [
0.692758,
0.673174,
0.31084,
-0.016093,
0.065971,
-0.121228
],
"network.2.weight": [
[
0.57438,
0.576218,
0.53029,
0.219544,
0.262341,
-0.581332
],
[
-0.65949,
-0.634313,
-0.031302,
0.130836,
-0.218025,
0.166749
],
[
0.380605,
0.349053,
0.220438,
0.689507,
0.73411,
-0.12643
],
[
-0.467985,
-0.553698,
-0.191315,
-0.032273,
0.400655,
0.511745
],
[
-0.519015,
-0.107456,
-0.602622,
-0.203195,
0.175333,
0.626074
],
[
-0.112379,
-0.378033,
-0.263262,
0.544732,
-0.401194,
-0.188167
]
],
"network.2.bias": [
0.171517,
-0.74109,
0.266068,
0.282902,
0.657812,
-0.372192
],
"network.4.weight": [
[
0.517248,
0.366811,
0.474461,
-0.586477,
-0.538393,
0.189903
],
[
0.369881,
-0.207982,
0.769733,
-0.650688,
-0.364273,
0.324629
],
[
-0.124064,
-0.340113,
-0.329222,
0.634395,
0.708403,
-0.586608
],
[
-0.316782,
-0.402439,
0.135556,
0.832068,
0.078744,
-0.610858
],
[
-0.55672,
-0.348199,
0.166331,
-0.015835,
0.630139,
-0.577851
],
[
0.204231,
0.022286,
-0.288439,
-0.310137,
-0.05585,
0.384916
]
],
"network.4.bias": [
-0.145597,
-0.015258,
0.788115,
0.272858,
0.401335,
-0.374371
],
"network.6.weight": [
[
-0.52,
-0.354857,
0.705469,
0.559468,
0.480621,
-0.62921
],
[
-0.720518,
-0.147595,
0.595432,
0.547071,
0.48362,
-0.6471
],
[
0.48171,
0.400626,
-0.254656,
-0.476169,
0.153746,
0.083138
],
[
0.201516,
0.737626,
-0.458875,
-0.174434,
-0.574713,
-0.132551
],
[
-0.492453,
-0.054087,
0.389742,
0.732779,
0.698992,
0.179092
],
[
0.193158,
0.820914,
-0.449294,
-0.231688,
-0.664975,
0.024184
]
],
"network.6.bias": [
0.617031,
0.570815,
0.403173,
0.05844,
0.334374,
0.391823
],
"network.8.weight": [
[
0.64799,
0.656925,
-0.028934,
-0.296338,
0.315623,
-0.643567
],
[
-0.292315,
-0.23336,
0.185667,
0.368147,
-0.705969,
-0.11522
],
[
0.884893,
0.457297,
-0.213295,
-0.652001,
0.450712,
-0.227509
],
[
-0.502506,
-0.152446,
0.620327,
0.587259,
-0.665021,
0.473446
],
[
-0.729536,
-0.377126,
0.333751,
0.176156,
-0.398551,
0.333992
],
[
-0.463589,
-0.644534,
0.68104,
0.296006,
-0.548959,
0.350463
]
],
"network.8.bias": [
0.386602,
0.2863,
0.340354,
0.063636,
0.280192,
0.584021
],
"network.10.weight": [
[
-0.233883,
-0.482278,
-0.333919,
0.67735,
0.535556,
0.073066
],
[
-0.047806,
-0.64355,
-0.401765,
0.058635,
0.554525,
0.502979
],
[
0.605734,
-0.43273,
0.492164,
-0.128911,
-0.366687,
-0.22068
],
[
-0.454421,
0.343445,
-0.202428,
0.617927,
0.623635,
0.620928
],
[
-0.298212,
0.090822,
-0.541279,
0.300519,
0.684595,
0.650648
],
[
-0.15952,
0.192373,
-0.376841,
0.116675,
0.348651,
0.4461
]
],
"network.10.bias": [
0.168822,
0.029012,
0.479378,
0.381675,
0.501922,
0.240955
],
"network.12.weight": [
[
-0.484905,
-0.474798,
0.560042,
-0.617076,
-0.672929,
-0.580633
]
],
"network.12.bias": [
-0.03763
]
}
## Activation Signature
### 0
mean: [0.907898, 0.830229, 0.100692, 0.765726, 1.495150, -0.839327]
std: [1.620708, 1.561799, 1.294308, 1.505011, 1.992525, 0.840829]
### 2
mean: [2.390841, -2.453026, 3.073430, -0.333625, -0.279981, -1.181736]
std: [1.930530, 1.658062, 2.362043, 0.591053, 0.791181, 1.374141]
### 4
mean: [2.443803, 3.179261, -0.428703, -0.032431, -0.320981, -0.774430]
std: [2.313431, 2.670312, 1.282489, 0.531289, 0.891571, 0.327822]
### 6
mean: [-1.418918, -1.329442, 2.764119, 2.721927, -0.844673, 3.262806]
std: [2.642578, 2.496546, 2.311316, 2.775897, 1.761795, 3.009281]
### 8
mean: [-1.932788, 0.951376, -2.033884, 4.456558, 2.119523, 3.726937]
std: [3.706806, 1.779521, 3.954406, 5.062441, 3.136961, 4.338863]
### 10
mean: [4.103285, 2.658182, -1.666453, 7.580576, 5.914869, 3.394180]
std: [4.597656, 3.166248, 4.073865, 7.642177, 6.410896, 3.830112]
### 12
mean: [-14.522885]
std: [14.192790]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
decreasing_pairs
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.598627,
-0.168119,
0.269685,
0.461029,
-0.231523
],
[
-0.624035,
0.196546,
-0.172766,
0.323372,
0.19696
],
[
-0.154035,
-0.006033,
-0.454197,
0.379071,
0.104111
],
[
-0.484359,
0.054547,
0.439236,
-0.164646,
0.584382
],
[
-0.713141,
0.254444,
0.155965,
0.591237,
0.210261
],
[
-0.226956,
-0.234816,
-0.136635,
0.075597,
0.093668
]
],
"network.0.bias": [
0.692758,
0.673174,
0.31084,
-0.016093,
0.065971,
-0.121228
],
"network.2.weight": [
[
0.57438,
0.576218,
0.53029,
0.219544,
0.262341,
-0.581332
],
[
-0.65949,
-0.634313,
-0.031302,
0.130836,
-0.218025,
0.166749
],
[
0.380605,
0.349053,
0.220438,
0.689507,
0.73411,
-0.12643
],
[
-0.467985,
-0.553698,
-0.191315,
-0.032273,
0.400655,
0.511745
],
[
-0.519015,
-0.107456,
-0.602622,
-0.203195,
0.175333,
0.626074
],
[
-0.112379,
-0.378033,
-0.263262,
0.544732,
-0.401194,
-0.188167
]
],
"network.2.bias": [
0.171517,
-0.74109,
0.266068,
0.282902,
0.657812,
-0.372192
],
"network.4.weight": [
[
0.517248,
0.366811,
0.474461,
-0.586477,
-0.538393,
0.189903
],
[
0.369881,
-0.207982,
0.769733,
-0.650688,
-0.364273,
0.324629
],
[
-0.124064,
-0.340113,
-0.329222,
0.634395,
0.708403,
-0.586608
],
[
-0.316782,
-0.402439,
0.135556,
0.832068,
0.078744,
-0.610858
],
[
-0.55672,
-0.348199,
0.166331,
-0.015835,
0.630139,
-0.577851
],
[
0.204231,
0.022286,
-0.288439,
-0.310137,
-0.05585,
0.384916
]
],
"network.4.bias": [
-0.145597,
-0.015258,
0.788115,
0.272858,
0.401335,
-0.374371
],
"network.6.weight": [
[
-0.52,
-0.354857,
0.705469,
0.559468,
0.480621,
-0.62921
],
[
-0.720518,
-0.147595,
0.595432,
0.547071,
0.48362,
-0.6471
],
[
0.48171,
0.400626,
-0.254656,
-0.476169,
0.153746,
0.083138
],
[
0.201516,
0.737626,
-0.458875,
-0.174434,
-0.574713,
-0.132551
],
[
-0.492453,
-0.054087,
0.389742,
0.732779,
0.698992,
0.179092
],
[
0.193158,
0.820914,
-0.449294,
-0.231688,
-0.664975,
0.024184
]
],
"network.6.bias": [
0.617031,
0.570815,
0.403173,
0.05844,
0.334374,
0.391823
],
"network.8.weight": [
[
0.64799,
0.656925,
-0.028934,
-0.296338,
0.315623,
-0.643567
],
[
-0.292315,
-0.23336,
0.185667,
0.368147,
-0.705969,
-0.11522
],
[
0.884893,
0.457297,
-0.213295,
-0.652001,
0.450712,
-0.227509
],
[
-0.502506,
-0.152446,
0.620327,
0.587259,
-0.665021,
0.473446
],
[
-0.729536,
-0.377126,
0.333751,
0.176156,
-0.398551,
0.333992
],
[
-0.463589,
-0.644534,
0.68104,
0.296006,
-0.548959,
0.350463
]
],
"network.8.bias": [
0.386602,
0.2863,
0.340354,
0.063636,
0.280192,
0.584021
],
"network.10.weight": [
[
-0.233883,
-0.482278,
-0.333919,
0.67735,
0.535556,
0.073066
],
[
-0.047806,
-0.64355,
-0.401765,
0.058635,
0.554525,
0.502979
],
[
0.605734,
-0.43273,
0.492164,
-0.128911,
-0.366687,
-0.22068
],
[
-0.454421,
0.343445,
-0.202428,
0.617927,
0.623635,
0.620928
],
[
-0.298212,
0.090822,
-0.541279,
0.300519,
0.684595,
0.650648
],
[
-0.15952,
0.192373,
-0.376841,
0.116675,
0.348651,
0.4461
]
],
"network.10.bias": [
0.168822,
0.029012,
0.479378,
0.381675,
0.501922,
0.240955
],
"network.12.weight": [
[
-0.484905,
-0.474798,
0.560042,
-0.617076,
-0.672929,
-0.580633
]
],
"network.12.bias": [
-0.03763
]
}
## Activation Signature
### 0
mean: [0.907898, 0.830229, 0.100692, 0.765726, 1.495150, -0.839327]
std: [1.620708, 1.561799, 1.294308, 1.505011, 1.992525, 0.840829]
### 2
mean: [2.390841, -2.453026, 3.073430, -0.333625, -0.279981, -1.181736]
std: [1.930530, 1.658062, 2.362043, 0.591053, 0.791181, 1.374141]
### 4
mean: [2.443803, 3.179261, -0.428703, -0.032431, -0.320981, -0.774430]
std: [2.313431, 2.670312, 1.282489, 0.531289, 0.891571, 0.327822]
### 6
mean: [-1.418918, -1.329442, 2.764119, 2.721927, -0.844673, 3.262806]
std: [2.642578, 2.496546, 2.311316, 2.775897, 1.761795, 3.009281]
### 8
mean: [-1.932788, 0.951376, -2.033884, 4.456558, 2.119523, 3.726937]
std: [3.706806, 1.779521, 3.954406, 5.062441, 3.136961, 4.338863]
### 10
mean: [4.103285, 2.658182, -1.666453, 7.580576, 5.914869, 3.394180]
std: [4.597656, 3.166248, 4.073865, 7.642177, 6.410896, 3.830112]
### 12
mean: [-14.522885]
std: [14.192790]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
decreasing_pairs
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.9078983664512634, "std": 1.6207081079483032}, "1": {"mean": 0.8302294611930847, "std": 1.5617989301681519}, "2": {"mean": 0.1006924957036972, "std": 1.2943083047866821}, "3": {"mean": 0.7657259702682495, "std": 1.5050106048583984}, "4": {"mean": 1.495150089263916, "std": 1.99252450466156}, "5": {"mean": -0.8393269777297974, "std": 0.8408288359642029}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 2.390840530395508, "std": 1.9305297136306763}, "1": {"mean": -2.4530258178710938, "std": 1.6580616235733032}, "2": {"mean": 3.073430299758911, "std": 2.3620433807373047}, "3": {"mean": -0.3336245119571686, "std": 0.5910526514053345}, "4": {"mean": -0.279981404542923, "std": 0.7911809086799622}, "5": {"mean": -1.1817364692687988, "std": 1.3741413354873657}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 2.44380259513855, "std": 2.3134312629699707}, "1": {"mean": 3.1792609691619873, "std": 2.67031192779541}, "2": {"mean": -0.42870303988456726, "std": 1.2824887037277222}, "3": {"mean": -0.032431285828351974, "std": 0.5312888622283936}, "4": {"mean": -0.3209807574748993, "std": 0.8915714025497437}, "5": {"mean": -0.774429976940155, "std": 0.3278222382068634}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -1.4189181327819824, "std": 2.642578125}, "1": {"mean": -1.3294415473937988, "std": 2.4965457916259766}, "2": {"mean": 2.7641191482543945, "std": 2.3113162517547607}, "3": {"mean": 2.7219271659851074, "std": 2.7758965492248535}, "4": {"mean": -0.8446725010871887, "std": 1.761794924736023}, "5": {"mean": 3.2628064155578613, "std": 3.0092811584472656}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.9327881336212158, "std": 3.70680570602417}, "1": {"mean": 0.9513764977455139, "std": 1.7795207500457764}, "2": {"mean": -2.033884286880493, "std": 3.9544055461883545}, "3": {"mean": 4.456557750701904, "std": 5.062440872192383}, "4": {"mean": 2.119523048400879, "std": 3.1369612216949463}, "5": {"mean": 3.7269365787506104, "std": 4.338863372802734}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": 4.10328483581543, "std": 4.59765625}, "1": {"mean": 2.65818190574646, "std": 3.166247606277466}, "2": {"mean": -1.6664533615112305, "std": 4.073864936828613}, "3": {"mean": 7.580575942993164, "std": 7.642177104949951}, "4": {"mean": 5.91486930847168, "std": 6.410895824432373}, "5": {"mean": 3.3941800594329834, "std": 3.830111503601074}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -14.5228853225708, "std": 14.192790031433105}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.598627, -0.168119, 0.269685, 0.461029, -0.231523], [-0.624035, 0.196546, -0.172766, 0.323372, 0.19696], [-0.154035, -0.006033, -0.454197, 0.379071, 0.104111], [-0.484359, 0.054547, 0.439236, -0.164646, 0.584382], [-0.713141, 0.254444, 0.155965, 0.591237, 0.210261], [-0.226956, -0.234816, -0.136635, 0.075597, 0.093668]], "network.0.bias": [0.692758, 0.673174, 0.31084, -0.016093, 0.065971, -0.121228], "network.2.weight": [[0.57438, 0.576218, 0.53029, 0.219544, 0.262341, -0.581332], [-0.65949, -0.634313, -0.031302, 0.130836, -0.218025, 0.166749], [0.380605, 0.349053, 0.220438, 0.689507, 0.73411, -0.12643], [-0.467985, -0.553698, -0.191315, -0.032273, 0.400655, 0.511745], [-0.519015, -0.107456, -0.602622, -0.203195, 0.175333, 0.626074], [-0.112379, -0.378033, -0.263262, 0.544732, -0.401194, -0.188167]], "network.2.bias": [0.171517, -0.74109, 0.266068, 0.282902, 0.657812, -0.372192], "network.4.weight": [[0.517248, 0.366811, 0.474461, -0.586477, -0.538393, 0.189903], [0.369881, -0.207982, 0.769733, -0.650688, -0.364273, 0.324629], [-0.124064, -0.340113, -0.329222, 0.634395, 0.708403, -0.586608], [-0.316782, -0.402439, 0.135556, 0.832068, 0.078744, -0.610858], [-0.55672, -0.348199, 0.166331, -0.015835, 0.630139, -0.577851], [0.204231, 0.022286, -0.288439, -0.310137, -0.05585, 0.384916]], "network.4.bias": [-0.145597, -0.015258, 0.788115, 0.272858, 0.401335, -0.374371], "network.6.weight": [[-0.52, -0.354857, 0.705469, 0.559468, 0.480621, -0.62921], [-0.720518, -0.147595, 0.595432, 0.547071, 0.48362, -0.6471], [0.48171, 0.400626, -0.254656, -0.476169, 0.153746, 0.083138], [0.201516, 0.737626, -0.458875, -0.174434, -0.574713, -0.132551], [-0.492453, -0.054087, 0.389742, 0.732779, 0.698992, 0.179092], [0.193158, 0.820914, -0.449294, -0.231688, -0.664975, 0.024184]], "network.6.bias": [0.617031, 0.570815, 0.403173, 0.05844, 0.334374, 0.391823], "network.8.weight": [[0.64799, 0.656925, -0.028934, -0.296338, 0.315623, -0.643567], [-0.292315, -0.23336, 0.185667, 0.368147, -0.705969, -0.11522], [0.884893, 0.457297, -0.213295, -0.652001, 0.450712, -0.227509], [-0.502506, -0.152446, 0.620327, 0.587259, -0.665021, 0.473446], [-0.729536, -0.377126, 0.333751, 0.176156, -0.398551, 0.333992], [-0.463589, -0.644534, 0.68104, 0.296006, -0.548959, 0.350463]], "network.8.bias": [0.386602, 0.2863, 0.340354, 0.063636, 0.280192, 0.584021], "network.10.weight": [[-0.233883, -0.482278, -0.333919, 0.67735, 0.535556, 0.073066], [-0.047806, -0.64355, -0.401765, 0.058635, 0.554525, 0.502979], [0.605734, -0.43273, 0.492164, -0.128911, -0.366687, -0.22068], [-0.454421, 0.343445, -0.202428, 0.617927, 0.623635, 0.620928], [-0.298212, 0.090822, -0.541279, 0.300519, 0.684595, 0.650648], [-0.15952, 0.192373, -0.376841, 0.116675, 0.348651, 0.4461]], "network.10.bias": [0.168822, 0.029012, 0.479378, 0.381675, 0.501922, 0.240955], "network.12.weight": [[-0.484905, -0.474798, 0.560042, -0.617076, -0.672929, -0.580633]], "network.12.bias": [-0.03763]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6811582744121552, "train_acc": 0.585, "val_loss": 0.7202267646789551, "val_acc": 0.44}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6768019795417786, "train_acc": 0.585, "val_loss": 0.7056785821914673, "val_acc": 0.44}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6672624349594116, "train_acc": 0.585, "val_loss": 0.6904460191726685, "val_acc": 0.44}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6446685791015625, "train_acc": 0.585, "val_loss": 0.6435732841491699, "val_acc": 0.44}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.5794844031333923, "train_acc": 0.515, "val_loss": 0.478657603263855, "val_acc": 0.94}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.42411115765571594, "train_acc": 0.915, "val_loss": 0.3543333113193512, "val_acc": 0.92}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.31074512004852295, "train_acc": 0.915, "val_loss": 0.2811303436756134, "val_acc": 0.92}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.2550388276576996, "train_acc": 0.93, "val_loss": 0.20099276304244995, "val_acc": 0.94}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.19756119698286057, "train_acc": 0.945, "val_loss": 0.13166336715221405, "val_acc": 0.96}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.1912015900015831, "train_acc": 0.95, "val_loss": 0.09049330651760101, "val_acc": 0.98}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.15945401042699814, "train_acc": 0.955, "val_loss": 0.09122191369533539, "val_acc": 0.98}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.26112569868564606, "train_acc": 0.95, "val_loss": 0.08831071853637695, "val_acc": 0.98}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.11579304607585073, "train_acc": 0.96, "val_loss": 0.08447137475013733, "val_acc": 0.98}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.1639416143298149, "train_acc": 0.955, "val_loss": 0.08267652243375778, "val_acc": 0.98}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["decreasing_pairs"], "degraded_stage": {"initial_val_loss": 0.7202267646789551, "final_val_loss": 0.6435732841491699, "initial_val_acc": 0.44, "final_val_acc": 0.44, "best_val_acc": 0.44}, "improved_stage": {"initial_val_loss": 0.478657603263855, "final_val_loss": 0.08267652243375778, "initial_val_acc": 0.94, "final_val_acc": 0.98, "best_val_acc": 0.98, "best_epoch": 9}, "improvement": 0.54, "first_improvement_epoch": 3}}
|
68
|
{"target_pattern": "ends_with", "degraded_accuracy": 0.46, "improved_accuracy": 0.88, "improvement": 0.42, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 6958, "learning_rate": 0.08815082734579661, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "ends_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["ends_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
1.130869,
0.008595,
-0.045251,
-0.321412,
0.790633
],
[
0.621127,
-0.189417,
-0.156909,
-0.463029,
0.96787
],
[
-0.256047,
-0.670645,
-0.157623,
-0.220294,
-0.213989
],
[
0.215934,
0.409858,
-0.165985,
0.601897,
-1.056108
],
[
-0.82963,
-0.188444,
-0.258617,
0.11324,
1.138488
],
[
0.021902,
-0.113747,
0.416592,
0.075892,
-0.040703
],
[
0.66618,
-0.276955,
-0.235735,
-0.43338,
0.879801
],
[
-0.480546,
0.160653,
-0.023016,
0.376309,
1.219302
]
],
"network.0.bias": [
0.047835,
-0.293407,
-0.562162,
0.295605,
-0.157422,
-0.087261,
-0.081568,
0.37456
],
"network.2.weight": [
[
-0.509268,
-0.457007,
0.024477,
0.477352,
-0.317344,
0.143013,
-0.080931,
-0.646011
],
[
0.805663,
0.418217,
-0.024189,
-0.390624,
0.412708,
0.232947,
0.660945,
1.041182
],
[
0.440006,
0.428386,
-0.229333,
0.014165,
-0.968488,
-0.250654,
0.487142,
0.051493
],
[
-0.341708,
-0.495928,
0.046603,
-0.291585,
-0.614049,
-0.635007,
-0.604605,
-0.079737
],
[
-0.522828,
-0.478523,
0.270772,
-0.294911,
-0.141503,
-0.179182,
0.035006,
-0.463899
],
[
-0.290837,
0.007798,
-0.062426,
-0.656736,
-0.515013,
0.296163,
-0.38526,
-0.602926
],
[
0.205325,
-0.046622,
0.612783,
-0.45288,
0.559081,
-0.128781,
-0.050504,
1.046938
],
[
-0.75411,
0.29265,
0.180781,
-0.511406,
0.250335,
-0.262818,
-0.482735,
-0.661624
]
],
"network.2.bias": [
0.93584,
0.171268,
-0.393705,
-0.328546,
-0.358992,
0.667034,
-0.227885,
-0.214742
],
"network.4.weight": [
[
-0.930121,
0.621117,
0.180193,
0.261239,
-0.018579,
-0.675249,
0.650122,
-0.426244
],
[
0.49205,
-0.188828,
-0.206465,
-0.329845,
0.31957,
-0.002987,
-0.207322,
0.287678
],
[
0.539848,
-0.283771,
-0.507946,
-0.489755,
-0.030784,
-0.059097,
-0.52795,
-0.074767
],
[
-0.443677,
0.316722,
-0.185478,
-0.168352,
-0.13007,
0.168883,
0.373129,
0.001102
],
[
-0.824881,
0.549305,
0.511041,
0.147826,
-0.071122,
-0.603288,
0.764596,
-0.566488
],
[
-0.288194,
-0.053132,
0.228758,
0.420218,
-0.059674,
-0.170599,
0.169506,
0.003986
],
[
-0.219393,
-0.084158,
-0.274589,
0.091211,
0.184392,
-0.263183,
-0.394603,
-0.214851
],
[
-0.557983,
0.388448,
0.153531,
-0.128743,
0.101005,
0.186908,
0.098641,
-0.293024
]
],
"network.4.bias": [
0.044925,
0.455766,
0.310688,
-0.584631,
0.148171,
-0.574621,
-0.479915,
-0.39109
],
"network.6.weight": [
[
0.14499,
0.00371,
-0.225416,
-0.344285,
-0.216547,
0.198,
-0.319047,
0.059571
],
[
0.516544,
-0.430603,
-0.719311,
0.360748,
0.521225,
0.123803,
-0.416785,
0.312161
],
[
-0.224004,
0.546419,
0.313725,
-0.182609,
-0.495137,
0.2565,
-0.085196,
0.035633
],
[
0.672599,
-0.410815,
0.050753,
0.536631,
0.183423,
-0.048659,
0.107128,
0.12242
],
[
-0.108305,
0.276272,
0.733867,
-0.566256,
-0.530261,
-0.248508,
-0.185214,
-0.739672
],
[
-0.188871,
-0.125579,
-0.090553,
-0.00355,
-0.297211,
0.251459,
-0.123678,
0.017742
],
[
0.516169,
-0.346135,
-0.431199,
0.532546,
0.333237,
-0.191182,
0.07058,
0.489607
],
[
0.129586,
-0.266865,
-0.068794,
-0.187657,
-0.120191,
0.02701,
0.092413,
-0.201012
]
],
"network.6.bias": [
-0.209949,
0.197044,
-0.09186,
0.128261,
0.529816,
-0.122503,
0.055172,
-0.352985
],
"network.8.weight": [
[
0.25083,
-0.068291,
-0.253078,
0.324947,
-0.165011,
-0.347798,
-0.108627,
-0.016467
],
[
-0.043722,
-0.085008,
-0.528217,
-0.320502,
-0.08776,
0.137274,
-0.646638,
-0.151908
],
[
0.324045,
0.574364,
0.251204,
0.582736,
-0.890863,
-0.23739,
0.243118,
0.213557
],
[
-0.255679,
0.115403,
-0.043592,
-0.019047,
-0.143446,
0.184625,
-0.362584,
-0.284161
],
[
-0.19317,
-0.943215,
0.300885,
-0.347557,
0.910941,
0.117797,
-0.488474,
-0.218245
],
[
-0.350622,
-0.107983,
0.074195,
0.074345,
-0.329501,
0.218417,
-0.307299,
0.062007
],
[
0.294483,
0.266514,
-0.46824,
-0.367332,
-0.251285,
-0.166525,
-0.053107,
-0.253588
],
[
0.312482,
-0.014019,
0.060671,
0.030815,
0.026742,
-0.043274,
-0.321949,
0.319799
]
],
"network.8.bias": [
-0.317913,
-0.736232,
0.211125,
-0.428568,
0.196981,
-0.035759,
-0.3038,
-0.351867
],
"network.10.weight": [
[
-0.323334,
0.257968,
-0.574574,
0.120543,
0.473922,
0.155662,
0.140329,
-0.066394
]
],
"network.10.bias": [
0.487666
]
}
## Activation Signature
### 0
mean: [1.662141, 0.007392, -3.168447, 0.952652, -0.426489, 0.715484, -0.093911, 2.331992]
std: [2.957995, 2.418061, 1.869929, 2.462304, 2.659692, 0.788509, 2.339046, 2.542563]
### 2
mean: [-1.511850, 5.109760, 0.587040, -3.545232, -3.633132, -2.844493, 2.216619, -4.267012]
std: [3.639780, 5.733577, 2.576861, 3.006835, 2.679237, 2.676011, 3.667738, 2.857402]
### 4
mean: [4.353376, -0.938936, -2.611514, 1.465013, 4.772486, -0.380013, -2.256696, 1.692595]
std: [6.271890, 2.201066, 4.155027, 3.051693, 6.574364, 0.675486, 1.991960, 2.956117]
### 6
mean: [-1.132608, 6.031157, -3.467623, 5.125063, -4.779941, -2.433872, 5.680100, -1.196126]
std: [1.220293, 8.661536, 5.118792, 7.212907, 7.935602, 2.921712, 8.226978, 0.960656]
### 8
mean: [0.170092, -6.874650, 7.849632, -2.014024, -9.837637, -2.276055, -1.081193, -2.140981]
std: [0.975524, 8.141447, 11.317319, 2.032393, 14.847471, 2.765924, 0.681717, 2.522832]
### 10
mean: [-4.052711]
std: [6.835497]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
ends_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
1.130869,
0.008595,
-0.045251,
-0.321412,
0.790633
],
[
0.621127,
-0.189417,
-0.156909,
-0.463029,
0.96787
],
[
-0.256047,
-0.670645,
-0.157623,
-0.220294,
-0.213989
],
[
0.215934,
0.409858,
-0.165985,
0.601897,
-1.056108
],
[
-0.82963,
-0.188444,
-0.258617,
0.11324,
1.138488
],
[
0.021902,
-0.113747,
0.416592,
0.075892,
-0.040703
],
[
0.66618,
-0.276955,
-0.235735,
-0.43338,
0.879801
],
[
-0.480546,
0.160653,
-0.023016,
0.376309,
1.219302
]
],
"network.0.bias": [
0.047835,
-0.293407,
-0.562162,
0.295605,
-0.157422,
-0.087261,
-0.081568,
0.37456
],
"network.2.weight": [
[
-0.509268,
-0.457007,
0.024477,
0.477352,
-0.317344,
0.143013,
-0.080931,
-0.646011
],
[
0.805663,
0.418217,
-0.024189,
-0.390624,
0.412708,
0.232947,
0.660945,
1.041182
],
[
0.440006,
0.428386,
-0.229333,
0.014165,
-0.968488,
-0.250654,
0.487142,
0.051493
],
[
-0.341708,
-0.495928,
0.046603,
-0.291585,
-0.614049,
-0.635007,
-0.604605,
-0.079737
],
[
-0.522828,
-0.478523,
0.270772,
-0.294911,
-0.141503,
-0.179182,
0.035006,
-0.463899
],
[
-0.290837,
0.007798,
-0.062426,
-0.656736,
-0.515013,
0.296163,
-0.38526,
-0.602926
],
[
0.205325,
-0.046622,
0.612783,
-0.45288,
0.559081,
-0.128781,
-0.050504,
1.046938
],
[
-0.75411,
0.29265,
0.180781,
-0.511406,
0.250335,
-0.262818,
-0.482735,
-0.661624
]
],
"network.2.bias": [
0.93584,
0.171268,
-0.393705,
-0.328546,
-0.358992,
0.667034,
-0.227885,
-0.214742
],
"network.4.weight": [
[
-0.930121,
0.621117,
0.180193,
0.261239,
-0.018579,
-0.675249,
0.650122,
-0.426244
],
[
0.49205,
-0.188828,
-0.206465,
-0.329845,
0.31957,
-0.002987,
-0.207322,
0.287678
],
[
0.539848,
-0.283771,
-0.507946,
-0.489755,
-0.030784,
-0.059097,
-0.52795,
-0.074767
],
[
-0.443677,
0.316722,
-0.185478,
-0.168352,
-0.13007,
0.168883,
0.373129,
0.001102
],
[
-0.824881,
0.549305,
0.511041,
0.147826,
-0.071122,
-0.603288,
0.764596,
-0.566488
],
[
-0.288194,
-0.053132,
0.228758,
0.420218,
-0.059674,
-0.170599,
0.169506,
0.003986
],
[
-0.219393,
-0.084158,
-0.274589,
0.091211,
0.184392,
-0.263183,
-0.394603,
-0.214851
],
[
-0.557983,
0.388448,
0.153531,
-0.128743,
0.101005,
0.186908,
0.098641,
-0.293024
]
],
"network.4.bias": [
0.044925,
0.455766,
0.310688,
-0.584631,
0.148171,
-0.574621,
-0.479915,
-0.39109
],
"network.6.weight": [
[
0.14499,
0.00371,
-0.225416,
-0.344285,
-0.216547,
0.198,
-0.319047,
0.059571
],
[
0.516544,
-0.430603,
-0.719311,
0.360748,
0.521225,
0.123803,
-0.416785,
0.312161
],
[
-0.224004,
0.546419,
0.313725,
-0.182609,
-0.495137,
0.2565,
-0.085196,
0.035633
],
[
0.672599,
-0.410815,
0.050753,
0.536631,
0.183423,
-0.048659,
0.107128,
0.12242
],
[
-0.108305,
0.276272,
0.733867,
-0.566256,
-0.530261,
-0.248508,
-0.185214,
-0.739672
],
[
-0.188871,
-0.125579,
-0.090553,
-0.00355,
-0.297211,
0.251459,
-0.123678,
0.017742
],
[
0.516169,
-0.346135,
-0.431199,
0.532546,
0.333237,
-0.191182,
0.07058,
0.489607
],
[
0.129586,
-0.266865,
-0.068794,
-0.187657,
-0.120191,
0.02701,
0.092413,
-0.201012
]
],
"network.6.bias": [
-0.209949,
0.197044,
-0.09186,
0.128261,
0.529816,
-0.122503,
0.055172,
-0.352985
],
"network.8.weight": [
[
0.25083,
-0.068291,
-0.253078,
0.324947,
-0.165011,
-0.347798,
-0.108627,
-0.016467
],
[
-0.043722,
-0.085008,
-0.528217,
-0.320502,
-0.08776,
0.137274,
-0.646638,
-0.151908
],
[
0.324045,
0.574364,
0.251204,
0.582736,
-0.890863,
-0.23739,
0.243118,
0.213557
],
[
-0.255679,
0.115403,
-0.043592,
-0.019047,
-0.143446,
0.184625,
-0.362584,
-0.284161
],
[
-0.19317,
-0.943215,
0.300885,
-0.347557,
0.910941,
0.117797,
-0.488474,
-0.218245
],
[
-0.350622,
-0.107983,
0.074195,
0.074345,
-0.329501,
0.218417,
-0.307299,
0.062007
],
[
0.294483,
0.266514,
-0.46824,
-0.367332,
-0.251285,
-0.166525,
-0.053107,
-0.253588
],
[
0.312482,
-0.014019,
0.060671,
0.030815,
0.026742,
-0.043274,
-0.321949,
0.319799
]
],
"network.8.bias": [
-0.317913,
-0.736232,
0.211125,
-0.428568,
0.196981,
-0.035759,
-0.3038,
-0.351867
],
"network.10.weight": [
[
-0.323334,
0.257968,
-0.574574,
0.120543,
0.473922,
0.155662,
0.140329,
-0.066394
]
],
"network.10.bias": [
0.487666
]
}
## Activation Signature
### 0
mean: [1.662141, 0.007392, -3.168447, 0.952652, -0.426489, 0.715484, -0.093911, 2.331992]
std: [2.957995, 2.418061, 1.869929, 2.462304, 2.659692, 0.788509, 2.339046, 2.542563]
### 2
mean: [-1.511850, 5.109760, 0.587040, -3.545232, -3.633132, -2.844493, 2.216619, -4.267012]
std: [3.639780, 5.733577, 2.576861, 3.006835, 2.679237, 2.676011, 3.667738, 2.857402]
### 4
mean: [4.353376, -0.938936, -2.611514, 1.465013, 4.772486, -0.380013, -2.256696, 1.692595]
std: [6.271890, 2.201066, 4.155027, 3.051693, 6.574364, 0.675486, 1.991960, 2.956117]
### 6
mean: [-1.132608, 6.031157, -3.467623, 5.125063, -4.779941, -2.433872, 5.680100, -1.196126]
std: [1.220293, 8.661536, 5.118792, 7.212907, 7.935602, 2.921712, 8.226978, 0.960656]
### 8
mean: [0.170092, -6.874650, 7.849632, -2.014024, -9.837637, -2.276055, -1.081193, -2.140981]
std: [0.975524, 8.141447, 11.317319, 2.032393, 14.847471, 2.765924, 0.681717, 2.522832]
### 10
mean: [-4.052711]
std: [6.835497]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
ends_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.6621410846710205, "std": 2.9579949378967285}, "1": {"mean": 0.007392088416963816, "std": 2.4180612564086914}, "2": {"mean": -3.1684467792510986, "std": 1.8699289560317993}, "3": {"mean": 0.9526515603065491, "std": 2.462303638458252}, "4": {"mean": -0.426488995552063, "std": 2.6596920490264893}, "5": {"mean": 0.7154842615127563, "std": 0.7885090708732605}, "6": {"mean": -0.09391114860773087, "std": 2.3390464782714844}, "7": {"mean": 2.3319921493530273, "std": 2.542562961578369}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -1.5118497610092163, "std": 3.639780044555664}, "1": {"mean": 5.109760284423828, "std": 5.733576774597168}, "2": {"mean": 0.5870401859283447, "std": 2.5768609046936035}, "3": {"mean": -3.545232057571411, "std": 3.0068352222442627}, "4": {"mean": -3.633131980895996, "std": 2.6792373657226562}, "5": {"mean": -2.8444933891296387, "std": 2.6760106086730957}, "6": {"mean": 2.216618776321411, "std": 3.6677377223968506}, "7": {"mean": -4.267011642456055, "std": 2.8574020862579346}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 4.353376388549805, "std": 6.271890163421631}, "1": {"mean": -0.938936173915863, "std": 2.201066255569458}, "2": {"mean": -2.611513614654541, "std": 4.155027389526367}, "3": {"mean": 1.465012788772583, "std": 3.0516932010650635}, "4": {"mean": 4.772486209869385, "std": 6.574364185333252}, "5": {"mean": -0.3800134062767029, "std": 0.6754857301712036}, "6": {"mean": -2.2566959857940674, "std": 1.9919596910476685}, "7": {"mean": 1.6925952434539795, "std": 2.9561166763305664}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -1.1326082944869995, "std": 1.220293402671814}, "1": {"mean": 6.03115701675415, "std": 8.66153621673584}, "2": {"mean": -3.467623233795166, "std": 5.118791580200195}, "3": {"mean": 5.125062942504883, "std": 7.212907314300537}, "4": {"mean": -4.779941082000732, "std": 7.935602188110352}, "5": {"mean": -2.4338724613189697, "std": 2.9217123985290527}, "6": {"mean": 5.680100440979004, "std": 8.226978302001953}, "7": {"mean": -1.1961263418197632, "std": 0.960656464099884}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.1700921356678009, "std": 0.9755235314369202}, "1": {"mean": -6.874650001525879, "std": 8.141447067260742}, "2": {"mean": 7.849632263183594, "std": 11.3173189163208}, "3": {"mean": -2.014023780822754, "std": 2.0323925018310547}, "4": {"mean": -9.837636947631836, "std": 14.847471237182617}, "5": {"mean": -2.276055097579956, "std": 2.765923500061035}, "6": {"mean": -1.0811933279037476, "std": 0.6817171573638916}, "7": {"mean": -2.1409809589385986, "std": 2.522831916809082}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -4.052711009979248, "std": 6.83549690246582}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[1.130869, 0.008595, -0.045251, -0.321412, 0.790633], [0.621127, -0.189417, -0.156909, -0.463029, 0.96787], [-0.256047, -0.670645, -0.157623, -0.220294, -0.213989], [0.215934, 0.409858, -0.165985, 0.601897, -1.056108], [-0.82963, -0.188444, -0.258617, 0.11324, 1.138488], [0.021902, -0.113747, 0.416592, 0.075892, -0.040703], [0.66618, -0.276955, -0.235735, -0.43338, 0.879801], [-0.480546, 0.160653, -0.023016, 0.376309, 1.219302]], "network.0.bias": [0.047835, -0.293407, -0.562162, 0.295605, -0.157422, -0.087261, -0.081568, 0.37456], "network.2.weight": [[-0.509268, -0.457007, 0.024477, 0.477352, -0.317344, 0.143013, -0.080931, -0.646011], [0.805663, 0.418217, -0.024189, -0.390624, 0.412708, 0.232947, 0.660945, 1.041182], [0.440006, 0.428386, -0.229333, 0.014165, -0.968488, -0.250654, 0.487142, 0.051493], [-0.341708, -0.495928, 0.046603, -0.291585, -0.614049, -0.635007, -0.604605, -0.079737], [-0.522828, -0.478523, 0.270772, -0.294911, -0.141503, -0.179182, 0.035006, -0.463899], [-0.290837, 0.007798, -0.062426, -0.656736, -0.515013, 0.296163, -0.38526, -0.602926], [0.205325, -0.046622, 0.612783, -0.45288, 0.559081, -0.128781, -0.050504, 1.046938], [-0.75411, 0.29265, 0.180781, -0.511406, 0.250335, -0.262818, -0.482735, -0.661624]], "network.2.bias": [0.93584, 0.171268, -0.393705, -0.328546, -0.358992, 0.667034, -0.227885, -0.214742], "network.4.weight": [[-0.930121, 0.621117, 0.180193, 0.261239, -0.018579, -0.675249, 0.650122, -0.426244], [0.49205, -0.188828, -0.206465, -0.329845, 0.31957, -0.002987, -0.207322, 0.287678], [0.539848, -0.283771, -0.507946, -0.489755, -0.030784, -0.059097, -0.52795, -0.074767], [-0.443677, 0.316722, -0.185478, -0.168352, -0.13007, 0.168883, 0.373129, 0.001102], [-0.824881, 0.549305, 0.511041, 0.147826, -0.071122, -0.603288, 0.764596, -0.566488], [-0.288194, -0.053132, 0.228758, 0.420218, -0.059674, -0.170599, 0.169506, 0.003986], [-0.219393, -0.084158, -0.274589, 0.091211, 0.184392, -0.263183, -0.394603, -0.214851], [-0.557983, 0.388448, 0.153531, -0.128743, 0.101005, 0.186908, 0.098641, -0.293024]], "network.4.bias": [0.044925, 0.455766, 0.310688, -0.584631, 0.148171, -0.574621, -0.479915, -0.39109], "network.6.weight": [[0.14499, 0.00371, -0.225416, -0.344285, -0.216547, 0.198, -0.319047, 0.059571], [0.516544, -0.430603, -0.719311, 0.360748, 0.521225, 0.123803, -0.416785, 0.312161], [-0.224004, 0.546419, 0.313725, -0.182609, -0.495137, 0.2565, -0.085196, 0.035633], [0.672599, -0.410815, 0.050753, 0.536631, 0.183423, -0.048659, 0.107128, 0.12242], [-0.108305, 0.276272, 0.733867, -0.566256, -0.530261, -0.248508, -0.185214, -0.739672], [-0.188871, -0.125579, -0.090553, -0.00355, -0.297211, 0.251459, -0.123678, 0.017742], [0.516169, -0.346135, -0.431199, 0.532546, 0.333237, -0.191182, 0.07058, 0.489607], [0.129586, -0.266865, -0.068794, -0.187657, -0.120191, 0.02701, 0.092413, -0.201012]], "network.6.bias": [-0.209949, 0.197044, -0.09186, 0.128261, 0.529816, -0.122503, 0.055172, -0.352985], "network.8.weight": [[0.25083, -0.068291, -0.253078, 0.324947, -0.165011, -0.347798, -0.108627, -0.016467], [-0.043722, -0.085008, -0.528217, -0.320502, -0.08776, 0.137274, -0.646638, -0.151908], [0.324045, 0.574364, 0.251204, 0.582736, -0.890863, -0.23739, 0.243118, 0.213557], [-0.255679, 0.115403, -0.043592, -0.019047, -0.143446, 0.184625, -0.362584, -0.284161], [-0.19317, -0.943215, 0.300885, -0.347557, 0.910941, 0.117797, -0.488474, -0.218245], [-0.350622, -0.107983, 0.074195, 0.074345, -0.329501, 0.218417, -0.307299, 0.062007], [0.294483, 0.266514, -0.46824, -0.367332, -0.251285, -0.166525, -0.053107, -0.253588], [0.312482, -0.014019, 0.060671, 0.030815, 0.026742, -0.043274, -0.321949, 0.319799]], "network.8.bias": [-0.317913, -0.736232, 0.211125, -0.428568, 0.196981, -0.035759, -0.3038, -0.351867], "network.10.weight": [[-0.323334, 0.257968, -0.574574, 0.120543, 0.473922, 0.155662, 0.140329, -0.066394]], "network.10.bias": [0.487666]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7229806780815125, "train_acc": 0.405, "val_loss": 0.6952873468399048, "val_acc": 0.46}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6766420602798462, "train_acc": 0.595, "val_loss": 0.7286233305931091, "val_acc": 0.46}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.5937195718288422, "train_acc": 0.595, "val_loss": 0.5643711090087891, "val_acc": 0.46}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.5735491216182709, "train_acc": 0.55, "val_loss": 0.5106047987937927, "val_acc": 0.76}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.510858952999115, "train_acc": 0.735, "val_loss": 0.4983022212982178, "val_acc": 0.76}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.4952864497900009, "train_acc": 0.735, "val_loss": 0.4637754559516907, "val_acc": 0.78}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.40155285596847534, "train_acc": 0.81, "val_loss": 0.3971730172634125, "val_acc": 0.84}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.3076358735561371, "train_acc": 0.885, "val_loss": 0.3060067296028137, "val_acc": 0.88}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.23174984753131866, "train_acc": 0.905, "val_loss": 0.46107766032218933, "val_acc": 0.74}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.38277725875377655, "train_acc": 0.845, "val_loss": 0.5743674039840698, "val_acc": 0.84}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.4604157358407974, "train_acc": 0.88, "val_loss": 0.46287956833839417, "val_acc": 0.84}], "summary": {"total_epochs": 11, "degraded_epochs": 3, "improved_epochs": 8, "patterns": ["ends_with"], "degraded_stage": {"initial_val_loss": 0.6952873468399048, "final_val_loss": 0.5643711090087891, "initial_val_acc": 0.46, "final_val_acc": 0.46, "best_val_acc": 0.46}, "improved_stage": {"initial_val_loss": 0.5106047987937927, "final_val_loss": 0.46287956833839417, "initial_val_acc": 0.76, "final_val_acc": 0.84, "best_val_acc": 0.88, "best_epoch": 7}, "improvement": 0.42, "first_improvement_epoch": 2}}
|
69
|
{"target_pattern": "decreasing_pairs", "degraded_accuracy": 0.62, "improved_accuracy": 0.98, "improvement": 0.36, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 7617, "learning_rate": 0.01839529955470709, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "decreasing_pairs", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["decreasing_pairs"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.101556,
-0.161518,
0.153132,
-0.149328,
0.307934
],
[
-0.292583,
0.388131,
-0.136571,
-0.210083,
-0.294516
],
[
-0.351882,
0.084912,
0.479277,
-0.067484,
-0.100113
],
[
0.507938,
0.313129,
0.202101,
-0.291326,
-0.037162
],
[
-0.307334,
-0.218279,
-0.09585,
-0.124478,
0.234885
],
[
0.237028,
-0.158461,
0.451675,
0.481993,
0.549941
],
[
-0.069829,
0.24676,
-0.239249,
0.438168,
-0.110908
]
],
"network.0.bias": [
0.452033,
0.561245,
0.29963,
-0.048806,
0.358963,
0.305415,
0.267526
],
"network.2.weight": [
[
-0.015723,
0.173583,
0.087373,
-0.118683,
-0.08279,
-0.036536,
-0.208821
],
[
0.030767,
0.159471,
0.283741,
-0.491486,
0.192642,
0.458297,
0.350974
],
[
-0.124836,
-0.253585,
-0.045332,
0.074252,
0.380424,
0.202316,
-0.111971
],
[
0.388004,
0.111447,
0.218818,
-0.340654,
-0.072816,
-0.239934,
-0.014803
],
[
-0.581146,
-0.05264,
0.319983,
-0.135679,
-0.293342,
-0.436958,
0.296537
],
[
-0.558868,
-0.254622,
-0.340538,
0.400504,
0.165453,
0.038422,
0.035918
],
[
-0.054634,
-0.127612,
0.013919,
0.394427,
0.086878,
-0.115889,
-0.122899
]
],
"network.2.bias": [
-0.094535,
0.362044,
-0.390503,
-0.037301,
-0.127581,
0.181996,
0.185857
],
"network.4.weight": [
[
0.107421,
-0.030958,
0.35389,
0.058009,
0.06546,
-0.308401,
-0.182185
],
[
0.034969,
0.301186,
0.076099,
-0.432218,
0.017777,
-0.014173,
-0.460212
],
[
-0.068372,
-0.195173,
-0.34394,
0.553566,
0.320341,
-0.40834,
0.191467
],
[
0.080235,
-0.227584,
0.194756,
0.158521,
-0.039729,
0.51782,
0.427074
],
[
0.156485,
0.100467,
-0.00891,
0.121087,
0.184021,
0.085037,
-0.169999
],
[
-0.24382,
0.334029,
-0.04966,
-0.308911,
-0.026027,
-0.423305,
-0.363348
],
[
-0.422606,
-0.188874,
0.180272,
-0.158127,
0.168404,
0.521194,
-0.053758
]
],
"network.4.bias": [
0.495019,
-0.122641,
-0.099683,
0.27665,
-0.332098,
0.01975,
0.081703
],
"network.6.weight": [
[
-0.491317,
-0.109689,
0.212636,
0.234606,
0.034183,
-0.470752,
-0.016705
],
[
-0.222868,
0.278816,
-0.069457,
0.388781,
0.155396,
-0.33794,
-0.263874
],
[
0.311173,
0.106101,
-0.397922,
-0.478728,
0.04663,
0.198337,
-0.256383
],
[
0.09876,
-0.387112,
-0.238054,
0.049592,
-0.411441,
-0.409779,
0.371756
],
[
0.360581,
-0.127743,
0.132731,
-0.16949,
0.16049,
0.325114,
-0.199129
],
[
0.369531,
0.35384,
-0.382273,
-0.423625,
-0.123077,
0.458099,
0.151397
],
[
-0.159262,
0.357932,
-0.080052,
-0.182123,
0.147139,
0.514723,
-0.171692
]
],
"network.6.bias": [
0.150254,
0.147178,
0.110088,
-0.187977,
0.416406,
0.088358,
0.496323
],
"network.8.weight": [
[
0.147247,
-0.197612,
0.470659,
-0.143807,
0.522217,
0.229661,
0.063608
],
[
0.242909,
-0.178817,
0.19516,
-0.285272,
0.346293,
0.258141,
0.083183
],
[
0.373456,
-0.184102,
-0.409404,
0.403549,
-0.234311,
-0.085581,
-0.25229
],
[
-0.017065,
-0.000762,
0.016622,
-0.520128,
-0.001037,
-0.241637,
-0.174107
],
[
-0.334168,
-0.166102,
0.488501,
-0.551021,
0.377972,
0.45731,
0.481101
],
[
0.073233,
-0.253715,
0.021116,
-0.526421,
0.109853,
0.413504,
0.279
],
[
-0.030766,
-0.262061,
0.273119,
-0.24176,
0.288915,
0.302578,
0.430383
]
],
"network.8.bias": [
-0.107985,
-0.178579,
0.235906,
0.279974,
0.22508,
0.279491,
0.320933
],
"network.10.weight": [
[
-0.270427,
-0.28806,
0.362148,
-0.317582,
-0.251381,
0.125285,
-0.171404
],
[
-0.219004,
0.378921,
-0.153942,
-0.102965,
0.442164,
-0.203033,
0.214379
],
[
0.213921,
-0.192241,
-0.351216,
-0.411897,
0.286589,
-0.199471,
0.107425
],
[
0.37709,
-0.076445,
-0.279817,
-0.161316,
0.365364,
0.547785,
0.438271
],
[
-0.282994,
0.048191,
-0.443896,
0.140703,
0.188858,
0.226648,
0.344082
],
[
-0.402463,
0.111827,
0.439856,
0.399542,
-0.478426,
-0.118008,
0.242869
],
[
0.286013,
0.316126,
-0.224495,
0.293604,
0.084103,
0.431125,
0.048607
]
],
"network.10.bias": [
0.123182,
0.055609,
-0.252605,
0.177422,
0.023725,
-0.304248,
-0.081112
],
"network.12.weight": [
[
0.254717,
-0.403498,
-0.373062,
-0.430695,
-0.306028,
0.183039,
-0.221673
]
],
"network.12.bias": [
-0.04509
]
}
## Activation Signature
### 0
mean: [0.410767, -0.194627, 0.749430, 0.905487, -0.598714, 2.966999, 0.933354]
std: [0.795711, 1.085666, 0.989796, 1.573714, 1.001905, 1.957567, 1.225773]
### 2
mean: [-0.436069, 1.814465, 0.038187, -0.745453, -1.286753, 0.192241, 0.082829]
std: [0.306838, 1.177632, 0.493691, 0.809724, 1.193054, 0.793276, 0.589952]
### 4
mean: [0.340439, 0.369476, -0.628227, 0.089063, -0.203312, 0.483761, -0.047567]
std: [0.215860, 0.527803, 0.356982, 0.696534, 0.139722, 0.715490, 0.474740]
### 6
mean: [-0.217124, 0.078504, 0.260652, -0.395896, 0.542392, 0.503849, 0.782011]
std: [0.406461, 0.201320, 0.493924, 0.449054, 0.294089, 0.523450, 0.479271]
### 8
mean: [0.357044, 0.180712, -0.214399, 0.102673, 1.054547, 0.713722, 0.924104]
std: [0.353142, 0.271314, 0.377294, 0.129466, 0.723713, 0.394302, 0.497597]
### 10
mean: [-0.330426, 0.530173, 0.017142, 1.295024, 0.569617, -0.730854, 0.435539]
std: [0.362962, 0.367759, 0.271810, 0.795074, 0.372672, 0.420113, 0.380755]
### 12
mean: [-0.999045]
std: [0.724825]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
decreasing_pairs
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.101556,
-0.161518,
0.153132,
-0.149328,
0.307934
],
[
-0.292583,
0.388131,
-0.136571,
-0.210083,
-0.294516
],
[
-0.351882,
0.084912,
0.479277,
-0.067484,
-0.100113
],
[
0.507938,
0.313129,
0.202101,
-0.291326,
-0.037162
],
[
-0.307334,
-0.218279,
-0.09585,
-0.124478,
0.234885
],
[
0.237028,
-0.158461,
0.451675,
0.481993,
0.549941
],
[
-0.069829,
0.24676,
-0.239249,
0.438168,
-0.110908
]
],
"network.0.bias": [
0.452033,
0.561245,
0.29963,
-0.048806,
0.358963,
0.305415,
0.267526
],
"network.2.weight": [
[
-0.015723,
0.173583,
0.087373,
-0.118683,
-0.08279,
-0.036536,
-0.208821
],
[
0.030767,
0.159471,
0.283741,
-0.491486,
0.192642,
0.458297,
0.350974
],
[
-0.124836,
-0.253585,
-0.045332,
0.074252,
0.380424,
0.202316,
-0.111971
],
[
0.388004,
0.111447,
0.218818,
-0.340654,
-0.072816,
-0.239934,
-0.014803
],
[
-0.581146,
-0.05264,
0.319983,
-0.135679,
-0.293342,
-0.436958,
0.296537
],
[
-0.558868,
-0.254622,
-0.340538,
0.400504,
0.165453,
0.038422,
0.035918
],
[
-0.054634,
-0.127612,
0.013919,
0.394427,
0.086878,
-0.115889,
-0.122899
]
],
"network.2.bias": [
-0.094535,
0.362044,
-0.390503,
-0.037301,
-0.127581,
0.181996,
0.185857
],
"network.4.weight": [
[
0.107421,
-0.030958,
0.35389,
0.058009,
0.06546,
-0.308401,
-0.182185
],
[
0.034969,
0.301186,
0.076099,
-0.432218,
0.017777,
-0.014173,
-0.460212
],
[
-0.068372,
-0.195173,
-0.34394,
0.553566,
0.320341,
-0.40834,
0.191467
],
[
0.080235,
-0.227584,
0.194756,
0.158521,
-0.039729,
0.51782,
0.427074
],
[
0.156485,
0.100467,
-0.00891,
0.121087,
0.184021,
0.085037,
-0.169999
],
[
-0.24382,
0.334029,
-0.04966,
-0.308911,
-0.026027,
-0.423305,
-0.363348
],
[
-0.422606,
-0.188874,
0.180272,
-0.158127,
0.168404,
0.521194,
-0.053758
]
],
"network.4.bias": [
0.495019,
-0.122641,
-0.099683,
0.27665,
-0.332098,
0.01975,
0.081703
],
"network.6.weight": [
[
-0.491317,
-0.109689,
0.212636,
0.234606,
0.034183,
-0.470752,
-0.016705
],
[
-0.222868,
0.278816,
-0.069457,
0.388781,
0.155396,
-0.33794,
-0.263874
],
[
0.311173,
0.106101,
-0.397922,
-0.478728,
0.04663,
0.198337,
-0.256383
],
[
0.09876,
-0.387112,
-0.238054,
0.049592,
-0.411441,
-0.409779,
0.371756
],
[
0.360581,
-0.127743,
0.132731,
-0.16949,
0.16049,
0.325114,
-0.199129
],
[
0.369531,
0.35384,
-0.382273,
-0.423625,
-0.123077,
0.458099,
0.151397
],
[
-0.159262,
0.357932,
-0.080052,
-0.182123,
0.147139,
0.514723,
-0.171692
]
],
"network.6.bias": [
0.150254,
0.147178,
0.110088,
-0.187977,
0.416406,
0.088358,
0.496323
],
"network.8.weight": [
[
0.147247,
-0.197612,
0.470659,
-0.143807,
0.522217,
0.229661,
0.063608
],
[
0.242909,
-0.178817,
0.19516,
-0.285272,
0.346293,
0.258141,
0.083183
],
[
0.373456,
-0.184102,
-0.409404,
0.403549,
-0.234311,
-0.085581,
-0.25229
],
[
-0.017065,
-0.000762,
0.016622,
-0.520128,
-0.001037,
-0.241637,
-0.174107
],
[
-0.334168,
-0.166102,
0.488501,
-0.551021,
0.377972,
0.45731,
0.481101
],
[
0.073233,
-0.253715,
0.021116,
-0.526421,
0.109853,
0.413504,
0.279
],
[
-0.030766,
-0.262061,
0.273119,
-0.24176,
0.288915,
0.302578,
0.430383
]
],
"network.8.bias": [
-0.107985,
-0.178579,
0.235906,
0.279974,
0.22508,
0.279491,
0.320933
],
"network.10.weight": [
[
-0.270427,
-0.28806,
0.362148,
-0.317582,
-0.251381,
0.125285,
-0.171404
],
[
-0.219004,
0.378921,
-0.153942,
-0.102965,
0.442164,
-0.203033,
0.214379
],
[
0.213921,
-0.192241,
-0.351216,
-0.411897,
0.286589,
-0.199471,
0.107425
],
[
0.37709,
-0.076445,
-0.279817,
-0.161316,
0.365364,
0.547785,
0.438271
],
[
-0.282994,
0.048191,
-0.443896,
0.140703,
0.188858,
0.226648,
0.344082
],
[
-0.402463,
0.111827,
0.439856,
0.399542,
-0.478426,
-0.118008,
0.242869
],
[
0.286013,
0.316126,
-0.224495,
0.293604,
0.084103,
0.431125,
0.048607
]
],
"network.10.bias": [
0.123182,
0.055609,
-0.252605,
0.177422,
0.023725,
-0.304248,
-0.081112
],
"network.12.weight": [
[
0.254717,
-0.403498,
-0.373062,
-0.430695,
-0.306028,
0.183039,
-0.221673
]
],
"network.12.bias": [
-0.04509
]
}
## Activation Signature
### 0
mean: [0.410767, -0.194627, 0.749430, 0.905487, -0.598714, 2.966999, 0.933354]
std: [0.795711, 1.085666, 0.989796, 1.573714, 1.001905, 1.957567, 1.225773]
### 2
mean: [-0.436069, 1.814465, 0.038187, -0.745453, -1.286753, 0.192241, 0.082829]
std: [0.306838, 1.177632, 0.493691, 0.809724, 1.193054, 0.793276, 0.589952]
### 4
mean: [0.340439, 0.369476, -0.628227, 0.089063, -0.203312, 0.483761, -0.047567]
std: [0.215860, 0.527803, 0.356982, 0.696534, 0.139722, 0.715490, 0.474740]
### 6
mean: [-0.217124, 0.078504, 0.260652, -0.395896, 0.542392, 0.503849, 0.782011]
std: [0.406461, 0.201320, 0.493924, 0.449054, 0.294089, 0.523450, 0.479271]
### 8
mean: [0.357044, 0.180712, -0.214399, 0.102673, 1.054547, 0.713722, 0.924104]
std: [0.353142, 0.271314, 0.377294, 0.129466, 0.723713, 0.394302, 0.497597]
### 10
mean: [-0.330426, 0.530173, 0.017142, 1.295024, 0.569617, -0.730854, 0.435539]
std: [0.362962, 0.367759, 0.271810, 0.795074, 0.372672, 0.420113, 0.380755]
### 12
mean: [-0.999045]
std: [0.724825]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
decreasing_pairs
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.4107673168182373, "std": 0.795711100101471}, "1": {"mean": -0.19462719559669495, "std": 1.0856660604476929}, "2": {"mean": 0.7494298219680786, "std": 0.9897957444190979}, "3": {"mean": 0.9054874777793884, "std": 1.5737144947052002}, "4": {"mean": -0.598714292049408, "std": 1.0019047260284424}, "5": {"mean": 2.9669992923736572, "std": 1.957566738128662}, "6": {"mean": 0.9333544969558716, "std": 1.2257729768753052}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.43606916069984436, "std": 0.3068382143974304}, "1": {"mean": 1.8144652843475342, "std": 1.1776323318481445}, "2": {"mean": 0.038186535239219666, "std": 0.4936912953853607}, "3": {"mean": -0.7454530596733093, "std": 0.8097237944602966}, "4": {"mean": -1.2867532968521118, "std": 1.1930538415908813}, "5": {"mean": 0.19224077463150024, "std": 0.7932759523391724}, "6": {"mean": 0.08282896131277084, "std": 0.5899524688720703}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.340438574552536, "std": 0.2158603072166443}, "1": {"mean": 0.3694758415222168, "std": 0.5278029441833496}, "2": {"mean": -0.628227174282074, "std": 0.3569820523262024}, "3": {"mean": 0.08906316012144089, "std": 0.6965338587760925}, "4": {"mean": -0.2033117115497589, "std": 0.13972210884094238}, "5": {"mean": 0.4837608337402344, "std": 0.7154898047447205}, "6": {"mean": -0.04756687581539154, "std": 0.47474023699760437}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.21712402999401093, "std": 0.406460702419281}, "1": {"mean": 0.07850430905818939, "std": 0.20131996273994446}, "2": {"mean": 0.2606521248817444, "std": 0.49392443895339966}, "3": {"mean": -0.3958961069583893, "std": 0.44905370473861694}, "4": {"mean": 0.5423919558525085, "std": 0.2940894663333893}, "5": {"mean": 0.5038493871688843, "std": 0.5234495401382446}, "6": {"mean": 0.782011091709137, "std": 0.4792709946632385}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.35704413056373596, "std": 0.35314157605171204}, "1": {"mean": 0.18071205914020538, "std": 0.27131387591362}, "2": {"mean": -0.21439868211746216, "std": 0.37729373574256897}, "3": {"mean": 0.10267269611358643, "std": 0.1294659674167633}, "4": {"mean": 1.05454683303833, "std": 0.7237125039100647}, "5": {"mean": 0.7137218713760376, "std": 0.39430174231529236}, "6": {"mean": 0.9241042137145996, "std": 0.4975971579551697}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -0.33042648434638977, "std": 0.36296218633651733}, "1": {"mean": 0.5301728844642639, "std": 0.3677586615085602}, "2": {"mean": 0.01714198850095272, "std": 0.2718101143836975}, "3": {"mean": 1.295023798942566, "std": 0.7950738668441772}, "4": {"mean": 0.5696167349815369, "std": 0.3726721405982971}, "5": {"mean": -0.7308538556098938, "std": 0.42011260986328125}, "6": {"mean": 0.43553903698921204, "std": 0.38075509667396545}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -0.9990453720092773, "std": 0.7248246669769287}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.101556, -0.161518, 0.153132, -0.149328, 0.307934], [-0.292583, 0.388131, -0.136571, -0.210083, -0.294516], [-0.351882, 0.084912, 0.479277, -0.067484, -0.100113], [0.507938, 0.313129, 0.202101, -0.291326, -0.037162], [-0.307334, -0.218279, -0.09585, -0.124478, 0.234885], [0.237028, -0.158461, 0.451675, 0.481993, 0.549941], [-0.069829, 0.24676, -0.239249, 0.438168, -0.110908]], "network.0.bias": [0.452033, 0.561245, 0.29963, -0.048806, 0.358963, 0.305415, 0.267526], "network.2.weight": [[-0.015723, 0.173583, 0.087373, -0.118683, -0.08279, -0.036536, -0.208821], [0.030767, 0.159471, 0.283741, -0.491486, 0.192642, 0.458297, 0.350974], [-0.124836, -0.253585, -0.045332, 0.074252, 0.380424, 0.202316, -0.111971], [0.388004, 0.111447, 0.218818, -0.340654, -0.072816, -0.239934, -0.014803], [-0.581146, -0.05264, 0.319983, -0.135679, -0.293342, -0.436958, 0.296537], [-0.558868, -0.254622, -0.340538, 0.400504, 0.165453, 0.038422, 0.035918], [-0.054634, -0.127612, 0.013919, 0.394427, 0.086878, -0.115889, -0.122899]], "network.2.bias": [-0.094535, 0.362044, -0.390503, -0.037301, -0.127581, 0.181996, 0.185857], "network.4.weight": [[0.107421, -0.030958, 0.35389, 0.058009, 0.06546, -0.308401, -0.182185], [0.034969, 0.301186, 0.076099, -0.432218, 0.017777, -0.014173, -0.460212], [-0.068372, -0.195173, -0.34394, 0.553566, 0.320341, -0.40834, 0.191467], [0.080235, -0.227584, 0.194756, 0.158521, -0.039729, 0.51782, 0.427074], [0.156485, 0.100467, -0.00891, 0.121087, 0.184021, 0.085037, -0.169999], [-0.24382, 0.334029, -0.04966, -0.308911, -0.026027, -0.423305, -0.363348], [-0.422606, -0.188874, 0.180272, -0.158127, 0.168404, 0.521194, -0.053758]], "network.4.bias": [0.495019, -0.122641, -0.099683, 0.27665, -0.332098, 0.01975, 0.081703], "network.6.weight": [[-0.491317, -0.109689, 0.212636, 0.234606, 0.034183, -0.470752, -0.016705], [-0.222868, 0.278816, -0.069457, 0.388781, 0.155396, -0.33794, -0.263874], [0.311173, 0.106101, -0.397922, -0.478728, 0.04663, 0.198337, -0.256383], [0.09876, -0.387112, -0.238054, 0.049592, -0.411441, -0.409779, 0.371756], [0.360581, -0.127743, 0.132731, -0.16949, 0.16049, 0.325114, -0.199129], [0.369531, 0.35384, -0.382273, -0.423625, -0.123077, 0.458099, 0.151397], [-0.159262, 0.357932, -0.080052, -0.182123, 0.147139, 0.514723, -0.171692]], "network.6.bias": [0.150254, 0.147178, 0.110088, -0.187977, 0.416406, 0.088358, 0.496323], "network.8.weight": [[0.147247, -0.197612, 0.470659, -0.143807, 0.522217, 0.229661, 0.063608], [0.242909, -0.178817, 0.19516, -0.285272, 0.346293, 0.258141, 0.083183], [0.373456, -0.184102, -0.409404, 0.403549, -0.234311, -0.085581, -0.25229], [-0.017065, -0.000762, 0.016622, -0.520128, -0.001037, -0.241637, -0.174107], [-0.334168, -0.166102, 0.488501, -0.551021, 0.377972, 0.45731, 0.481101], [0.073233, -0.253715, 0.021116, -0.526421, 0.109853, 0.413504, 0.279], [-0.030766, -0.262061, 0.273119, -0.24176, 0.288915, 0.302578, 0.430383]], "network.8.bias": [-0.107985, -0.178579, 0.235906, 0.279974, 0.22508, 0.279491, 0.320933], "network.10.weight": [[-0.270427, -0.28806, 0.362148, -0.317582, -0.251381, 0.125285, -0.171404], [-0.219004, 0.378921, -0.153942, -0.102965, 0.442164, -0.203033, 0.214379], [0.213921, -0.192241, -0.351216, -0.411897, 0.286589, -0.199471, 0.107425], [0.37709, -0.076445, -0.279817, -0.161316, 0.365364, 0.547785, 0.438271], [-0.282994, 0.048191, -0.443896, 0.140703, 0.188858, 0.226648, 0.344082], [-0.402463, 0.111827, 0.439856, 0.399542, -0.478426, -0.118008, 0.242869], [0.286013, 0.316126, -0.224495, 0.293604, 0.084103, 0.431125, 0.048607]], "network.10.bias": [0.123182, 0.055609, -0.252605, 0.177422, 0.023725, -0.304248, -0.081112], "network.12.weight": [[0.254717, -0.403498, -0.373062, -0.430695, -0.306028, 0.183039, -0.221673]], "network.12.bias": [-0.04509]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6925990581512451, "train_acc": 0.55, "val_loss": 0.6800563931465149, "val_acc": 0.62}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6862382292747498, "train_acc": 0.55, "val_loss": 0.6720321774482727, "val_acc": 0.62}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.683437168598175, "train_acc": 0.55, "val_loss": 0.6548855304718018, "val_acc": 0.62}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6728981733322144, "train_acc": 0.55, "val_loss": 0.6142439842224121, "val_acc": 0.62}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6595568060874939, "train_acc": 0.47, "val_loss": 0.5035485625267029, "val_acc": 0.62}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.5558302402496338, "train_acc": 0.56, "val_loss": 0.3484804630279541, "val_acc": 0.98}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.4287234991788864, "train_acc": 0.915, "val_loss": 0.24693666398525238, "val_acc": 0.96}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.34573327004909515, "train_acc": 0.925, "val_loss": 0.19625455141067505, "val_acc": 0.96}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.27985048294067383, "train_acc": 0.92, "val_loss": 0.15912316739559174, "val_acc": 0.96}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.23970410227775574, "train_acc": 0.915, "val_loss": 0.13455447554588318, "val_acc": 0.96}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.20737704634666443, "train_acc": 0.935, "val_loss": 0.12060924619436264, "val_acc": 0.96}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.21174722909927368, "train_acc": 0.93, "val_loss": 0.11693243682384491, "val_acc": 0.96}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.19121137261390686, "train_acc": 0.94, "val_loss": 0.11890479177236557, "val_acc": 0.96}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.16210828348994255, "train_acc": 0.94, "val_loss": 0.11811548471450806, "val_acc": 0.96}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["decreasing_pairs"], "degraded_stage": {"initial_val_loss": 0.6800563931465149, "final_val_loss": 0.6142439842224121, "initial_val_acc": 0.62, "final_val_acc": 0.62, "best_val_acc": 0.62}, "improved_stage": {"initial_val_loss": 0.5035485625267029, "final_val_loss": 0.11811548471450806, "initial_val_acc": 0.62, "final_val_acc": 0.96, "best_val_acc": 0.98, "best_epoch": 5}, "improvement": 0.36, "first_improvement_epoch": 3}}
|
70
|
{"target_pattern": "has_majority", "degraded_accuracy": 0.44, "improved_accuracy": 0.88, "improvement": 0.44, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 7511, "learning_rate": 0.058202799735515814, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "has_majority", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["has_majority"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.232249,
-0.089599,
-0.381569,
-0.124121,
-0.138739
],
[
0.445523,
-1.00762,
-0.014844,
0.586418,
-0.779475
],
[
-1.601324,
-0.02321,
-0.033977,
0.560053,
0.196896
],
[
0.266328,
-0.264567,
-0.388869,
0.56452,
-0.345835
],
[
0.502196,
0.040841,
0.439413,
0.77978,
0.028347
]
],
"network.0.bias": [
-0.498983,
0.263536,
0.05192,
-0.218804,
-0.722873
],
"network.2.weight": [
[
-0.471198,
-0.671588,
-0.814914,
-0.091239,
0.079644
],
[
-0.132745,
-0.008575,
-0.111056,
-0.340297,
0.001706
],
[
-0.032354,
-0.171465,
0.061535,
-0.503325,
0.106531
],
[
0.127947,
-1.144693,
0.49982,
-0.741061,
0.631805
],
[
0.123831,
-0.636336,
-0.801801,
-0.075307,
-0.063449
]
],
"network.2.bias": [
-0.177967,
-0.620576,
-0.297306,
-0.313573,
-0.102971
],
"network.4.weight": [
[
-0.118089,
-0.1402,
-0.245145,
-0.012331,
-0.023531
],
[
-0.227654,
0.257579,
0.327795,
-0.93565,
-0.529092
],
[
0.543037,
-0.067098,
0.04929,
0.871296,
0.213327
],
[
-0.138288,
0.003014,
-1.077467,
-0.936167,
-0.144617
],
[
0.419409,
-0.419134,
-1.071376,
-0.653672,
-0.149053
]
],
"network.4.bias": [
-0.399192,
-0.192668,
-0.136078,
-0.170394,
-0.007044
],
"network.6.weight": [
[
-0.089762,
-0.316624,
0.693059,
0.250023,
-0.306192
],
[
0.08986,
0.03797,
-0.998821,
-0.259401,
0.149863
],
[
0.341852,
-0.429884,
-0.060929,
-0.268395,
0.252624
],
[
-0.034759,
0.045345,
0.752405,
-0.097549,
0.11688
],
[
-0.035388,
-0.409472,
0.739312,
-0.282251,
-0.26178
]
],
"network.6.bias": [
-0.278419,
-0.313106,
-0.448171,
-0.348942,
-0.279146
],
"network.8.weight": [
[
-0.652333,
0.111567,
0.189972,
-0.558595,
-0.226005
]
],
"network.8.bias": [
0.949981
]
}
## Activation Signature
### 0
mean: [-2.185612, -0.752961, -0.610267, -0.397725, 2.601546]
std: [1.192980, 2.037385, 3.509292, 1.423793, 2.252939]
### 2
mean: [-1.092784, -0.856924, -0.225099, 1.045384, -1.367736]
std: [1.227323, 0.315742, 0.444768, 1.764528, 1.279640]
### 4
mean: [-0.436697, -1.398194, 1.023705, -1.469950, -0.916551]
std: [0.063469, 1.339699, 1.316436, 1.526371, 1.077438]
### 6
mean: [0.459726, -1.376904, -0.513063, 0.452409, 0.508260]
std: [0.887869, 1.279578, 0.078055, 0.963897, 0.947123]
### 8
mean: [0.111061]
std: [1.201157]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
has_majority
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.232249,
-0.089599,
-0.381569,
-0.124121,
-0.138739
],
[
0.445523,
-1.00762,
-0.014844,
0.586418,
-0.779475
],
[
-1.601324,
-0.02321,
-0.033977,
0.560053,
0.196896
],
[
0.266328,
-0.264567,
-0.388869,
0.56452,
-0.345835
],
[
0.502196,
0.040841,
0.439413,
0.77978,
0.028347
]
],
"network.0.bias": [
-0.498983,
0.263536,
0.05192,
-0.218804,
-0.722873
],
"network.2.weight": [
[
-0.471198,
-0.671588,
-0.814914,
-0.091239,
0.079644
],
[
-0.132745,
-0.008575,
-0.111056,
-0.340297,
0.001706
],
[
-0.032354,
-0.171465,
0.061535,
-0.503325,
0.106531
],
[
0.127947,
-1.144693,
0.49982,
-0.741061,
0.631805
],
[
0.123831,
-0.636336,
-0.801801,
-0.075307,
-0.063449
]
],
"network.2.bias": [
-0.177967,
-0.620576,
-0.297306,
-0.313573,
-0.102971
],
"network.4.weight": [
[
-0.118089,
-0.1402,
-0.245145,
-0.012331,
-0.023531
],
[
-0.227654,
0.257579,
0.327795,
-0.93565,
-0.529092
],
[
0.543037,
-0.067098,
0.04929,
0.871296,
0.213327
],
[
-0.138288,
0.003014,
-1.077467,
-0.936167,
-0.144617
],
[
0.419409,
-0.419134,
-1.071376,
-0.653672,
-0.149053
]
],
"network.4.bias": [
-0.399192,
-0.192668,
-0.136078,
-0.170394,
-0.007044
],
"network.6.weight": [
[
-0.089762,
-0.316624,
0.693059,
0.250023,
-0.306192
],
[
0.08986,
0.03797,
-0.998821,
-0.259401,
0.149863
],
[
0.341852,
-0.429884,
-0.060929,
-0.268395,
0.252624
],
[
-0.034759,
0.045345,
0.752405,
-0.097549,
0.11688
],
[
-0.035388,
-0.409472,
0.739312,
-0.282251,
-0.26178
]
],
"network.6.bias": [
-0.278419,
-0.313106,
-0.448171,
-0.348942,
-0.279146
],
"network.8.weight": [
[
-0.652333,
0.111567,
0.189972,
-0.558595,
-0.226005
]
],
"network.8.bias": [
0.949981
]
}
## Activation Signature
### 0
mean: [-2.185612, -0.752961, -0.610267, -0.397725, 2.601546]
std: [1.192980, 2.037385, 3.509292, 1.423793, 2.252939]
### 2
mean: [-1.092784, -0.856924, -0.225099, 1.045384, -1.367736]
std: [1.227323, 0.315742, 0.444768, 1.764528, 1.279640]
### 4
mean: [-0.436697, -1.398194, 1.023705, -1.469950, -0.916551]
std: [0.063469, 1.339699, 1.316436, 1.526371, 1.077438]
### 6
mean: [0.459726, -1.376904, -0.513063, 0.452409, 0.508260]
std: [0.887869, 1.279578, 0.078055, 0.963897, 0.947123]
### 8
mean: [0.111061]
std: [1.201157]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
has_majority
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -2.185612440109253, "std": 1.1929796934127808}, "1": {"mean": -0.7529614567756653, "std": 2.037384510040283}, "2": {"mean": -0.6102674007415771, "std": 3.509291648864746}, "3": {"mean": -0.3977254629135132, "std": 1.4237933158874512}, "4": {"mean": 2.601545810699463, "std": 2.252939224243164}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -1.0927841663360596, "std": 1.2273225784301758}, "1": {"mean": -0.8569237589836121, "std": 0.31574153900146484}, "2": {"mean": -0.22509869933128357, "std": 0.44476792216300964}, "3": {"mean": 1.045384407043457, "std": 1.7645282745361328}, "4": {"mean": -1.3677362203598022, "std": 1.2796403169631958}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.4366969168186188, "std": 0.0634685680270195}, "1": {"mean": -1.3981940746307373, "std": 1.339699149131775}, "2": {"mean": 1.023705244064331, "std": 1.3164361715316772}, "3": {"mean": -1.4699498414993286, "std": 1.5263710021972656}, "4": {"mean": -0.9165512919425964, "std": 1.0774378776550293}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.45972582697868347, "std": 0.8878693580627441}, "1": {"mean": -1.3769035339355469, "std": 1.2795779705047607}, "2": {"mean": -0.5130633115768433, "std": 0.07805509865283966}, "3": {"mean": 0.4524090588092804, "std": 0.9638966917991638}, "4": {"mean": 0.5082603693008423, "std": 0.9471234083175659}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.1110605001449585, "std": 1.2011570930480957}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.232249, -0.089599, -0.381569, -0.124121, -0.138739], [0.445523, -1.00762, -0.014844, 0.586418, -0.779475], [-1.601324, -0.02321, -0.033977, 0.560053, 0.196896], [0.266328, -0.264567, -0.388869, 0.56452, -0.345835], [0.502196, 0.040841, 0.439413, 0.77978, 0.028347]], "network.0.bias": [-0.498983, 0.263536, 0.05192, -0.218804, -0.722873], "network.2.weight": [[-0.471198, -0.671588, -0.814914, -0.091239, 0.079644], [-0.132745, -0.008575, -0.111056, -0.340297, 0.001706], [-0.032354, -0.171465, 0.061535, -0.503325, 0.106531], [0.127947, -1.144693, 0.49982, -0.741061, 0.631805], [0.123831, -0.636336, -0.801801, -0.075307, -0.063449]], "network.2.bias": [-0.177967, -0.620576, -0.297306, -0.313573, -0.102971], "network.4.weight": [[-0.118089, -0.1402, -0.245145, -0.012331, -0.023531], [-0.227654, 0.257579, 0.327795, -0.93565, -0.529092], [0.543037, -0.067098, 0.04929, 0.871296, 0.213327], [-0.138288, 0.003014, -1.077467, -0.936167, -0.144617], [0.419409, -0.419134, -1.071376, -0.653672, -0.149053]], "network.4.bias": [-0.399192, -0.192668, -0.136078, -0.170394, -0.007044], "network.6.weight": [[-0.089762, -0.316624, 0.693059, 0.250023, -0.306192], [0.08986, 0.03797, -0.998821, -0.259401, 0.149863], [0.341852, -0.429884, -0.060929, -0.268395, 0.252624], [-0.034759, 0.045345, 0.752405, -0.097549, 0.11688], [-0.035388, -0.409472, 0.739312, -0.282251, -0.26178]], "network.6.bias": [-0.278419, -0.313106, -0.448171, -0.348942, -0.279146], "network.8.weight": [[-0.652333, 0.111567, 0.189972, -0.558595, -0.226005]], "network.8.bias": [0.949981]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6921209990978241, "train_acc": 0.475, "val_loss": 0.7133306860923767, "val_acc": 0.44}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6447307169437408, "train_acc": 0.595, "val_loss": 0.8337544798851013, "val_acc": 0.44}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6655393838882446, "train_acc": 0.595, "val_loss": 0.745745062828064, "val_acc": 0.44}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6355307400226593, "train_acc": 0.595, "val_loss": 0.6932868361473083, "val_acc": 0.44}, {"stage": "degraded", "epoch": 4, "global_epoch": 4, "train_loss": 0.6295746862888336, "train_acc": 0.595, "val_loss": 0.6717107892036438, "val_acc": 0.44}, {"stage": "improved", "epoch": 0, "global_epoch": 5, "train_loss": 0.6267906427383423, "train_acc": 0.525, "val_loss": 0.6188070178031921, "val_acc": 0.58}, {"stage": "improved", "epoch": 1, "global_epoch": 6, "train_loss": 0.5988602936267853, "train_acc": 0.705, "val_loss": 0.5572695732116699, "val_acc": 0.8}, {"stage": "improved", "epoch": 2, "global_epoch": 7, "train_loss": 0.5761023461818695, "train_acc": 0.75, "val_loss": 0.536298394203186, "val_acc": 0.8}, {"stage": "improved", "epoch": 3, "global_epoch": 8, "train_loss": 0.5500390529632568, "train_acc": 0.75, "val_loss": 0.5385434627532959, "val_acc": 0.72}, {"stage": "improved", "epoch": 4, "global_epoch": 9, "train_loss": 0.5561944544315338, "train_acc": 0.75, "val_loss": 0.5029022097587585, "val_acc": 0.72}, {"stage": "improved", "epoch": 5, "global_epoch": 10, "train_loss": 0.5208597332239151, "train_acc": 0.76, "val_loss": 0.45959243178367615, "val_acc": 0.86}, {"stage": "improved", "epoch": 6, "global_epoch": 11, "train_loss": 0.5183462202548981, "train_acc": 0.76, "val_loss": 0.45031723380088806, "val_acc": 0.84}, {"stage": "improved", "epoch": 7, "global_epoch": 12, "train_loss": 0.5114623755216599, "train_acc": 0.78, "val_loss": 0.4318181574344635, "val_acc": 0.86}, {"stage": "improved", "epoch": 8, "global_epoch": 13, "train_loss": 0.5126883834600449, "train_acc": 0.77, "val_loss": 0.4329599142074585, "val_acc": 0.78}, {"stage": "improved", "epoch": 9, "global_epoch": 14, "train_loss": 0.4938611090183258, "train_acc": 0.76, "val_loss": 0.39702948927879333, "val_acc": 0.88}], "summary": {"total_epochs": 15, "degraded_epochs": 5, "improved_epochs": 10, "patterns": ["has_majority"], "degraded_stage": {"initial_val_loss": 0.7133306860923767, "final_val_loss": 0.6717107892036438, "initial_val_acc": 0.44, "final_val_acc": 0.44, "best_val_acc": 0.44}, "improved_stage": {"initial_val_loss": 0.6188070178031921, "final_val_loss": 0.39702948927879333, "initial_val_acc": 0.58, "final_val_acc": 0.88, "best_val_acc": 0.88, "best_epoch": 14}, "improvement": 0.44, "first_improvement_epoch": 4}}
|
71
|
{"target_pattern": "sorted_descending", "degraded_accuracy": 0.8, "improved_accuracy": 0.98, "improvement": 0.17999999999999994, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 1256, "learning_rate": 0.06504650220298641, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_descending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_descending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.596685,
0.037322,
0.551303,
0.373688,
0.039439
],
[
0.748693,
0.477041,
-0.116139,
-0.135784,
-0.072735
],
[
-0.056876,
0.391504,
-0.35265,
0.114321,
-0.049017
],
[
-0.359228,
-0.558949,
-0.037674,
0.041378,
-0.021103
],
[
-0.614111,
0.292091,
0.383786,
-0.183442,
-0.328547
],
[
0.456196,
0.329536,
-0.17682,
-0.195156,
0.769087
],
[
0.471661,
0.301351,
-0.128717,
-0.236226,
-0.011343
],
[
0.148346,
-0.057018,
-0.186736,
-0.394909,
0.224521
]
],
"network.0.bias": [
0.422919,
0.450482,
0.44889,
-0.349599,
0.012565,
-0.16223,
-0.375073,
0.08353
],
"network.2.weight": [
[
0.166651,
0.494397,
-0.360376,
0.209997,
-0.212002,
0.412034,
0.45656,
-0.130972
],
[
0.15028,
-0.313623,
0.413113,
-0.261263,
-0.204943,
-0.149763,
-0.189402,
0.069665
],
[
0.385218,
-0.210699,
0.13944,
-0.453658,
0.411735,
0.216621,
-0.288482,
0.398986
],
[
-0.108968,
0.475195,
0.107534,
0.253196,
0.084434,
0.29867,
0.286772,
-0.401211
],
[
0.54223,
-0.411303,
0.100642,
-0.563381,
0.231576,
0.341064,
-0.163439,
0.170127
],
[
-0.1426,
0.348399,
-0.146927,
-0.1796,
-0.248161,
-0.084045,
0.021219,
-0.355645
],
[
-0.169763,
-0.098606,
0.184815,
0.225433,
-0.113711,
-0.224904,
0.008735,
0.083016
],
[
0.522402,
-0.295362,
0.592868,
-0.225236,
0.477608,
0.44059,
-0.230601,
-0.325832
]
],
"network.2.bias": [
0.077268,
0.431542,
0.170135,
0.316435,
0.528086,
0.060742,
-0.388864,
0.020817
],
"network.4.weight": [
[
0.228492,
-0.010306,
0.348078,
-0.369195,
0.3438,
0.053056,
-0.193726,
0.452355
],
[
-0.355944,
-0.026799,
-0.075358,
-0.128872,
0.108559,
0.396993,
-0.226876,
-0.042438
],
[
-0.349535,
0.380079,
0.048152,
-0.23079,
0.563758,
-0.385421,
-0.428698,
0.311185
],
[
-0.268207,
0.267502,
0.70252,
-0.073729,
0.582154,
0.092543,
0.059399,
0.32088
],
[
0.286625,
-0.65225,
-0.125949,
0.327745,
-0.265147,
0.40045,
-0.401831,
-0.316858
],
[
0.126149,
-0.391586,
-0.006104,
-0.001948,
-0.409954,
0.016461,
-0.169084,
-0.055509
],
[
-0.012996,
0.447129,
-0.14257,
-0.126231,
0.041852,
-0.147404,
0.338622,
0.377803
],
[
0.344338,
-0.475338,
-0.174868,
0.351229,
-0.230307,
0.343101,
-0.290287,
-0.421797
]
],
"network.4.bias": [
-0.140274,
-0.193103,
0.545662,
0.200871,
-0.007214,
-0.22033,
-0.354633,
0.128872
],
"network.6.weight": [
[
0.033602,
0.371752,
-0.273336,
-0.04661,
0.535241,
0.364457,
-0.241527,
0.274275
],
[
0.261011,
0.014355,
0.364383,
0.416911,
-0.251296,
0.176869,
0.026032,
-0.031497
],
[
0.02947,
-0.450542,
-0.118221,
-0.524823,
0.595308,
0.159965,
-0.02282,
0.440339
],
[
-0.087883,
0.141098,
-0.012427,
0.109693,
0.27302,
0.214712,
0.084015,
-0.117522
],
[
0.450468,
-0.225995,
0.387767,
0.539388,
-0.135085,
-0.091759,
-0.060804,
-0.571555
],
[
-0.173524,
0.158937,
0.031365,
0.015448,
0.462213,
-0.118876,
-0.024998,
0.373553
],
[
0.101823,
-0.409898,
-0.527839,
-0.368866,
0.280022,
-0.030438,
-0.372392,
0.602351
],
[
0.215972,
-0.11372,
0.646212,
0.314336,
-0.63938,
-0.352097,
0.348661,
-0.549552
]
],
"network.6.bias": [
0.072764,
-0.268166,
-0.02529,
-0.360862,
0.212107,
-0.305367,
0.07643,
0.512659
],
"network.8.weight": [
[
0.199193,
-0.306008,
0.39077,
0.020452,
-0.689295,
0.071793,
0.453362,
-0.073286
],
[
0.333467,
0.219446,
-0.224445,
0.447165,
0.507589,
-0.470334,
0.037772,
0.079297
],
[
0.082462,
0.038107,
-0.229476,
0.076182,
0.47695,
-0.482394,
-0.138385,
0.427764
],
[
-0.205784,
-0.090951,
0.611525,
-0.505474,
-0.453161,
-0.121246,
0.0928,
-0.39564
],
[
-0.192048,
0.684448,
-0.111127,
0.094889,
0.067834,
0.064493,
-0.161234,
0.261972
],
[
-0.049653,
-0.654982,
0.032983,
-0.146327,
-0.16595,
-0.1284,
0.229533,
-0.052876
],
[
-0.393405,
0.570288,
-0.421465,
-0.04276,
0.507338,
-0.516563,
-0.466888,
0.537194
],
[
-0.40626,
0.085595,
0.026109,
0.102987,
0.333405,
-0.426527,
-0.433712,
0.603357
]
],
"network.8.bias": [
0.31101,
-0.160468,
0.283581,
0.223941,
0.149017,
-0.198597,
0.236432,
0.140071
],
"network.10.weight": [
[
0.270746,
-0.123024,
-0.374353,
0.231973,
-0.172803,
0.352092,
-0.548568,
-0.55429
]
],
"network.10.bias": [
0.158551
]
}
## Activation Signature
### 0
mean: [1.749992, 1.627972, 0.539571, -1.831160, -0.216311, 1.166652, -0.028428, -0.795784]
std: [1.597454, 1.909696, 0.991359, 1.461503, 1.421382, 1.889577, 1.240295, 0.989044]
### 2
mean: [1.544388, 0.135887, 0.974510, 1.436533, 1.426282, 0.092060, -1.088270, 1.526039]
std: [1.820748, 1.040950, 1.029235, 1.612382, 1.317880, 0.751589, 0.495199, 1.300224]
### 4
mean: [1.262920, -0.763281, 1.185466, 1.857787, -0.244371, -0.853752, 0.041870, -0.108130]
std: [1.352718, 0.686214, 2.066153, 2.006952, 1.959574, 0.779739, 0.717463, 2.100794]
### 6
mean: [0.010586, 1.309285, -0.505229, -0.196897, 2.021831, 0.093916, -0.799470, 1.763204]
std: [1.548731, 1.767228, 2.275219, 0.188334, 2.750690, 1.254406, 2.351443, 3.153372]
### 8
mean: [-1.315246, 1.431028, 2.172262, -1.632306, 1.728912, -1.605527, 2.671731, 1.943646]
std: [3.200770, 1.777376, 2.591404, 2.297334, 2.049294, 1.523551, 4.739734, 3.187331]
### 10
mean: [-4.534178]
std: [4.544732]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_descending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.596685,
0.037322,
0.551303,
0.373688,
0.039439
],
[
0.748693,
0.477041,
-0.116139,
-0.135784,
-0.072735
],
[
-0.056876,
0.391504,
-0.35265,
0.114321,
-0.049017
],
[
-0.359228,
-0.558949,
-0.037674,
0.041378,
-0.021103
],
[
-0.614111,
0.292091,
0.383786,
-0.183442,
-0.328547
],
[
0.456196,
0.329536,
-0.17682,
-0.195156,
0.769087
],
[
0.471661,
0.301351,
-0.128717,
-0.236226,
-0.011343
],
[
0.148346,
-0.057018,
-0.186736,
-0.394909,
0.224521
]
],
"network.0.bias": [
0.422919,
0.450482,
0.44889,
-0.349599,
0.012565,
-0.16223,
-0.375073,
0.08353
],
"network.2.weight": [
[
0.166651,
0.494397,
-0.360376,
0.209997,
-0.212002,
0.412034,
0.45656,
-0.130972
],
[
0.15028,
-0.313623,
0.413113,
-0.261263,
-0.204943,
-0.149763,
-0.189402,
0.069665
],
[
0.385218,
-0.210699,
0.13944,
-0.453658,
0.411735,
0.216621,
-0.288482,
0.398986
],
[
-0.108968,
0.475195,
0.107534,
0.253196,
0.084434,
0.29867,
0.286772,
-0.401211
],
[
0.54223,
-0.411303,
0.100642,
-0.563381,
0.231576,
0.341064,
-0.163439,
0.170127
],
[
-0.1426,
0.348399,
-0.146927,
-0.1796,
-0.248161,
-0.084045,
0.021219,
-0.355645
],
[
-0.169763,
-0.098606,
0.184815,
0.225433,
-0.113711,
-0.224904,
0.008735,
0.083016
],
[
0.522402,
-0.295362,
0.592868,
-0.225236,
0.477608,
0.44059,
-0.230601,
-0.325832
]
],
"network.2.bias": [
0.077268,
0.431542,
0.170135,
0.316435,
0.528086,
0.060742,
-0.388864,
0.020817
],
"network.4.weight": [
[
0.228492,
-0.010306,
0.348078,
-0.369195,
0.3438,
0.053056,
-0.193726,
0.452355
],
[
-0.355944,
-0.026799,
-0.075358,
-0.128872,
0.108559,
0.396993,
-0.226876,
-0.042438
],
[
-0.349535,
0.380079,
0.048152,
-0.23079,
0.563758,
-0.385421,
-0.428698,
0.311185
],
[
-0.268207,
0.267502,
0.70252,
-0.073729,
0.582154,
0.092543,
0.059399,
0.32088
],
[
0.286625,
-0.65225,
-0.125949,
0.327745,
-0.265147,
0.40045,
-0.401831,
-0.316858
],
[
0.126149,
-0.391586,
-0.006104,
-0.001948,
-0.409954,
0.016461,
-0.169084,
-0.055509
],
[
-0.012996,
0.447129,
-0.14257,
-0.126231,
0.041852,
-0.147404,
0.338622,
0.377803
],
[
0.344338,
-0.475338,
-0.174868,
0.351229,
-0.230307,
0.343101,
-0.290287,
-0.421797
]
],
"network.4.bias": [
-0.140274,
-0.193103,
0.545662,
0.200871,
-0.007214,
-0.22033,
-0.354633,
0.128872
],
"network.6.weight": [
[
0.033602,
0.371752,
-0.273336,
-0.04661,
0.535241,
0.364457,
-0.241527,
0.274275
],
[
0.261011,
0.014355,
0.364383,
0.416911,
-0.251296,
0.176869,
0.026032,
-0.031497
],
[
0.02947,
-0.450542,
-0.118221,
-0.524823,
0.595308,
0.159965,
-0.02282,
0.440339
],
[
-0.087883,
0.141098,
-0.012427,
0.109693,
0.27302,
0.214712,
0.084015,
-0.117522
],
[
0.450468,
-0.225995,
0.387767,
0.539388,
-0.135085,
-0.091759,
-0.060804,
-0.571555
],
[
-0.173524,
0.158937,
0.031365,
0.015448,
0.462213,
-0.118876,
-0.024998,
0.373553
],
[
0.101823,
-0.409898,
-0.527839,
-0.368866,
0.280022,
-0.030438,
-0.372392,
0.602351
],
[
0.215972,
-0.11372,
0.646212,
0.314336,
-0.63938,
-0.352097,
0.348661,
-0.549552
]
],
"network.6.bias": [
0.072764,
-0.268166,
-0.02529,
-0.360862,
0.212107,
-0.305367,
0.07643,
0.512659
],
"network.8.weight": [
[
0.199193,
-0.306008,
0.39077,
0.020452,
-0.689295,
0.071793,
0.453362,
-0.073286
],
[
0.333467,
0.219446,
-0.224445,
0.447165,
0.507589,
-0.470334,
0.037772,
0.079297
],
[
0.082462,
0.038107,
-0.229476,
0.076182,
0.47695,
-0.482394,
-0.138385,
0.427764
],
[
-0.205784,
-0.090951,
0.611525,
-0.505474,
-0.453161,
-0.121246,
0.0928,
-0.39564
],
[
-0.192048,
0.684448,
-0.111127,
0.094889,
0.067834,
0.064493,
-0.161234,
0.261972
],
[
-0.049653,
-0.654982,
0.032983,
-0.146327,
-0.16595,
-0.1284,
0.229533,
-0.052876
],
[
-0.393405,
0.570288,
-0.421465,
-0.04276,
0.507338,
-0.516563,
-0.466888,
0.537194
],
[
-0.40626,
0.085595,
0.026109,
0.102987,
0.333405,
-0.426527,
-0.433712,
0.603357
]
],
"network.8.bias": [
0.31101,
-0.160468,
0.283581,
0.223941,
0.149017,
-0.198597,
0.236432,
0.140071
],
"network.10.weight": [
[
0.270746,
-0.123024,
-0.374353,
0.231973,
-0.172803,
0.352092,
-0.548568,
-0.55429
]
],
"network.10.bias": [
0.158551
]
}
## Activation Signature
### 0
mean: [1.749992, 1.627972, 0.539571, -1.831160, -0.216311, 1.166652, -0.028428, -0.795784]
std: [1.597454, 1.909696, 0.991359, 1.461503, 1.421382, 1.889577, 1.240295, 0.989044]
### 2
mean: [1.544388, 0.135887, 0.974510, 1.436533, 1.426282, 0.092060, -1.088270, 1.526039]
std: [1.820748, 1.040950, 1.029235, 1.612382, 1.317880, 0.751589, 0.495199, 1.300224]
### 4
mean: [1.262920, -0.763281, 1.185466, 1.857787, -0.244371, -0.853752, 0.041870, -0.108130]
std: [1.352718, 0.686214, 2.066153, 2.006952, 1.959574, 0.779739, 0.717463, 2.100794]
### 6
mean: [0.010586, 1.309285, -0.505229, -0.196897, 2.021831, 0.093916, -0.799470, 1.763204]
std: [1.548731, 1.767228, 2.275219, 0.188334, 2.750690, 1.254406, 2.351443, 3.153372]
### 8
mean: [-1.315246, 1.431028, 2.172262, -1.632306, 1.728912, -1.605527, 2.671731, 1.943646]
std: [3.200770, 1.777376, 2.591404, 2.297334, 2.049294, 1.523551, 4.739734, 3.187331]
### 10
mean: [-4.534178]
std: [4.544732]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_descending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.7499916553497314, "std": 1.5974535942077637}, "1": {"mean": 1.627971887588501, "std": 1.909696340560913}, "2": {"mean": 0.5395708680152893, "std": 0.9913590550422668}, "3": {"mean": -1.8311598300933838, "std": 1.4615031480789185}, "4": {"mean": -0.2163105458021164, "std": 1.4213817119598389}, "5": {"mean": 1.166651964187622, "std": 1.8895773887634277}, "6": {"mean": -0.028427982702851295, "std": 1.2402946949005127}, "7": {"mean": -0.7957836389541626, "std": 0.9890443086624146}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 1.544387936592102, "std": 1.8207483291625977}, "1": {"mean": 0.1358865201473236, "std": 1.0409502983093262}, "2": {"mean": 0.9745096564292908, "std": 1.029234528541565}, "3": {"mean": 1.436532974243164, "std": 1.612382173538208}, "4": {"mean": 1.4262816905975342, "std": 1.3178801536560059}, "5": {"mean": 0.0920599177479744, "std": 0.7515890002250671}, "6": {"mean": -1.0882701873779297, "std": 0.49519863724708557}, "7": {"mean": 1.526039481163025, "std": 1.3002243041992188}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 1.2629199028015137, "std": 1.3527183532714844}, "1": {"mean": -0.7632806301116943, "std": 0.6862136721611023}, "2": {"mean": 1.185465931892395, "std": 2.066152811050415}, "3": {"mean": 1.8577865362167358, "std": 2.0069515705108643}, "4": {"mean": -0.24437077343463898, "std": 1.9595739841461182}, "5": {"mean": -0.8537519574165344, "std": 0.7797390818595886}, "6": {"mean": 0.04187003895640373, "std": 0.7174625396728516}, "7": {"mean": -0.10812953114509583, "std": 2.1007940769195557}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.010586048476397991, "std": 1.5487310886383057}, "1": {"mean": 1.3092849254608154, "std": 1.7672278881072998}, "2": {"mean": -0.505228579044342, "std": 2.275219202041626}, "3": {"mean": -0.1968967616558075, "std": 0.18833421170711517}, "4": {"mean": 2.0218310356140137, "std": 2.750690460205078}, "5": {"mean": 0.09391649812459946, "std": 1.2544063329696655}, "6": {"mean": -0.7994695901870728, "std": 2.35144305229187}, "7": {"mean": 1.7632040977478027, "std": 3.153372287750244}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.3152459859848022, "std": 3.2007696628570557}, "1": {"mean": 1.4310275316238403, "std": 1.7773758172988892}, "2": {"mean": 2.172261953353882, "std": 2.5914037227630615}, "3": {"mean": -1.6323058605194092, "std": 2.2973341941833496}, "4": {"mean": 1.728912353515625, "std": 2.0492939949035645}, "5": {"mean": -1.6055270433425903, "std": 1.5235514640808105}, "6": {"mean": 2.6717309951782227, "std": 4.739734172821045}, "7": {"mean": 1.9436460733413696, "std": 3.187331199645996}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -4.534177780151367, "std": 4.544731616973877}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.596685, 0.037322, 0.551303, 0.373688, 0.039439], [0.748693, 0.477041, -0.116139, -0.135784, -0.072735], [-0.056876, 0.391504, -0.35265, 0.114321, -0.049017], [-0.359228, -0.558949, -0.037674, 0.041378, -0.021103], [-0.614111, 0.292091, 0.383786, -0.183442, -0.328547], [0.456196, 0.329536, -0.17682, -0.195156, 0.769087], [0.471661, 0.301351, -0.128717, -0.236226, -0.011343], [0.148346, -0.057018, -0.186736, -0.394909, 0.224521]], "network.0.bias": [0.422919, 0.450482, 0.44889, -0.349599, 0.012565, -0.16223, -0.375073, 0.08353], "network.2.weight": [[0.166651, 0.494397, -0.360376, 0.209997, -0.212002, 0.412034, 0.45656, -0.130972], [0.15028, -0.313623, 0.413113, -0.261263, -0.204943, -0.149763, -0.189402, 0.069665], [0.385218, -0.210699, 0.13944, -0.453658, 0.411735, 0.216621, -0.288482, 0.398986], [-0.108968, 0.475195, 0.107534, 0.253196, 0.084434, 0.29867, 0.286772, -0.401211], [0.54223, -0.411303, 0.100642, -0.563381, 0.231576, 0.341064, -0.163439, 0.170127], [-0.1426, 0.348399, -0.146927, -0.1796, -0.248161, -0.084045, 0.021219, -0.355645], [-0.169763, -0.098606, 0.184815, 0.225433, -0.113711, -0.224904, 0.008735, 0.083016], [0.522402, -0.295362, 0.592868, -0.225236, 0.477608, 0.44059, -0.230601, -0.325832]], "network.2.bias": [0.077268, 0.431542, 0.170135, 0.316435, 0.528086, 0.060742, -0.388864, 0.020817], "network.4.weight": [[0.228492, -0.010306, 0.348078, -0.369195, 0.3438, 0.053056, -0.193726, 0.452355], [-0.355944, -0.026799, -0.075358, -0.128872, 0.108559, 0.396993, -0.226876, -0.042438], [-0.349535, 0.380079, 0.048152, -0.23079, 0.563758, -0.385421, -0.428698, 0.311185], [-0.268207, 0.267502, 0.70252, -0.073729, 0.582154, 0.092543, 0.059399, 0.32088], [0.286625, -0.65225, -0.125949, 0.327745, -0.265147, 0.40045, -0.401831, -0.316858], [0.126149, -0.391586, -0.006104, -0.001948, -0.409954, 0.016461, -0.169084, -0.055509], [-0.012996, 0.447129, -0.14257, -0.126231, 0.041852, -0.147404, 0.338622, 0.377803], [0.344338, -0.475338, -0.174868, 0.351229, -0.230307, 0.343101, -0.290287, -0.421797]], "network.4.bias": [-0.140274, -0.193103, 0.545662, 0.200871, -0.007214, -0.22033, -0.354633, 0.128872], "network.6.weight": [[0.033602, 0.371752, -0.273336, -0.04661, 0.535241, 0.364457, -0.241527, 0.274275], [0.261011, 0.014355, 0.364383, 0.416911, -0.251296, 0.176869, 0.026032, -0.031497], [0.02947, -0.450542, -0.118221, -0.524823, 0.595308, 0.159965, -0.02282, 0.440339], [-0.087883, 0.141098, -0.012427, 0.109693, 0.27302, 0.214712, 0.084015, -0.117522], [0.450468, -0.225995, 0.387767, 0.539388, -0.135085, -0.091759, -0.060804, -0.571555], [-0.173524, 0.158937, 0.031365, 0.015448, 0.462213, -0.118876, -0.024998, 0.373553], [0.101823, -0.409898, -0.527839, -0.368866, 0.280022, -0.030438, -0.372392, 0.602351], [0.215972, -0.11372, 0.646212, 0.314336, -0.63938, -0.352097, 0.348661, -0.549552]], "network.6.bias": [0.072764, -0.268166, -0.02529, -0.360862, 0.212107, -0.305367, 0.07643, 0.512659], "network.8.weight": [[0.199193, -0.306008, 0.39077, 0.020452, -0.689295, 0.071793, 0.453362, -0.073286], [0.333467, 0.219446, -0.224445, 0.447165, 0.507589, -0.470334, 0.037772, 0.079297], [0.082462, 0.038107, -0.229476, 0.076182, 0.47695, -0.482394, -0.138385, 0.427764], [-0.205784, -0.090951, 0.611525, -0.505474, -0.453161, -0.121246, 0.0928, -0.39564], [-0.192048, 0.684448, -0.111127, 0.094889, 0.067834, 0.064493, -0.161234, 0.261972], [-0.049653, -0.654982, 0.032983, -0.146327, -0.16595, -0.1284, 0.229533, -0.052876], [-0.393405, 0.570288, -0.421465, -0.04276, 0.507338, -0.516563, -0.466888, 0.537194], [-0.40626, 0.085595, 0.026109, 0.102987, 0.333405, -0.426527, -0.433712, 0.603357]], "network.8.bias": [0.31101, -0.160468, 0.283581, 0.223941, 0.149017, -0.198597, 0.236432, 0.140071], "network.10.weight": [[0.270746, -0.123024, -0.374353, 0.231973, -0.172803, 0.352092, -0.548568, -0.55429]], "network.10.bias": [0.158551]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7026453018188477, "train_acc": 0.445, "val_loss": 0.6509823799133301, "val_acc": 0.54}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6007198393344879, "train_acc": 0.575, "val_loss": 0.39959633350372314, "val_acc": 0.8}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.4010734409093857, "train_acc": 0.785, "val_loss": 0.12947005033493042, "val_acc": 0.98}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.32154275476932526, "train_acc": 0.91, "val_loss": 0.10970264673233032, "val_acc": 0.98}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.32425881922245026, "train_acc": 0.91, "val_loss": 0.14544883370399475, "val_acc": 0.98}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.2784509211778641, "train_acc": 0.9, "val_loss": 0.1939259171485901, "val_acc": 0.98}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.2696443349123001, "train_acc": 0.925, "val_loss": 0.19841693341732025, "val_acc": 0.98}], "summary": {"total_epochs": 7, "degraded_epochs": 2, "improved_epochs": 5, "patterns": ["sorted_descending"], "degraded_stage": {"initial_val_loss": 0.6509823799133301, "final_val_loss": 0.39959633350372314, "initial_val_acc": 0.54, "final_val_acc": 0.8, "best_val_acc": 0.8}, "improved_stage": {"initial_val_loss": 0.12947005033493042, "final_val_loss": 0.19841693341732025, "initial_val_acc": 0.98, "final_val_acc": 0.98, "best_val_acc": 0.98, "best_epoch": 2}, "improvement": 0.17999999999999994, "first_improvement_epoch": 1}}
|
72
|
{"target_pattern": "first_last_match", "degraded_accuracy": 0.5, "improved_accuracy": 0.82, "improvement": 0.31999999999999995, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 2617, "learning_rate": 0.03671955834795707, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "first_last_match", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["first_last_match"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.061402,
-0.539979,
-0.301582,
0.062127,
-0.382179
],
[
-0.526495,
0.075349,
0.234699,
-0.660663,
-0.091098
],
[
0.544276,
0.092542,
-0.106422,
-0.211367,
-0.054051
],
[
-0.275536,
0.251985,
0.175351,
0.356682,
-0.557291
],
[
0.171428,
-0.232364,
0.336346,
0.502101,
0.636576
],
[
0.125534,
-0.265525,
0.335271,
0.233784,
-0.456593
],
[
-0.625502,
-0.202451,
-0.456845,
-0.070338,
0.159243
]
],
"network.0.bias": [
0.126051,
0.251975,
0.105147,
0.052979,
-0.345109,
0.361127,
-0.330503
],
"network.2.weight": [
[
0.277286,
0.050435,
0.227302,
-0.151232,
0.525679,
-0.359594,
-0.066744
],
[
0.449886,
0.129804,
-0.30103,
-0.153,
-0.12505,
-0.107083,
0.337075
],
[
-0.110874,
-0.594766,
-0.066758,
0.18376,
-0.245425,
-0.459937,
-0.008792
],
[
0.356128,
-0.011534,
-0.540613,
0.504914,
0.152244,
0.067588,
0.252981
],
[
0.204005,
-0.265807,
-0.523937,
0.021056,
-0.131794,
0.346983,
0.181182
],
[
0.031416,
-0.030197,
0.586417,
-0.452953,
0.029597,
0.088031,
-0.524497
],
[
-0.249682,
-0.116707,
-0.316699,
0.40476,
-0.121422,
0.43929,
0.112219
]
],
"network.2.bias": [
-0.071448,
-0.161408,
-0.205377,
0.322627,
-0.036181,
0.452703,
0.509819
],
"network.4.weight": [
[
0.352654,
0.070746,
-0.022907,
0.187998,
-0.12705,
0.175294,
-0.430051
],
[
0.515195,
-0.303799,
-0.298662,
-0.178247,
-0.060773,
0.423981,
-0.180448
],
[
0.173546,
-0.345679,
-0.037123,
-0.009477,
-0.489889,
0.542702,
-0.10542
],
[
0.491912,
-0.409568,
-0.385634,
0.046023,
-0.755931,
0.418804,
0.128732
],
[
0.490514,
-0.112851,
-0.134828,
-0.211843,
-0.375892,
0.511791,
-0.029572
],
[
0.379544,
-0.392994,
-0.163876,
0.16602,
-0.649836,
0.128325,
-0.280772
],
[
-0.635526,
-0.300475,
-0.039085,
0.426948,
0.260295,
-0.701499,
0.503013
]
],
"network.4.bias": [
0.309971,
0.411857,
-0.011034,
0.456562,
0.152772,
0.501498,
0.019375
],
"network.6.weight": [
[
0.246206,
0.250145,
0.359046,
0.031473,
0.070385,
0.488992,
-0.505569
],
[
-0.2395,
-0.289319,
-0.051755,
-0.221213,
-0.131364,
0.103439,
0.267281
],
[
-0.006289,
-0.422078,
-0.237106,
0.210193,
-0.528331,
-0.375473,
0.459153
],
[
0.498259,
0.348299,
0.022752,
0.360933,
0.83631,
0.586242,
-0.523245
],
[
0.210865,
-0.087679,
0.240401,
-0.202645,
0.155481,
-0.369855,
0.057707
],
[
-0.443918,
-0.514027,
-0.278174,
0.054617,
-0.761217,
-0.024865,
0.390501
],
[
-0.101467,
-0.455174,
-0.186443,
0.233443,
-0.344226,
-0.161118,
0.232445
]
],
"network.6.bias": [
0.309134,
0.141367,
0.04286,
-0.042668,
0.107121,
0.344412,
0.029631
],
"network.8.weight": [
[
0.52687,
-0.419017,
-0.047539,
0.645399,
-0.34217,
-0.359562,
0.112181
],
[
0.761181,
0.198452,
-0.244942,
0.576437,
0.165253,
-0.240716,
-0.115303
],
[
-0.116864,
0.362221,
0.368643,
-0.061728,
0.046349,
0.552548,
0.186458
],
[
0.354594,
-0.244243,
-0.243833,
0.622026,
0.09783,
-0.330105,
-0.356442
],
[
0.663341,
-0.23285,
-0.254573,
0.593507,
-0.203401,
-0.038417,
-0.524096
],
[
0.605931,
-0.127466,
-0.210905,
0.550326,
-0.193815,
-0.608604,
-0.142199
],
[
0.499598,
0.403515,
-0.427485,
0.164167,
0.157335,
-0.562749,
-0.233993
]
],
"network.8.bias": [
-0.02264,
-0.061796,
0.319876,
0.466949,
0.388628,
0.512047,
0.419899
],
"network.10.weight": [
[
-0.041589,
-0.103941,
0.36678,
-0.006833,
0.033576,
-0.166818,
0.308698
],
[
0.25469,
0.388908,
-0.301865,
0.597294,
0.537167,
0.193058,
0.646257
],
[
0.204361,
-0.290486,
0.400023,
-0.255669,
-0.573081,
-0.460171,
-0.550333
],
[
0.682125,
0.219787,
0.001545,
0.017017,
0.180547,
0.65568,
0.381042
],
[
0.642075,
0.142328,
-0.317932,
0.335878,
0.207896,
0.486071,
0.317124
],
[
-0.021983,
0.293549,
-0.150462,
0.642878,
0.365983,
0.599289,
0.597694
],
[
-0.282959,
-0.304494,
0.284807,
-0.388752,
-0.370813,
-0.468201,
-0.279479
]
],
"network.10.bias": [
-0.264399,
0.074336,
0.27518,
-0.053512,
0.019763,
0.050142,
0.297729
],
"network.12.weight": [
[
0.101121,
-0.67525,
0.445356,
-0.260761,
-0.556258,
-0.57634,
0.509278
]
],
"network.12.bias": [
0.180069
]
}
## Activation Signature
### 0
mean: [-1.902419, -1.304766, 0.208867, 0.613108, 2.009232, 0.672051, -2.386551]
std: [1.532663, 1.657228, 1.201374, 1.461722, 1.913497, 0.978838, 1.892316]
### 2
mean: [0.702409, -0.771040, -0.950633, 0.815610, -0.321216, 0.487878, 0.756038]
std: [1.120631, 0.413409, 0.647862, 0.905589, 0.635502, 0.848964, 0.851865]
### 4
mean: [0.471962, 0.781856, 0.353427, 1.238314, 0.590818, 0.832144, -0.024930]
std: [0.650109, 0.932323, 0.612019, 0.696912, 0.912293, 0.624243, 1.580647]
### 6
mean: [0.872188, -0.297499, -0.449404, 1.463765, -0.155886, -0.510786, -0.344853]
std: [1.234204, 0.784854, 1.285491, 2.223279, 0.160497, 1.674161, 0.902269]
### 8
mean: [1.394364, 1.446287, 0.397341, 1.599795, 1.856886, 1.723308, 0.879271]
std: [2.088452, 2.132052, 0.793093, 1.943749, 2.153765, 2.122089, 1.210269]
### 10
mean: [-0.294861, 3.930802, -2.578936, 3.265737, 3.235517, 3.845544, -2.966269]
std: [0.444134, 4.675668, 3.529103, 3.846758, 4.014588, 4.358839, 3.990905]
### 12
mean: [-7.359863]
std: [8.894181]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
first_last_match
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.061402,
-0.539979,
-0.301582,
0.062127,
-0.382179
],
[
-0.526495,
0.075349,
0.234699,
-0.660663,
-0.091098
],
[
0.544276,
0.092542,
-0.106422,
-0.211367,
-0.054051
],
[
-0.275536,
0.251985,
0.175351,
0.356682,
-0.557291
],
[
0.171428,
-0.232364,
0.336346,
0.502101,
0.636576
],
[
0.125534,
-0.265525,
0.335271,
0.233784,
-0.456593
],
[
-0.625502,
-0.202451,
-0.456845,
-0.070338,
0.159243
]
],
"network.0.bias": [
0.126051,
0.251975,
0.105147,
0.052979,
-0.345109,
0.361127,
-0.330503
],
"network.2.weight": [
[
0.277286,
0.050435,
0.227302,
-0.151232,
0.525679,
-0.359594,
-0.066744
],
[
0.449886,
0.129804,
-0.30103,
-0.153,
-0.12505,
-0.107083,
0.337075
],
[
-0.110874,
-0.594766,
-0.066758,
0.18376,
-0.245425,
-0.459937,
-0.008792
],
[
0.356128,
-0.011534,
-0.540613,
0.504914,
0.152244,
0.067588,
0.252981
],
[
0.204005,
-0.265807,
-0.523937,
0.021056,
-0.131794,
0.346983,
0.181182
],
[
0.031416,
-0.030197,
0.586417,
-0.452953,
0.029597,
0.088031,
-0.524497
],
[
-0.249682,
-0.116707,
-0.316699,
0.40476,
-0.121422,
0.43929,
0.112219
]
],
"network.2.bias": [
-0.071448,
-0.161408,
-0.205377,
0.322627,
-0.036181,
0.452703,
0.509819
],
"network.4.weight": [
[
0.352654,
0.070746,
-0.022907,
0.187998,
-0.12705,
0.175294,
-0.430051
],
[
0.515195,
-0.303799,
-0.298662,
-0.178247,
-0.060773,
0.423981,
-0.180448
],
[
0.173546,
-0.345679,
-0.037123,
-0.009477,
-0.489889,
0.542702,
-0.10542
],
[
0.491912,
-0.409568,
-0.385634,
0.046023,
-0.755931,
0.418804,
0.128732
],
[
0.490514,
-0.112851,
-0.134828,
-0.211843,
-0.375892,
0.511791,
-0.029572
],
[
0.379544,
-0.392994,
-0.163876,
0.16602,
-0.649836,
0.128325,
-0.280772
],
[
-0.635526,
-0.300475,
-0.039085,
0.426948,
0.260295,
-0.701499,
0.503013
]
],
"network.4.bias": [
0.309971,
0.411857,
-0.011034,
0.456562,
0.152772,
0.501498,
0.019375
],
"network.6.weight": [
[
0.246206,
0.250145,
0.359046,
0.031473,
0.070385,
0.488992,
-0.505569
],
[
-0.2395,
-0.289319,
-0.051755,
-0.221213,
-0.131364,
0.103439,
0.267281
],
[
-0.006289,
-0.422078,
-0.237106,
0.210193,
-0.528331,
-0.375473,
0.459153
],
[
0.498259,
0.348299,
0.022752,
0.360933,
0.83631,
0.586242,
-0.523245
],
[
0.210865,
-0.087679,
0.240401,
-0.202645,
0.155481,
-0.369855,
0.057707
],
[
-0.443918,
-0.514027,
-0.278174,
0.054617,
-0.761217,
-0.024865,
0.390501
],
[
-0.101467,
-0.455174,
-0.186443,
0.233443,
-0.344226,
-0.161118,
0.232445
]
],
"network.6.bias": [
0.309134,
0.141367,
0.04286,
-0.042668,
0.107121,
0.344412,
0.029631
],
"network.8.weight": [
[
0.52687,
-0.419017,
-0.047539,
0.645399,
-0.34217,
-0.359562,
0.112181
],
[
0.761181,
0.198452,
-0.244942,
0.576437,
0.165253,
-0.240716,
-0.115303
],
[
-0.116864,
0.362221,
0.368643,
-0.061728,
0.046349,
0.552548,
0.186458
],
[
0.354594,
-0.244243,
-0.243833,
0.622026,
0.09783,
-0.330105,
-0.356442
],
[
0.663341,
-0.23285,
-0.254573,
0.593507,
-0.203401,
-0.038417,
-0.524096
],
[
0.605931,
-0.127466,
-0.210905,
0.550326,
-0.193815,
-0.608604,
-0.142199
],
[
0.499598,
0.403515,
-0.427485,
0.164167,
0.157335,
-0.562749,
-0.233993
]
],
"network.8.bias": [
-0.02264,
-0.061796,
0.319876,
0.466949,
0.388628,
0.512047,
0.419899
],
"network.10.weight": [
[
-0.041589,
-0.103941,
0.36678,
-0.006833,
0.033576,
-0.166818,
0.308698
],
[
0.25469,
0.388908,
-0.301865,
0.597294,
0.537167,
0.193058,
0.646257
],
[
0.204361,
-0.290486,
0.400023,
-0.255669,
-0.573081,
-0.460171,
-0.550333
],
[
0.682125,
0.219787,
0.001545,
0.017017,
0.180547,
0.65568,
0.381042
],
[
0.642075,
0.142328,
-0.317932,
0.335878,
0.207896,
0.486071,
0.317124
],
[
-0.021983,
0.293549,
-0.150462,
0.642878,
0.365983,
0.599289,
0.597694
],
[
-0.282959,
-0.304494,
0.284807,
-0.388752,
-0.370813,
-0.468201,
-0.279479
]
],
"network.10.bias": [
-0.264399,
0.074336,
0.27518,
-0.053512,
0.019763,
0.050142,
0.297729
],
"network.12.weight": [
[
0.101121,
-0.67525,
0.445356,
-0.260761,
-0.556258,
-0.57634,
0.509278
]
],
"network.12.bias": [
0.180069
]
}
## Activation Signature
### 0
mean: [-1.902419, -1.304766, 0.208867, 0.613108, 2.009232, 0.672051, -2.386551]
std: [1.532663, 1.657228, 1.201374, 1.461722, 1.913497, 0.978838, 1.892316]
### 2
mean: [0.702409, -0.771040, -0.950633, 0.815610, -0.321216, 0.487878, 0.756038]
std: [1.120631, 0.413409, 0.647862, 0.905589, 0.635502, 0.848964, 0.851865]
### 4
mean: [0.471962, 0.781856, 0.353427, 1.238314, 0.590818, 0.832144, -0.024930]
std: [0.650109, 0.932323, 0.612019, 0.696912, 0.912293, 0.624243, 1.580647]
### 6
mean: [0.872188, -0.297499, -0.449404, 1.463765, -0.155886, -0.510786, -0.344853]
std: [1.234204, 0.784854, 1.285491, 2.223279, 0.160497, 1.674161, 0.902269]
### 8
mean: [1.394364, 1.446287, 0.397341, 1.599795, 1.856886, 1.723308, 0.879271]
std: [2.088452, 2.132052, 0.793093, 1.943749, 2.153765, 2.122089, 1.210269]
### 10
mean: [-0.294861, 3.930802, -2.578936, 3.265737, 3.235517, 3.845544, -2.966269]
std: [0.444134, 4.675668, 3.529103, 3.846758, 4.014588, 4.358839, 3.990905]
### 12
mean: [-7.359863]
std: [8.894181]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
first_last_match
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.9024193286895752, "std": 1.5326627492904663}, "1": {"mean": -1.3047659397125244, "std": 1.6572279930114746}, "2": {"mean": 0.20886731147766113, "std": 1.2013740539550781}, "3": {"mean": 0.613108217716217, "std": 1.4617215394973755}, "4": {"mean": 2.0092315673828125, "std": 1.913496971130371}, "5": {"mean": 0.6720507144927979, "std": 0.978837788105011}, "6": {"mean": -2.3865506649017334, "std": 1.8923158645629883}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.7024086117744446, "std": 1.120631217956543}, "1": {"mean": -0.7710404396057129, "std": 0.41340941190719604}, "2": {"mean": -0.9506327509880066, "std": 0.6478621959686279}, "3": {"mean": 0.8156101703643799, "std": 0.9055891036987305}, "4": {"mean": -0.3212164044380188, "std": 0.6355018019676208}, "5": {"mean": 0.4878782629966736, "std": 0.8489644527435303}, "6": {"mean": 0.7560376524925232, "std": 0.8518645763397217}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.47196224331855774, "std": 0.650109052658081}, "1": {"mean": 0.781855583190918, "std": 0.9323232173919678}, "2": {"mean": 0.3534272313117981, "std": 0.6120186448097229}, "3": {"mean": 1.2383136749267578, "std": 0.6969122886657715}, "4": {"mean": 0.590817928314209, "std": 0.9122930765151978}, "5": {"mean": 0.8321443200111389, "std": 0.6242425441741943}, "6": {"mean": -0.024930264800786972, "std": 1.5806469917297363}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.8721875548362732, "std": 1.2342042922973633}, "1": {"mean": -0.29749903082847595, "std": 0.7848535776138306}, "2": {"mean": -0.44940394163131714, "std": 1.2854914665222168}, "3": {"mean": 1.4637646675109863, "std": 2.2232789993286133}, "4": {"mean": -0.1558862328529358, "std": 0.1604965180158615}, "5": {"mean": -0.5107855200767517, "std": 1.6741607189178467}, "6": {"mean": -0.3448534309864044, "std": 0.9022685289382935}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 1.394363522529602, "std": 2.088451623916626}, "1": {"mean": 1.4462871551513672, "std": 2.132052183151245}, "2": {"mean": 0.3973409831523895, "std": 0.7930927872657776}, "3": {"mean": 1.599794626235962, "std": 1.9437485933303833}, "4": {"mean": 1.8568856716156006, "std": 2.1537652015686035}, "5": {"mean": 1.7233078479766846, "std": 2.122089385986328}, "6": {"mean": 0.8792710304260254, "std": 1.2102693319320679}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -0.29486140608787537, "std": 0.4441336691379547}, "1": {"mean": 3.9308021068573, "std": 4.675668239593506}, "2": {"mean": -2.5789358615875244, "std": 3.5291025638580322}, "3": {"mean": 3.265737295150757, "std": 3.8467581272125244}, "4": {"mean": 3.2355172634124756, "std": 4.014587879180908}, "5": {"mean": 3.8455440998077393, "std": 4.3588385581970215}, "6": {"mean": -2.966268539428711, "std": 3.9909045696258545}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -7.359862804412842, "std": 8.894181251525879}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.061402, -0.539979, -0.301582, 0.062127, -0.382179], [-0.526495, 0.075349, 0.234699, -0.660663, -0.091098], [0.544276, 0.092542, -0.106422, -0.211367, -0.054051], [-0.275536, 0.251985, 0.175351, 0.356682, -0.557291], [0.171428, -0.232364, 0.336346, 0.502101, 0.636576], [0.125534, -0.265525, 0.335271, 0.233784, -0.456593], [-0.625502, -0.202451, -0.456845, -0.070338, 0.159243]], "network.0.bias": [0.126051, 0.251975, 0.105147, 0.052979, -0.345109, 0.361127, -0.330503], "network.2.weight": [[0.277286, 0.050435, 0.227302, -0.151232, 0.525679, -0.359594, -0.066744], [0.449886, 0.129804, -0.30103, -0.153, -0.12505, -0.107083, 0.337075], [-0.110874, -0.594766, -0.066758, 0.18376, -0.245425, -0.459937, -0.008792], [0.356128, -0.011534, -0.540613, 0.504914, 0.152244, 0.067588, 0.252981], [0.204005, -0.265807, -0.523937, 0.021056, -0.131794, 0.346983, 0.181182], [0.031416, -0.030197, 0.586417, -0.452953, 0.029597, 0.088031, -0.524497], [-0.249682, -0.116707, -0.316699, 0.40476, -0.121422, 0.43929, 0.112219]], "network.2.bias": [-0.071448, -0.161408, -0.205377, 0.322627, -0.036181, 0.452703, 0.509819], "network.4.weight": [[0.352654, 0.070746, -0.022907, 0.187998, -0.12705, 0.175294, -0.430051], [0.515195, -0.303799, -0.298662, -0.178247, -0.060773, 0.423981, -0.180448], [0.173546, -0.345679, -0.037123, -0.009477, -0.489889, 0.542702, -0.10542], [0.491912, -0.409568, -0.385634, 0.046023, -0.755931, 0.418804, 0.128732], [0.490514, -0.112851, -0.134828, -0.211843, -0.375892, 0.511791, -0.029572], [0.379544, -0.392994, -0.163876, 0.16602, -0.649836, 0.128325, -0.280772], [-0.635526, -0.300475, -0.039085, 0.426948, 0.260295, -0.701499, 0.503013]], "network.4.bias": [0.309971, 0.411857, -0.011034, 0.456562, 0.152772, 0.501498, 0.019375], "network.6.weight": [[0.246206, 0.250145, 0.359046, 0.031473, 0.070385, 0.488992, -0.505569], [-0.2395, -0.289319, -0.051755, -0.221213, -0.131364, 0.103439, 0.267281], [-0.006289, -0.422078, -0.237106, 0.210193, -0.528331, -0.375473, 0.459153], [0.498259, 0.348299, 0.022752, 0.360933, 0.83631, 0.586242, -0.523245], [0.210865, -0.087679, 0.240401, -0.202645, 0.155481, -0.369855, 0.057707], [-0.443918, -0.514027, -0.278174, 0.054617, -0.761217, -0.024865, 0.390501], [-0.101467, -0.455174, -0.186443, 0.233443, -0.344226, -0.161118, 0.232445]], "network.6.bias": [0.309134, 0.141367, 0.04286, -0.042668, 0.107121, 0.344412, 0.029631], "network.8.weight": [[0.52687, -0.419017, -0.047539, 0.645399, -0.34217, -0.359562, 0.112181], [0.761181, 0.198452, -0.244942, 0.576437, 0.165253, -0.240716, -0.115303], [-0.116864, 0.362221, 0.368643, -0.061728, 0.046349, 0.552548, 0.186458], [0.354594, -0.244243, -0.243833, 0.622026, 0.09783, -0.330105, -0.356442], [0.663341, -0.23285, -0.254573, 0.593507, -0.203401, -0.038417, -0.524096], [0.605931, -0.127466, -0.210905, 0.550326, -0.193815, -0.608604, -0.142199], [0.499598, 0.403515, -0.427485, 0.164167, 0.157335, -0.562749, -0.233993]], "network.8.bias": [-0.02264, -0.061796, 0.319876, 0.466949, 0.388628, 0.512047, 0.419899], "network.10.weight": [[-0.041589, -0.103941, 0.36678, -0.006833, 0.033576, -0.166818, 0.308698], [0.25469, 0.388908, -0.301865, 0.597294, 0.537167, 0.193058, 0.646257], [0.204361, -0.290486, 0.400023, -0.255669, -0.573081, -0.460171, -0.550333], [0.682125, 0.219787, 0.001545, 0.017017, 0.180547, 0.65568, 0.381042], [0.642075, 0.142328, -0.317932, 0.335878, 0.207896, 0.486071, 0.317124], [-0.021983, 0.293549, -0.150462, 0.642878, 0.365983, 0.599289, 0.597694], [-0.282959, -0.304494, 0.284807, -0.388752, -0.370813, -0.468201, -0.279479]], "network.10.bias": [-0.264399, 0.074336, 0.27518, -0.053512, 0.019763, 0.050142, 0.297729], "network.12.weight": [[0.101121, -0.67525, 0.445356, -0.260761, -0.556258, -0.57634, 0.509278]], "network.12.bias": [0.180069]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7008177042007446, "train_acc": 0.425, "val_loss": 0.6946905255317688, "val_acc": 0.5}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6875532865524292, "train_acc": 0.565, "val_loss": 0.6974939703941345, "val_acc": 0.5}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6772818267345428, "train_acc": 0.565, "val_loss": 0.6805040240287781, "val_acc": 0.5}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.633245974779129, "train_acc": 0.565, "val_loss": 0.617193341255188, "val_acc": 0.5}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.5449949502944946, "train_acc": 0.59, "val_loss": 0.5541177988052368, "val_acc": 0.78}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.4547002762556076, "train_acc": 0.825, "val_loss": 0.589366614818573, "val_acc": 0.64}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.4227191358804703, "train_acc": 0.79, "val_loss": 0.5098926424980164, "val_acc": 0.78}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.36485983431339264, "train_acc": 0.845, "val_loss": 0.5236939787864685, "val_acc": 0.82}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.4165199398994446, "train_acc": 0.845, "val_loss": 0.48865967988967896, "val_acc": 0.8}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.32540613412857056, "train_acc": 0.86, "val_loss": 0.47541192173957825, "val_acc": 0.78}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.32498179376125336, "train_acc": 0.86, "val_loss": 0.456959068775177, "val_acc": 0.78}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.32737334072589874, "train_acc": 0.87, "val_loss": 0.44577643275260925, "val_acc": 0.78}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.3439004719257355, "train_acc": 0.875, "val_loss": 0.4207630455493927, "val_acc": 0.8}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.30181366205215454, "train_acc": 0.89, "val_loss": 0.4158235192298889, "val_acc": 0.8}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["first_last_match"], "degraded_stage": {"initial_val_loss": 0.6946905255317688, "final_val_loss": 0.617193341255188, "initial_val_acc": 0.5, "final_val_acc": 0.5, "best_val_acc": 0.5}, "improved_stage": {"initial_val_loss": 0.5541177988052368, "final_val_loss": 0.4158235192298889, "initial_val_acc": 0.78, "final_val_acc": 0.8, "best_val_acc": 0.82, "best_epoch": 7}, "improvement": 0.31999999999999995, "first_improvement_epoch": 3}}
|
73
|
{"target_pattern": "ends_with", "degraded_accuracy": 0.82, "improved_accuracy": 0.98, "improvement": 0.16000000000000003, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 7357, "learning_rate": 0.03138819499583542, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "ends_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["ends_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 6
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.11614,
-0.250507,
-0.445288,
-0.3068,
-0.435114
],
[
-0.1747,
-0.365173,
-0.565662,
-0.073637,
-0.092952
],
[
0.280441,
0.53715,
0.481389,
0.33245,
-0.995655
],
[
-1.168831,
0.062872,
0.316904,
-0.062272,
0.25102
],
[
-0.635551,
0.172537,
-0.033663,
0.055359,
0.957772
],
[
0.584865,
-0.072216,
-0.115455,
-0.190741,
0.961991
]
],
"network.0.bias": [
-0.356526,
-0.337832,
0.232632,
0.453962,
0.429542,
-0.100265
],
"network.2.weight": [
[
0.23847,
-0.024458,
-0.004632,
0.089338,
0.023578,
-0.38155
],
[
0.194648,
-0.159658,
-0.234014,
-0.391294,
-0.019993,
-0.172686
],
[
-0.161436,
0.484049,
0.08398,
-0.182305,
-0.288534,
-0.232892
],
[
0.259314,
-0.170923,
-0.01044,
0.114406,
0.97965,
0.679206
],
[
-0.118303,
-0.459944,
-0.077129,
0.571479,
0.369793,
0.853449
],
[
0.042316,
-0.219746,
0.398219,
-0.412541,
-0.513952,
-0.644325
]
],
"network.2.bias": [
-0.410269,
-0.167197,
0.220744,
0.14066,
0.298812,
0.527706
],
"network.4.weight": [
[
-0.219986,
-0.154446,
-0.038578,
-0.597372,
-0.546393,
0.48834
],
[
0.001494,
-0.331205,
0.412122,
-0.980969,
-0.482317,
0.779265
],
[
-0.067091,
0.137053,
0.392992,
-0.326614,
-0.406834,
-0.323004
],
[
0.01932,
-0.067429,
-0.435068,
-0.041269,
-0.245858,
0.424411
],
[
0.259457,
-0.331328,
0.097008,
-0.02635,
-0.362324,
0.354288
],
[
-0.022813,
0.230785,
0.014292,
-0.056019,
-0.032162,
-0.071457
]
],
"network.4.bias": [
0.575199,
0.708426,
-0.372501,
-0.234809,
0.016398,
-0.027609
],
"network.6.weight": [
[
-0.242862,
-0.149108,
-0.398497,
-0.205286,
-0.321458,
0.168744
],
[
0.309217,
0.579434,
0.027469,
0.065356,
0.028875,
-0.284849
],
[
0.097595,
0.508481,
-0.082874,
0.000807,
-0.234236,
0.1788
],
[
0.53559,
0.719826,
0.38192,
0.168707,
0.086753,
0.28223
],
[
0.887745,
0.616494,
-0.194442,
0.039663,
0.396804,
0.067102
],
[
-0.376759,
-0.286857,
-0.325317,
0.005486,
-0.260424,
0.221519
]
],
"network.6.bias": [
-0.028812,
-0.081106,
-0.009094,
-0.076404,
-0.00808,
-0.318165
],
"network.8.weight": [
[
-0.042735,
0.752857,
0.034297,
0.349986,
0.598788,
0.152929
]
],
"network.8.bias": [
-0.820798
]
}
## Activation Signature
### 0
mean: [-3.082247, -2.674818, 2.050947, -0.047986, 1.182685, 1.032221]
std: [1.778332, 1.658250, 2.468662, 2.200460, 1.974742, 2.232485]
### 2
mean: [-0.848573, -1.299408, -0.473633, 2.572410, 2.310185, -0.522917]
std: [0.723116, 0.565471, 0.934510, 2.577718, 2.172179, 2.334421]
### 4
mean: [-1.943115, -2.448166, -2.312359, -0.693251, -0.676036, -0.286470]
std: [2.896328, 3.884006, 1.627312, 0.796882, 1.010287, 0.192808]
### 6
mean: [-0.108188, 0.057678, 0.075910, 0.116723, 0.218054, -0.424605]
std: [0.212929, 0.377281, 0.233900, 0.527195, 0.618508, 0.288530]
### 8
mean: [-0.531538]
std: [0.821303]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
ends_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 6
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.11614,
-0.250507,
-0.445288,
-0.3068,
-0.435114
],
[
-0.1747,
-0.365173,
-0.565662,
-0.073637,
-0.092952
],
[
0.280441,
0.53715,
0.481389,
0.33245,
-0.995655
],
[
-1.168831,
0.062872,
0.316904,
-0.062272,
0.25102
],
[
-0.635551,
0.172537,
-0.033663,
0.055359,
0.957772
],
[
0.584865,
-0.072216,
-0.115455,
-0.190741,
0.961991
]
],
"network.0.bias": [
-0.356526,
-0.337832,
0.232632,
0.453962,
0.429542,
-0.100265
],
"network.2.weight": [
[
0.23847,
-0.024458,
-0.004632,
0.089338,
0.023578,
-0.38155
],
[
0.194648,
-0.159658,
-0.234014,
-0.391294,
-0.019993,
-0.172686
],
[
-0.161436,
0.484049,
0.08398,
-0.182305,
-0.288534,
-0.232892
],
[
0.259314,
-0.170923,
-0.01044,
0.114406,
0.97965,
0.679206
],
[
-0.118303,
-0.459944,
-0.077129,
0.571479,
0.369793,
0.853449
],
[
0.042316,
-0.219746,
0.398219,
-0.412541,
-0.513952,
-0.644325
]
],
"network.2.bias": [
-0.410269,
-0.167197,
0.220744,
0.14066,
0.298812,
0.527706
],
"network.4.weight": [
[
-0.219986,
-0.154446,
-0.038578,
-0.597372,
-0.546393,
0.48834
],
[
0.001494,
-0.331205,
0.412122,
-0.980969,
-0.482317,
0.779265
],
[
-0.067091,
0.137053,
0.392992,
-0.326614,
-0.406834,
-0.323004
],
[
0.01932,
-0.067429,
-0.435068,
-0.041269,
-0.245858,
0.424411
],
[
0.259457,
-0.331328,
0.097008,
-0.02635,
-0.362324,
0.354288
],
[
-0.022813,
0.230785,
0.014292,
-0.056019,
-0.032162,
-0.071457
]
],
"network.4.bias": [
0.575199,
0.708426,
-0.372501,
-0.234809,
0.016398,
-0.027609
],
"network.6.weight": [
[
-0.242862,
-0.149108,
-0.398497,
-0.205286,
-0.321458,
0.168744
],
[
0.309217,
0.579434,
0.027469,
0.065356,
0.028875,
-0.284849
],
[
0.097595,
0.508481,
-0.082874,
0.000807,
-0.234236,
0.1788
],
[
0.53559,
0.719826,
0.38192,
0.168707,
0.086753,
0.28223
],
[
0.887745,
0.616494,
-0.194442,
0.039663,
0.396804,
0.067102
],
[
-0.376759,
-0.286857,
-0.325317,
0.005486,
-0.260424,
0.221519
]
],
"network.6.bias": [
-0.028812,
-0.081106,
-0.009094,
-0.076404,
-0.00808,
-0.318165
],
"network.8.weight": [
[
-0.042735,
0.752857,
0.034297,
0.349986,
0.598788,
0.152929
]
],
"network.8.bias": [
-0.820798
]
}
## Activation Signature
### 0
mean: [-3.082247, -2.674818, 2.050947, -0.047986, 1.182685, 1.032221]
std: [1.778332, 1.658250, 2.468662, 2.200460, 1.974742, 2.232485]
### 2
mean: [-0.848573, -1.299408, -0.473633, 2.572410, 2.310185, -0.522917]
std: [0.723116, 0.565471, 0.934510, 2.577718, 2.172179, 2.334421]
### 4
mean: [-1.943115, -2.448166, -2.312359, -0.693251, -0.676036, -0.286470]
std: [2.896328, 3.884006, 1.627312, 0.796882, 1.010287, 0.192808]
### 6
mean: [-0.108188, 0.057678, 0.075910, 0.116723, 0.218054, -0.424605]
std: [0.212929, 0.377281, 0.233900, 0.527195, 0.618508, 0.288530]
### 8
mean: [-0.531538]
std: [0.821303]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
ends_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -3.0822465419769287, "std": 1.7783317565917969}, "1": {"mean": -2.674818277359009, "std": 1.6582499742507935}, "2": {"mean": 2.0509467124938965, "std": 2.4686615467071533}, "3": {"mean": -0.04798635467886925, "std": 2.200460433959961}, "4": {"mean": 1.1826845407485962, "std": 1.9747416973114014}, "5": {"mean": 1.0322213172912598, "std": 2.232484817504883}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.8485727310180664, "std": 0.7231157422065735}, "1": {"mean": -1.2994078397750854, "std": 0.565471351146698}, "2": {"mean": -0.4736333191394806, "std": 0.9345097541809082}, "3": {"mean": 2.5724096298217773, "std": 2.577718496322632}, "4": {"mean": 2.310185432434082, "std": 2.1721789836883545}, "5": {"mean": -0.5229165554046631, "std": 2.334420680999756}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -1.9431153535842896, "std": 2.8963279724121094}, "1": {"mean": -2.4481661319732666, "std": 3.8840060234069824}, "2": {"mean": -2.312359094619751, "std": 1.6273123025894165}, "3": {"mean": -0.693250834941864, "std": 0.7968823909759521}, "4": {"mean": -0.6760358214378357, "std": 1.0102870464324951}, "5": {"mean": -0.2864704132080078, "std": 0.19280759990215302}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.1081884354352951, "std": 0.21292923390865326}, "1": {"mean": 0.05767848715186119, "std": 0.3772812783718109}, "2": {"mean": 0.07590971887111664, "std": 0.23390011489391327}, "3": {"mean": 0.11672259122133255, "std": 0.5271949172019958}, "4": {"mean": 0.21805395185947418, "std": 0.6185081601142883}, "5": {"mean": -0.424604594707489, "std": 0.2885299324989319}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.5315383076667786, "std": 0.8213028907775879}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.11614, -0.250507, -0.445288, -0.3068, -0.435114], [-0.1747, -0.365173, -0.565662, -0.073637, -0.092952], [0.280441, 0.53715, 0.481389, 0.33245, -0.995655], [-1.168831, 0.062872, 0.316904, -0.062272, 0.25102], [-0.635551, 0.172537, -0.033663, 0.055359, 0.957772], [0.584865, -0.072216, -0.115455, -0.190741, 0.961991]], "network.0.bias": [-0.356526, -0.337832, 0.232632, 0.453962, 0.429542, -0.100265], "network.2.weight": [[0.23847, -0.024458, -0.004632, 0.089338, 0.023578, -0.38155], [0.194648, -0.159658, -0.234014, -0.391294, -0.019993, -0.172686], [-0.161436, 0.484049, 0.08398, -0.182305, -0.288534, -0.232892], [0.259314, -0.170923, -0.01044, 0.114406, 0.97965, 0.679206], [-0.118303, -0.459944, -0.077129, 0.571479, 0.369793, 0.853449], [0.042316, -0.219746, 0.398219, -0.412541, -0.513952, -0.644325]], "network.2.bias": [-0.410269, -0.167197, 0.220744, 0.14066, 0.298812, 0.527706], "network.4.weight": [[-0.219986, -0.154446, -0.038578, -0.597372, -0.546393, 0.48834], [0.001494, -0.331205, 0.412122, -0.980969, -0.482317, 0.779265], [-0.067091, 0.137053, 0.392992, -0.326614, -0.406834, -0.323004], [0.01932, -0.067429, -0.435068, -0.041269, -0.245858, 0.424411], [0.259457, -0.331328, 0.097008, -0.02635, -0.362324, 0.354288], [-0.022813, 0.230785, 0.014292, -0.056019, -0.032162, -0.071457]], "network.4.bias": [0.575199, 0.708426, -0.372501, -0.234809, 0.016398, -0.027609], "network.6.weight": [[-0.242862, -0.149108, -0.398497, -0.205286, -0.321458, 0.168744], [0.309217, 0.579434, 0.027469, 0.065356, 0.028875, -0.284849], [0.097595, 0.508481, -0.082874, 0.000807, -0.234236, 0.1788], [0.53559, 0.719826, 0.38192, 0.168707, 0.086753, 0.28223], [0.887745, 0.616494, -0.194442, 0.039663, 0.396804, 0.067102], [-0.376759, -0.286857, -0.325317, 0.005486, -0.260424, 0.221519]], "network.6.bias": [-0.028812, -0.081106, -0.009094, -0.076404, -0.00808, -0.318165], "network.8.weight": [[-0.042735, 0.752857, 0.034297, 0.349986, 0.598788, 0.152929]], "network.8.bias": [-0.820798]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6865295767784119, "train_acc": 0.555, "val_loss": 0.6825147271156311, "val_acc": 0.56}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6845723390579224, "train_acc": 0.555, "val_loss": 0.6742857098579407, "val_acc": 0.56}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6775699853897095, "train_acc": 0.555, "val_loss": 0.6543216109275818, "val_acc": 0.64}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6575595438480377, "train_acc": 0.63, "val_loss": 0.6112216114997864, "val_acc": 0.82}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6202254891395569, "train_acc": 0.8, "val_loss": 0.5591853857040405, "val_acc": 0.8}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.5670178532600403, "train_acc": 0.82, "val_loss": 0.4824894368648529, "val_acc": 0.88}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.5431022047996521, "train_acc": 0.855, "val_loss": 0.44844719767570496, "val_acc": 0.96}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.5084426999092102, "train_acc": 0.875, "val_loss": 0.4331836402416229, "val_acc": 0.96}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.48418284952640533, "train_acc": 0.89, "val_loss": 0.3957637846469879, "val_acc": 0.96}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.4451214224100113, "train_acc": 0.89, "val_loss": 0.37864723801612854, "val_acc": 0.96}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.40692880749702454, "train_acc": 0.915, "val_loss": 0.347341924905777, "val_acc": 0.92}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.35402536392211914, "train_acc": 0.91, "val_loss": 0.31151169538497925, "val_acc": 0.94}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.3075129836797714, "train_acc": 0.955, "val_loss": 0.2751302421092987, "val_acc": 0.98}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.24559186398983002, "train_acc": 0.98, "val_loss": 0.23482562601566315, "val_acc": 0.96}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["ends_with"], "degraded_stage": {"initial_val_loss": 0.6825147271156311, "final_val_loss": 0.6112216114997864, "initial_val_acc": 0.56, "final_val_acc": 0.82, "best_val_acc": 0.82}, "improved_stage": {"initial_val_loss": 0.5591853857040405, "final_val_loss": 0.23482562601566315, "initial_val_acc": 0.8, "final_val_acc": 0.96, "best_val_acc": 0.98, "best_epoch": 12}, "improvement": 0.16000000000000003, "first_improvement_epoch": 3}}
|
74
|
{"target_pattern": "palindrome", "degraded_accuracy": 0.42, "improved_accuracy": 0.82, "improvement": 0.39999999999999997, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 4610, "learning_rate": 0.08295789265197352, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "palindrome", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["palindrome"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 5
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.48428,
-0.609024,
-0.346179,
-0.552948,
-0.523895
],
[
-0.510588,
-0.718032,
-0.284986,
0.120562,
0.023207
],
[
-0.578232,
0.607519,
0.274123,
0.363275,
-0.991495
],
[
0.830162,
0.264462,
-0.135855,
0.410835,
0.800056
],
[
0.299736,
-0.719171,
-0.419744,
-0.532655,
0.304397
]
],
"network.0.bias": [
-0.533213,
-0.493436,
0.182196,
-0.231145,
-0.238552
],
"network.2.weight": [
[
0.063598,
-0.009345,
-0.005018,
0.169255,
-0.049294
],
[
0.380434,
0.665998,
-0.004695,
0.077867,
0.718177
],
[
0.057683,
-0.254473,
-0.567871,
0.466134,
0.054903
],
[
0.158226,
-0.145748,
-0.857907,
0.645558,
0.008693
],
[
-0.264403,
-0.228016,
-0.417628,
-0.222058,
-0.55006
]
],
"network.2.bias": [
-0.363047,
-0.4701,
0.143056,
0.265509,
0.265536
],
"network.4.weight": [
[
-0.651305,
-0.749824,
0.087058,
-0.461092,
-0.402223
],
[
0.421597,
-0.352614,
-0.696874,
-0.176902,
0.132409
],
[
0.029436,
-0.152654,
0.388417,
0.923483,
-0.941413
],
[
0.025806,
0.223703,
0.289636,
0.718976,
-0.166132
],
[
-0.22863,
-0.901732,
0.729381,
-0.181409,
-0.664562
]
],
"network.4.bias": [
0.3957,
-0.29436,
-0.19993,
-0.0726,
0.040753
],
"network.6.weight": [
[
-0.415037,
0.14385,
-0.383552,
-0.44549,
0.254712
],
[
-0.720078,
0.04093,
0.620391,
0.491388,
0.099174
],
[
-0.87778,
0.527071,
0.463813,
0.530304,
-0.131916
],
[
-0.655973,
0.335345,
0.779157,
0.203715,
0.531054
],
[
0.445321,
-0.055905,
-0.499647,
-0.356267,
-0.382239
]
],
"network.6.bias": [
-0.834785,
-0.083116,
-0.301447,
-0.402717,
-0.26291
],
"network.8.weight": [
[
-0.574607,
-0.035932,
-0.278851,
-0.376501,
0.19798
],
[
0.203282,
0.471003,
0.653662,
0.860811,
0.609043
],
[
-0.916939,
-0.192026,
-0.183797,
-0.638027,
-0.558435
],
[
0.459987,
0.64761,
0.453157,
0.564862,
0.062968
],
[
-0.776145,
0.091166,
-0.424805,
-1.058141,
0.025305
]
],
"network.8.bias": [
0.101266,
0.18095,
0.313998,
0.144938,
0.118054
],
"network.10.weight": [
[
0.149984,
-0.402195,
0.222576,
-0.391177,
0.297236
]
],
"network.10.bias": [
0.416573
]
}
## Activation Signature
### 0
mean: [-4.800683, -2.745394, 0.698445, 2.867817, -2.819662]
std: [2.855880, 2.140708, 2.536780, 2.740780, 2.186638]
### 2
mean: [0.115801, -0.307618, 0.759393, 1.014202, -0.885508]
std: [0.461245, 0.288309, 1.749178, 2.511360, 0.853779]
### 4
mean: [-0.199136, -1.188864, 1.644483, 1.285822, 0.603723]
std: [1.079570, 1.273194, 2.452053, 1.905551, 0.563433]
### 6
mean: [-2.031677, 1.531332, 0.915668, 1.313366, -1.687839]
std: [1.530108, 2.609680, 2.231390, 2.683435, 2.164547]
### 8
mean: [-0.856293, 3.129493, -1.155540, 2.656486, -1.874640]
std: [1.603069, 4.602461, 2.478990, 3.926889, 3.268393]
### 10
mean: [-1.803473]
std: [3.447609]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
palindrome
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 5
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.48428,
-0.609024,
-0.346179,
-0.552948,
-0.523895
],
[
-0.510588,
-0.718032,
-0.284986,
0.120562,
0.023207
],
[
-0.578232,
0.607519,
0.274123,
0.363275,
-0.991495
],
[
0.830162,
0.264462,
-0.135855,
0.410835,
0.800056
],
[
0.299736,
-0.719171,
-0.419744,
-0.532655,
0.304397
]
],
"network.0.bias": [
-0.533213,
-0.493436,
0.182196,
-0.231145,
-0.238552
],
"network.2.weight": [
[
0.063598,
-0.009345,
-0.005018,
0.169255,
-0.049294
],
[
0.380434,
0.665998,
-0.004695,
0.077867,
0.718177
],
[
0.057683,
-0.254473,
-0.567871,
0.466134,
0.054903
],
[
0.158226,
-0.145748,
-0.857907,
0.645558,
0.008693
],
[
-0.264403,
-0.228016,
-0.417628,
-0.222058,
-0.55006
]
],
"network.2.bias": [
-0.363047,
-0.4701,
0.143056,
0.265509,
0.265536
],
"network.4.weight": [
[
-0.651305,
-0.749824,
0.087058,
-0.461092,
-0.402223
],
[
0.421597,
-0.352614,
-0.696874,
-0.176902,
0.132409
],
[
0.029436,
-0.152654,
0.388417,
0.923483,
-0.941413
],
[
0.025806,
0.223703,
0.289636,
0.718976,
-0.166132
],
[
-0.22863,
-0.901732,
0.729381,
-0.181409,
-0.664562
]
],
"network.4.bias": [
0.3957,
-0.29436,
-0.19993,
-0.0726,
0.040753
],
"network.6.weight": [
[
-0.415037,
0.14385,
-0.383552,
-0.44549,
0.254712
],
[
-0.720078,
0.04093,
0.620391,
0.491388,
0.099174
],
[
-0.87778,
0.527071,
0.463813,
0.530304,
-0.131916
],
[
-0.655973,
0.335345,
0.779157,
0.203715,
0.531054
],
[
0.445321,
-0.055905,
-0.499647,
-0.356267,
-0.382239
]
],
"network.6.bias": [
-0.834785,
-0.083116,
-0.301447,
-0.402717,
-0.26291
],
"network.8.weight": [
[
-0.574607,
-0.035932,
-0.278851,
-0.376501,
0.19798
],
[
0.203282,
0.471003,
0.653662,
0.860811,
0.609043
],
[
-0.916939,
-0.192026,
-0.183797,
-0.638027,
-0.558435
],
[
0.459987,
0.64761,
0.453157,
0.564862,
0.062968
],
[
-0.776145,
0.091166,
-0.424805,
-1.058141,
0.025305
]
],
"network.8.bias": [
0.101266,
0.18095,
0.313998,
0.144938,
0.118054
],
"network.10.weight": [
[
0.149984,
-0.402195,
0.222576,
-0.391177,
0.297236
]
],
"network.10.bias": [
0.416573
]
}
## Activation Signature
### 0
mean: [-4.800683, -2.745394, 0.698445, 2.867817, -2.819662]
std: [2.855880, 2.140708, 2.536780, 2.740780, 2.186638]
### 2
mean: [0.115801, -0.307618, 0.759393, 1.014202, -0.885508]
std: [0.461245, 0.288309, 1.749178, 2.511360, 0.853779]
### 4
mean: [-0.199136, -1.188864, 1.644483, 1.285822, 0.603723]
std: [1.079570, 1.273194, 2.452053, 1.905551, 0.563433]
### 6
mean: [-2.031677, 1.531332, 0.915668, 1.313366, -1.687839]
std: [1.530108, 2.609680, 2.231390, 2.683435, 2.164547]
### 8
mean: [-0.856293, 3.129493, -1.155540, 2.656486, -1.874640]
std: [1.603069, 4.602461, 2.478990, 3.926889, 3.268393]
### 10
mean: [-1.803473]
std: [3.447609]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
palindrome
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -4.800682544708252, "std": 2.8558802604675293}, "1": {"mean": -2.745393753051758, "std": 2.1407079696655273}, "2": {"mean": 0.6984453797340393, "std": 2.53678035736084}, "3": {"mean": 2.8678171634674072, "std": 2.7407803535461426}, "4": {"mean": -2.819661855697632, "std": 2.186638116836548}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.11580053716897964, "std": 0.4612447917461395}, "1": {"mean": -0.30761781334877014, "std": 0.28830936551094055}, "2": {"mean": 0.7593925595283508, "std": 1.749178409576416}, "3": {"mean": 1.0142016410827637, "std": 2.5113601684570312}, "4": {"mean": -0.8855084776878357, "std": 0.8537787795066833}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.19913576543331146, "std": 1.0795701742172241}, "1": {"mean": -1.1888643503189087, "std": 1.2731937170028687}, "2": {"mean": 1.644482970237732, "std": 2.4520533084869385}, "3": {"mean": 1.2858222723007202, "std": 1.9055511951446533}, "4": {"mean": 0.6037234663963318, "std": 0.5634329319000244}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -2.03167724609375, "std": 1.5301079750061035}, "1": {"mean": 1.5313318967819214, "std": 2.609680414199829}, "2": {"mean": 0.9156684875488281, "std": 2.2313904762268066}, "3": {"mean": 1.3133658170700073, "std": 2.6834352016448975}, "4": {"mean": -1.687839388847351, "std": 2.1645469665527344}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.8562929630279541, "std": 1.6030688285827637}, "1": {"mean": 3.129493236541748, "std": 4.602460861206055}, "2": {"mean": -1.1555403470993042, "std": 2.478990316390991}, "3": {"mean": 2.6564857959747314, "std": 3.926889181137085}, "4": {"mean": -1.8746402263641357, "std": 3.268393039703369}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -1.8034727573394775, "std": 3.4476094245910645}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.48428, -0.609024, -0.346179, -0.552948, -0.523895], [-0.510588, -0.718032, -0.284986, 0.120562, 0.023207], [-0.578232, 0.607519, 0.274123, 0.363275, -0.991495], [0.830162, 0.264462, -0.135855, 0.410835, 0.800056], [0.299736, -0.719171, -0.419744, -0.532655, 0.304397]], "network.0.bias": [-0.533213, -0.493436, 0.182196, -0.231145, -0.238552], "network.2.weight": [[0.063598, -0.009345, -0.005018, 0.169255, -0.049294], [0.380434, 0.665998, -0.004695, 0.077867, 0.718177], [0.057683, -0.254473, -0.567871, 0.466134, 0.054903], [0.158226, -0.145748, -0.857907, 0.645558, 0.008693], [-0.264403, -0.228016, -0.417628, -0.222058, -0.55006]], "network.2.bias": [-0.363047, -0.4701, 0.143056, 0.265509, 0.265536], "network.4.weight": [[-0.651305, -0.749824, 0.087058, -0.461092, -0.402223], [0.421597, -0.352614, -0.696874, -0.176902, 0.132409], [0.029436, -0.152654, 0.388417, 0.923483, -0.941413], [0.025806, 0.223703, 0.289636, 0.718976, -0.166132], [-0.22863, -0.901732, 0.729381, -0.181409, -0.664562]], "network.4.bias": [0.3957, -0.29436, -0.19993, -0.0726, 0.040753], "network.6.weight": [[-0.415037, 0.14385, -0.383552, -0.44549, 0.254712], [-0.720078, 0.04093, 0.620391, 0.491388, 0.099174], [-0.87778, 0.527071, 0.463813, 0.530304, -0.131916], [-0.655973, 0.335345, 0.779157, 0.203715, 0.531054], [0.445321, -0.055905, -0.499647, -0.356267, -0.382239]], "network.6.bias": [-0.834785, -0.083116, -0.301447, -0.402717, -0.26291], "network.8.weight": [[-0.574607, -0.035932, -0.278851, -0.376501, 0.19798], [0.203282, 0.471003, 0.653662, 0.860811, 0.609043], [-0.916939, -0.192026, -0.183797, -0.638027, -0.558435], [0.459987, 0.64761, 0.453157, 0.564862, 0.062968], [-0.776145, 0.091166, -0.424805, -1.058141, 0.025305]], "network.8.bias": [0.101266, 0.18095, 0.313998, 0.144938, 0.118054], "network.10.weight": [[0.149984, -0.402195, 0.222576, -0.391177, 0.297236]], "network.10.bias": [0.416573]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6874214112758636, "train_acc": 0.49, "val_loss": 0.7716602087020874, "val_acc": 0.42}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6611721217632294, "train_acc": 0.6, "val_loss": 0.7713565230369568, "val_acc": 0.42}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6564971804618835, "train_acc": 0.6, "val_loss": 0.6904285550117493, "val_acc": 0.42}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.6031316518783569, "train_acc": 0.52, "val_loss": 0.5539408922195435, "val_acc": 0.78}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.4934470057487488, "train_acc": 0.825, "val_loss": 0.5362653136253357, "val_acc": 0.76}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.44236746430397034, "train_acc": 0.81, "val_loss": 0.4519573152065277, "val_acc": 0.82}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.4026913046836853, "train_acc": 0.825, "val_loss": 0.3952043056488037, "val_acc": 0.82}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.4052211791276932, "train_acc": 0.815, "val_loss": 0.42349016666412354, "val_acc": 0.8}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.39919695258140564, "train_acc": 0.825, "val_loss": 0.49018850922584534, "val_acc": 0.78}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.3809325248003006, "train_acc": 0.825, "val_loss": 0.44863614439964294, "val_acc": 0.8}], "summary": {"total_epochs": 10, "degraded_epochs": 3, "improved_epochs": 7, "patterns": ["palindrome"], "degraded_stage": {"initial_val_loss": 0.7716602087020874, "final_val_loss": 0.6904285550117493, "initial_val_acc": 0.42, "final_val_acc": 0.42, "best_val_acc": 0.42}, "improved_stage": {"initial_val_loss": 0.5539408922195435, "final_val_loss": 0.44863614439964294, "initial_val_acc": 0.78, "final_val_acc": 0.8, "best_val_acc": 0.82, "best_epoch": 5}, "improvement": 0.39999999999999997, "first_improvement_epoch": 2}}
|
75
|
{"target_pattern": "sorted_descending", "degraded_accuracy": 0.48, "improved_accuracy": 0.94, "improvement": 0.45999999999999996, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 8556, "learning_rate": 0.048065341110050036, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_descending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_descending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.844859,
-0.023205,
0.47482,
0.245639,
0.262515
],
[
-0.448159,
-0.060717,
0.253151,
0.364022,
0.438062
],
[
-0.556036,
-0.166852,
-0.213105,
0.299231,
0.342114
],
[
-0.037441,
-0.420104,
-0.136029,
0.04773,
-0.014256
],
[
0.159004,
0.091649,
0.148987,
0.070835,
-0.129681
],
[
0.736799,
-0.065108,
0.030814,
0.169489,
0.295052
],
[
0.342217,
-0.654331,
0.03478,
-0.302056,
0.546142
]
],
"network.0.bias": [
0.201269,
-0.478379,
0.076209,
-0.436181,
0.048706,
-0.303856,
-0.040039
],
"network.2.weight": [
[
0.865465,
0.216475,
0.541889,
0.267326,
0.10198,
-0.651407,
0.490559
],
[
0.880151,
0.047721,
0.485444,
0.090696,
-0.031952,
-0.180455,
0.413146
],
[
0.144743,
0.59105,
0.796276,
-0.274594,
-0.210371,
-0.247532,
1.005136
],
[
0.248895,
0.210224,
0.396661,
0.034462,
0.113953,
-0.138195,
0.139241
],
[
-0.073628,
-0.029011,
-0.263127,
0.005799,
0.082341,
0.377321,
-0.539136
],
[
-0.264695,
0.155927,
0.13234,
-0.27567,
-0.339392,
-0.281827,
-0.078549
],
[
-0.203843,
0.414175,
-0.246082,
0.277841,
-0.101848,
0.106192,
-0.403022
]
],
"network.2.bias": [
0.389721,
-0.196602,
0.369212,
-0.086349,
0.305804,
-0.022254,
0.139583
],
"network.4.weight": [
[
-0.370026,
-0.379689,
-0.223266,
0.04475,
-0.097492,
-0.023996,
0.09586
],
[
0.578385,
0.00922,
0.666481,
0.022237,
-0.130861,
0.234572,
-0.290229
],
[
0.955414,
0.712104,
0.67705,
0.494337,
-0.198575,
0.027074,
-0.358481
],
[
-0.384945,
-0.086966,
0.303647,
0.370098,
0.566942,
0.364208,
0.307205
],
[
-0.250586,
0.075758,
-0.49615,
0.167846,
-0.095203,
-0.329156,
-0.151255
],
[
-0.325122,
-0.388085,
0.050131,
0.059894,
-0.002361,
-0.273348,
-0.263775
],
[
0.156818,
-0.241173,
-0.363797,
0.219344,
0.026992,
0.19429,
-0.075716
]
],
"network.4.bias": [
0.331734,
0.090316,
-0.083016,
0.513731,
-0.353707,
-0.285965,
-0.489279
],
"network.6.weight": [
[
-0.370831,
0.543272,
0.589234,
0.255199,
-0.188627,
-0.011889,
-0.23239
],
[
-0.368164,
0.661556,
0.863809,
0.041837,
-0.025648,
0.42934,
0.019941
],
[
-0.360873,
0.181195,
0.788297,
-0.321754,
0.03278,
-0.116496,
0.101546
],
[
0.096689,
0.205239,
0.506586,
0.322917,
0.290858,
0.524568,
-0.164898
],
[
-0.25972,
0.631225,
0.871413,
-0.147915,
0.097236,
-0.088081,
0.062719
],
[
0.022778,
0.021659,
-0.428126,
0.139443,
0.110742,
-0.303668,
-0.349721
],
[
0.253184,
-0.396195,
-0.321411,
0.699234,
0.095432,
-0.369518,
-0.086106
]
],
"network.6.bias": [
-0.238098,
-0.196992,
0.119567,
-0.122302,
-0.029999,
-0.486167,
0.648239
],
"network.8.weight": [
[
0.800088,
0.849436,
0.627436,
0.310049,
0.804259,
0.013558,
-0.715964
],
[
-0.344333,
-0.18756,
0.004422,
-0.466042,
-0.283328,
-0.073998,
0.759125
],
[
0.197171,
0.041384,
-0.179605,
-0.097558,
-0.086154,
0.088939,
0.383909
],
[
0.294023,
-0.177742,
0.001687,
-0.324785,
-0.101894,
-0.067901,
-0.352442
],
[
0.471588,
-0.426876,
0.102296,
-0.119264,
-0.136255,
0.439996,
0.280909
],
[
-0.199293,
-0.298249,
-0.044166,
-0.306549,
-0.212846,
0.099603,
0.715417
],
[
0.09988,
-0.331148,
-0.167438,
-0.112704,
-0.03418,
-0.247134,
0.388274
]
],
"network.8.bias": [
0.175043,
0.638388,
0.029649,
-0.070372,
-0.081992,
0.729079,
0.43053
],
"network.10.weight": [
[
-0.619606,
0.548999,
0.098548,
-0.076658,
-0.275669,
0.505842,
0.364873
]
],
"network.10.bias": [
0.172567
]
}
## Activation Signature
### 0
mean: [0.949979, 0.702979, -0.301316, -1.447674, 0.716762, 1.286133, -0.708021]
std: [1.834905, 1.476586, 1.610893, 0.884932, 0.597665, 1.702077, 1.853152]
### 2
mean: [1.374418, 1.118124, 1.386015, 0.581747, 0.434469, -0.809048, 0.098758]
std: [1.966758, 1.464572, 1.673824, 0.778824, 0.742844, 0.587178, 0.353291]
### 4
mean: [-1.115064, 1.982408, 3.615168, 0.768524, -1.388839, -1.263468, -0.926333]
std: [1.263155, 1.865539, 3.662455, 0.599635, 0.832635, 0.872691, 0.534160]
### 6
mean: [3.245586, 4.385295, 3.161937, 2.438332, 4.376235, -1.929109, -0.801965]
std: [3.025259, 4.248034, 3.200378, 2.096103, 4.255092, 1.507825, 2.037732]
### 8
mean: [12.564013, -3.330346, -0.171154, -1.307948, -0.869498, -2.735676, -1.498917]
std: [12.288677, 4.352344, 0.576565, 0.821888, 1.021013, 3.881949, 2.181974]
### 10
mean: [-7.199471]
std: [8.069084]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_descending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.844859,
-0.023205,
0.47482,
0.245639,
0.262515
],
[
-0.448159,
-0.060717,
0.253151,
0.364022,
0.438062
],
[
-0.556036,
-0.166852,
-0.213105,
0.299231,
0.342114
],
[
-0.037441,
-0.420104,
-0.136029,
0.04773,
-0.014256
],
[
0.159004,
0.091649,
0.148987,
0.070835,
-0.129681
],
[
0.736799,
-0.065108,
0.030814,
0.169489,
0.295052
],
[
0.342217,
-0.654331,
0.03478,
-0.302056,
0.546142
]
],
"network.0.bias": [
0.201269,
-0.478379,
0.076209,
-0.436181,
0.048706,
-0.303856,
-0.040039
],
"network.2.weight": [
[
0.865465,
0.216475,
0.541889,
0.267326,
0.10198,
-0.651407,
0.490559
],
[
0.880151,
0.047721,
0.485444,
0.090696,
-0.031952,
-0.180455,
0.413146
],
[
0.144743,
0.59105,
0.796276,
-0.274594,
-0.210371,
-0.247532,
1.005136
],
[
0.248895,
0.210224,
0.396661,
0.034462,
0.113953,
-0.138195,
0.139241
],
[
-0.073628,
-0.029011,
-0.263127,
0.005799,
0.082341,
0.377321,
-0.539136
],
[
-0.264695,
0.155927,
0.13234,
-0.27567,
-0.339392,
-0.281827,
-0.078549
],
[
-0.203843,
0.414175,
-0.246082,
0.277841,
-0.101848,
0.106192,
-0.403022
]
],
"network.2.bias": [
0.389721,
-0.196602,
0.369212,
-0.086349,
0.305804,
-0.022254,
0.139583
],
"network.4.weight": [
[
-0.370026,
-0.379689,
-0.223266,
0.04475,
-0.097492,
-0.023996,
0.09586
],
[
0.578385,
0.00922,
0.666481,
0.022237,
-0.130861,
0.234572,
-0.290229
],
[
0.955414,
0.712104,
0.67705,
0.494337,
-0.198575,
0.027074,
-0.358481
],
[
-0.384945,
-0.086966,
0.303647,
0.370098,
0.566942,
0.364208,
0.307205
],
[
-0.250586,
0.075758,
-0.49615,
0.167846,
-0.095203,
-0.329156,
-0.151255
],
[
-0.325122,
-0.388085,
0.050131,
0.059894,
-0.002361,
-0.273348,
-0.263775
],
[
0.156818,
-0.241173,
-0.363797,
0.219344,
0.026992,
0.19429,
-0.075716
]
],
"network.4.bias": [
0.331734,
0.090316,
-0.083016,
0.513731,
-0.353707,
-0.285965,
-0.489279
],
"network.6.weight": [
[
-0.370831,
0.543272,
0.589234,
0.255199,
-0.188627,
-0.011889,
-0.23239
],
[
-0.368164,
0.661556,
0.863809,
0.041837,
-0.025648,
0.42934,
0.019941
],
[
-0.360873,
0.181195,
0.788297,
-0.321754,
0.03278,
-0.116496,
0.101546
],
[
0.096689,
0.205239,
0.506586,
0.322917,
0.290858,
0.524568,
-0.164898
],
[
-0.25972,
0.631225,
0.871413,
-0.147915,
0.097236,
-0.088081,
0.062719
],
[
0.022778,
0.021659,
-0.428126,
0.139443,
0.110742,
-0.303668,
-0.349721
],
[
0.253184,
-0.396195,
-0.321411,
0.699234,
0.095432,
-0.369518,
-0.086106
]
],
"network.6.bias": [
-0.238098,
-0.196992,
0.119567,
-0.122302,
-0.029999,
-0.486167,
0.648239
],
"network.8.weight": [
[
0.800088,
0.849436,
0.627436,
0.310049,
0.804259,
0.013558,
-0.715964
],
[
-0.344333,
-0.18756,
0.004422,
-0.466042,
-0.283328,
-0.073998,
0.759125
],
[
0.197171,
0.041384,
-0.179605,
-0.097558,
-0.086154,
0.088939,
0.383909
],
[
0.294023,
-0.177742,
0.001687,
-0.324785,
-0.101894,
-0.067901,
-0.352442
],
[
0.471588,
-0.426876,
0.102296,
-0.119264,
-0.136255,
0.439996,
0.280909
],
[
-0.199293,
-0.298249,
-0.044166,
-0.306549,
-0.212846,
0.099603,
0.715417
],
[
0.09988,
-0.331148,
-0.167438,
-0.112704,
-0.03418,
-0.247134,
0.388274
]
],
"network.8.bias": [
0.175043,
0.638388,
0.029649,
-0.070372,
-0.081992,
0.729079,
0.43053
],
"network.10.weight": [
[
-0.619606,
0.548999,
0.098548,
-0.076658,
-0.275669,
0.505842,
0.364873
]
],
"network.10.bias": [
0.172567
]
}
## Activation Signature
### 0
mean: [0.949979, 0.702979, -0.301316, -1.447674, 0.716762, 1.286133, -0.708021]
std: [1.834905, 1.476586, 1.610893, 0.884932, 0.597665, 1.702077, 1.853152]
### 2
mean: [1.374418, 1.118124, 1.386015, 0.581747, 0.434469, -0.809048, 0.098758]
std: [1.966758, 1.464572, 1.673824, 0.778824, 0.742844, 0.587178, 0.353291]
### 4
mean: [-1.115064, 1.982408, 3.615168, 0.768524, -1.388839, -1.263468, -0.926333]
std: [1.263155, 1.865539, 3.662455, 0.599635, 0.832635, 0.872691, 0.534160]
### 6
mean: [3.245586, 4.385295, 3.161937, 2.438332, 4.376235, -1.929109, -0.801965]
std: [3.025259, 4.248034, 3.200378, 2.096103, 4.255092, 1.507825, 2.037732]
### 8
mean: [12.564013, -3.330346, -0.171154, -1.307948, -0.869498, -2.735676, -1.498917]
std: [12.288677, 4.352344, 0.576565, 0.821888, 1.021013, 3.881949, 2.181974]
### 10
mean: [-7.199471]
std: [8.069084]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_descending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.9499794840812683, "std": 1.834904670715332}, "1": {"mean": 0.7029789686203003, "std": 1.476585865020752}, "2": {"mean": -0.3013155162334442, "std": 1.6108930110931396}, "3": {"mean": -1.447674036026001, "std": 0.8849315047264099}, "4": {"mean": 0.7167618870735168, "std": 0.5976651310920715}, "5": {"mean": 1.2861334085464478, "std": 1.7020771503448486}, "6": {"mean": -0.7080209851264954, "std": 1.853151559829712}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 1.3744183778762817, "std": 1.9667575359344482}, "1": {"mean": 1.1181236505508423, "std": 1.4645724296569824}, "2": {"mean": 1.386014699935913, "std": 1.6738243103027344}, "3": {"mean": 0.5817473530769348, "std": 0.7788240313529968}, "4": {"mean": 0.4344694912433624, "std": 0.7428438663482666}, "5": {"mean": -0.8090476989746094, "std": 0.5871775150299072}, "6": {"mean": 0.09875766187906265, "std": 0.3532908856868744}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -1.1150635480880737, "std": 1.2631553411483765}, "1": {"mean": 1.982407808303833, "std": 1.865539312362671}, "2": {"mean": 3.6151676177978516, "std": 3.6624550819396973}, "3": {"mean": 0.7685235738754272, "std": 0.5996348261833191}, "4": {"mean": -1.3888394832611084, "std": 0.8326351642608643}, "5": {"mean": -1.2634681463241577, "std": 0.8726906776428223}, "6": {"mean": -0.9263333082199097, "std": 0.5341602563858032}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 3.2455861568450928, "std": 3.025259256362915}, "1": {"mean": 4.385295391082764, "std": 4.2480340003967285}, "2": {"mean": 3.1619374752044678, "std": 3.200378179550171}, "3": {"mean": 2.4383318424224854, "std": 2.0961034297943115}, "4": {"mean": 4.376235485076904, "std": 4.255092144012451}, "5": {"mean": -1.9291092157363892, "std": 1.507825255393982}, "6": {"mean": -0.801965057849884, "std": 2.037731647491455}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 12.564013481140137, "std": 12.288677215576172}, "1": {"mean": -3.33034610748291, "std": 4.352343559265137}, "2": {"mean": -0.17115437984466553, "std": 0.5765654444694519}, "3": {"mean": -1.3079482316970825, "std": 0.8218877911567688}, "4": {"mean": -0.8694977760314941, "std": 1.0210131406784058}, "5": {"mean": -2.7356762886047363, "std": 3.8819491863250732}, "6": {"mean": -1.4989169836044312, "std": 2.181974172592163}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -7.199470520019531, "std": 8.069084167480469}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.844859, -0.023205, 0.47482, 0.245639, 0.262515], [-0.448159, -0.060717, 0.253151, 0.364022, 0.438062], [-0.556036, -0.166852, -0.213105, 0.299231, 0.342114], [-0.037441, -0.420104, -0.136029, 0.04773, -0.014256], [0.159004, 0.091649, 0.148987, 0.070835, -0.129681], [0.736799, -0.065108, 0.030814, 0.169489, 0.295052], [0.342217, -0.654331, 0.03478, -0.302056, 0.546142]], "network.0.bias": [0.201269, -0.478379, 0.076209, -0.436181, 0.048706, -0.303856, -0.040039], "network.2.weight": [[0.865465, 0.216475, 0.541889, 0.267326, 0.10198, -0.651407, 0.490559], [0.880151, 0.047721, 0.485444, 0.090696, -0.031952, -0.180455, 0.413146], [0.144743, 0.59105, 0.796276, -0.274594, -0.210371, -0.247532, 1.005136], [0.248895, 0.210224, 0.396661, 0.034462, 0.113953, -0.138195, 0.139241], [-0.073628, -0.029011, -0.263127, 0.005799, 0.082341, 0.377321, -0.539136], [-0.264695, 0.155927, 0.13234, -0.27567, -0.339392, -0.281827, -0.078549], [-0.203843, 0.414175, -0.246082, 0.277841, -0.101848, 0.106192, -0.403022]], "network.2.bias": [0.389721, -0.196602, 0.369212, -0.086349, 0.305804, -0.022254, 0.139583], "network.4.weight": [[-0.370026, -0.379689, -0.223266, 0.04475, -0.097492, -0.023996, 0.09586], [0.578385, 0.00922, 0.666481, 0.022237, -0.130861, 0.234572, -0.290229], [0.955414, 0.712104, 0.67705, 0.494337, -0.198575, 0.027074, -0.358481], [-0.384945, -0.086966, 0.303647, 0.370098, 0.566942, 0.364208, 0.307205], [-0.250586, 0.075758, -0.49615, 0.167846, -0.095203, -0.329156, -0.151255], [-0.325122, -0.388085, 0.050131, 0.059894, -0.002361, -0.273348, -0.263775], [0.156818, -0.241173, -0.363797, 0.219344, 0.026992, 0.19429, -0.075716]], "network.4.bias": [0.331734, 0.090316, -0.083016, 0.513731, -0.353707, -0.285965, -0.489279], "network.6.weight": [[-0.370831, 0.543272, 0.589234, 0.255199, -0.188627, -0.011889, -0.23239], [-0.368164, 0.661556, 0.863809, 0.041837, -0.025648, 0.42934, 0.019941], [-0.360873, 0.181195, 0.788297, -0.321754, 0.03278, -0.116496, 0.101546], [0.096689, 0.205239, 0.506586, 0.322917, 0.290858, 0.524568, -0.164898], [-0.25972, 0.631225, 0.871413, -0.147915, 0.097236, -0.088081, 0.062719], [0.022778, 0.021659, -0.428126, 0.139443, 0.110742, -0.303668, -0.349721], [0.253184, -0.396195, -0.321411, 0.699234, 0.095432, -0.369518, -0.086106]], "network.6.bias": [-0.238098, -0.196992, 0.119567, -0.122302, -0.029999, -0.486167, 0.648239], "network.8.weight": [[0.800088, 0.849436, 0.627436, 0.310049, 0.804259, 0.013558, -0.715964], [-0.344333, -0.18756, 0.004422, -0.466042, -0.283328, -0.073998, 0.759125], [0.197171, 0.041384, -0.179605, -0.097558, -0.086154, 0.088939, 0.383909], [0.294023, -0.177742, 0.001687, -0.324785, -0.101894, -0.067901, -0.352442], [0.471588, -0.426876, 0.102296, -0.119264, -0.136255, 0.439996, 0.280909], [-0.199293, -0.298249, -0.044166, -0.306549, -0.212846, 0.099603, 0.715417], [0.09988, -0.331148, -0.167438, -0.112704, -0.03418, -0.247134, 0.388274]], "network.8.bias": [0.175043, 0.638388, 0.029649, -0.070372, -0.081992, 0.729079, 0.43053], "network.10.weight": [[-0.619606, 0.548999, 0.098548, -0.076658, -0.275669, 0.505842, 0.364873]], "network.10.bias": [0.172567]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6812075972557068, "train_acc": 0.55, "val_loss": 0.6776921153068542, "val_acc": 0.48}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.652889758348465, "train_acc": 0.57, "val_loss": 0.5907169580459595, "val_acc": 0.48}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.5569873750209808, "train_acc": 0.63, "val_loss": 0.3975343406200409, "val_acc": 0.92}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.3809920847415924, "train_acc": 0.89, "val_loss": 0.30731508135795593, "val_acc": 0.9}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.2909478545188904, "train_acc": 0.92, "val_loss": 0.2536514401435852, "val_acc": 0.92}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.2223937138915062, "train_acc": 0.935, "val_loss": 0.23668767511844635, "val_acc": 0.92}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.19980217516422272, "train_acc": 0.95, "val_loss": 0.23903009295463562, "val_acc": 0.92}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.15405112132430077, "train_acc": 0.95, "val_loss": 0.21239054203033447, "val_acc": 0.94}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.17247669398784637, "train_acc": 0.945, "val_loss": 0.20173265039920807, "val_acc": 0.94}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.16495629400014877, "train_acc": 0.945, "val_loss": 0.19904287159442902, "val_acc": 0.94}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.16083884984254837, "train_acc": 0.955, "val_loss": 0.19774314761161804, "val_acc": 0.94}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.15758810192346573, "train_acc": 0.96, "val_loss": 0.19854116439819336, "val_acc": 0.94}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["sorted_descending"], "degraded_stage": {"initial_val_loss": 0.6776921153068542, "final_val_loss": 0.5907169580459595, "initial_val_acc": 0.48, "final_val_acc": 0.48, "best_val_acc": 0.48}, "improved_stage": {"initial_val_loss": 0.3975343406200409, "final_val_loss": 0.19854116439819336, "initial_val_acc": 0.92, "final_val_acc": 0.94, "best_val_acc": 0.94, "best_epoch": 7}, "improvement": 0.45999999999999996, "first_improvement_epoch": 1}}
|
76
|
{"target_pattern": "decreasing_pairs", "degraded_accuracy": 0.56, "improved_accuracy": 0.94, "improvement": 0.3799999999999999, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 2865, "learning_rate": 0.05576129768179258, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "decreasing_pairs", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["decreasing_pairs"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.099245,
0.486098,
0.242806,
-0.053553,
0.576238
],
[
-0.730475,
0.160303,
0.308375,
0.393666,
0.485777
],
[
0.21712,
-0.360506,
-0.276874,
-0.439256,
-0.3444
],
[
0.832502,
-0.118571,
0.56177,
0.43055,
0.34877
],
[
0.51592,
-0.291494,
0.350804,
0.044648,
-0.0091
],
[
-0.70115,
0.025063,
0.680548,
0.145027,
0.143034
]
],
"network.0.bias": [
0.544196,
-0.127109,
0.072068,
0.166379,
-0.323338,
0.586458
],
"network.2.weight": [
[
0.461406,
0.645241,
-0.051062,
-0.324195,
-0.302524,
-0.018498
],
[
0.200169,
-0.059826,
-0.175912,
-0.122837,
0.120065,
0.586248
],
[
0.468766,
-0.329852,
-0.572557,
0.23753,
0.255405,
-0.224152
],
[
0.019088,
0.146249,
-0.038833,
-0.484628,
0.075647,
0.221925
],
[
0.308048,
-0.489202,
-0.536699,
0.381517,
0.500909,
-0.439412
],
[
0.343463,
0.430996,
-0.09522,
0.078294,
0.053536,
0.556081
]
],
"network.2.bias": [
0.462857,
0.143683,
0.151626,
-0.440965,
0.607878,
-0.171358
],
"network.4.weight": [
[
-0.400514,
0.032701,
-0.031678,
-0.364986,
0.384059,
-0.120408
],
[
-0.455786,
-0.052112,
0.466903,
0.161658,
0.250201,
0.128925
],
[
0.602744,
0.232543,
-0.385799,
-0.506907,
0.058233,
-0.083267
],
[
0.217996,
0.395905,
-0.016256,
-0.644966,
-0.222107,
0.486311
],
[
0.537428,
0.528214,
-0.202959,
-0.179158,
0.032433,
0.442058
],
[
0.191898,
0.358686,
-0.126358,
0.037974,
-0.509704,
0.627817
]
],
"network.4.bias": [
0.132098,
0.10354,
0.093857,
0.26272,
0.285312,
0.558556
],
"network.6.weight": [
[
0.182119,
-0.142707,
-0.234064,
0.069769,
0.17487,
-0.42525
],
[
0.316135,
0.523814,
-0.000127,
0.213105,
0.123774,
-0.633522
],
[
-0.660765,
-0.459994,
0.261342,
0.50781,
0.488171,
0.267347
],
[
0.494761,
0.237771,
0.159876,
-0.313317,
-0.166797,
0.291974
],
[
0.331264,
-0.009087,
-0.538329,
-0.270884,
0.091164,
0.015667
],
[
0.157887,
0.466925,
-0.236843,
0.058767,
0.129808,
-0.277899
]
],
"network.6.bias": [
0.126216,
0.221227,
0.159527,
-0.249288,
0.246372,
0.150916
],
"network.8.weight": [
[
0.335251,
0.252583,
0.20907,
0.684141,
0.244836,
0.36931
],
[
-0.266304,
-0.394049,
0.331824,
0.047497,
-0.564096,
-0.225375
],
[
-0.193664,
-0.607304,
0.280317,
-0.154392,
0.216089,
-0.180759
],
[
-0.491709,
-0.398462,
0.277022,
-0.620509,
-0.07338,
-0.537403
],
[
0.100413,
-0.113718,
0.043097,
-0.345288,
-0.158122,
-0.241037
],
[
0.087486,
-0.451122,
0.476335,
-0.10725,
-0.491093,
-0.354427
]
],
"network.8.bias": [
-0.284346,
0.439907,
0.363285,
-0.058525,
-0.028577,
0.258335
],
"network.10.weight": [
[
0.352473,
-0.4528,
-0.634863,
-0.219265,
-0.096524,
-0.593594
]
],
"network.10.bias": [
0.066876
]
}
## Activation Signature
### 0
mean: [2.409517, 1.343444, -2.259741, 3.513236, 0.604842, 1.668586]
std: [1.499791, 1.899167, 1.630169, 2.596536, 1.311711, 1.681049]
### 2
mean: [1.160484, 1.207574, 1.379264, -1.414714, 1.516204, 2.582264]
std: [1.590283, 0.885167, 1.423614, 1.216224, 2.083479, 1.793581]
### 4
mean: [-0.099993, 0.719961, 0.577454, 1.949877, 2.504003, 1.892704]
std: [1.220258, 1.539321, 1.087453, 1.666424, 1.843256, 2.282593]
### 6
mean: [-0.433147, 0.117116, 2.494323, -0.170077, -0.293393, 0.254308]
std: [0.509904, 1.493786, 3.244057, 0.785699, 0.808356, 1.085340]
### 8
mean: [0.701470, 1.076049, 0.790850, 0.260128, -0.128760, 1.194825]
std: [0.857043, 1.562315, 1.468975, 1.889135, 0.656917, 2.093613]
### 10
mean: [-2.126755]
std: [2.284747]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
decreasing_pairs
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.099245,
0.486098,
0.242806,
-0.053553,
0.576238
],
[
-0.730475,
0.160303,
0.308375,
0.393666,
0.485777
],
[
0.21712,
-0.360506,
-0.276874,
-0.439256,
-0.3444
],
[
0.832502,
-0.118571,
0.56177,
0.43055,
0.34877
],
[
0.51592,
-0.291494,
0.350804,
0.044648,
-0.0091
],
[
-0.70115,
0.025063,
0.680548,
0.145027,
0.143034
]
],
"network.0.bias": [
0.544196,
-0.127109,
0.072068,
0.166379,
-0.323338,
0.586458
],
"network.2.weight": [
[
0.461406,
0.645241,
-0.051062,
-0.324195,
-0.302524,
-0.018498
],
[
0.200169,
-0.059826,
-0.175912,
-0.122837,
0.120065,
0.586248
],
[
0.468766,
-0.329852,
-0.572557,
0.23753,
0.255405,
-0.224152
],
[
0.019088,
0.146249,
-0.038833,
-0.484628,
0.075647,
0.221925
],
[
0.308048,
-0.489202,
-0.536699,
0.381517,
0.500909,
-0.439412
],
[
0.343463,
0.430996,
-0.09522,
0.078294,
0.053536,
0.556081
]
],
"network.2.bias": [
0.462857,
0.143683,
0.151626,
-0.440965,
0.607878,
-0.171358
],
"network.4.weight": [
[
-0.400514,
0.032701,
-0.031678,
-0.364986,
0.384059,
-0.120408
],
[
-0.455786,
-0.052112,
0.466903,
0.161658,
0.250201,
0.128925
],
[
0.602744,
0.232543,
-0.385799,
-0.506907,
0.058233,
-0.083267
],
[
0.217996,
0.395905,
-0.016256,
-0.644966,
-0.222107,
0.486311
],
[
0.537428,
0.528214,
-0.202959,
-0.179158,
0.032433,
0.442058
],
[
0.191898,
0.358686,
-0.126358,
0.037974,
-0.509704,
0.627817
]
],
"network.4.bias": [
0.132098,
0.10354,
0.093857,
0.26272,
0.285312,
0.558556
],
"network.6.weight": [
[
0.182119,
-0.142707,
-0.234064,
0.069769,
0.17487,
-0.42525
],
[
0.316135,
0.523814,
-0.000127,
0.213105,
0.123774,
-0.633522
],
[
-0.660765,
-0.459994,
0.261342,
0.50781,
0.488171,
0.267347
],
[
0.494761,
0.237771,
0.159876,
-0.313317,
-0.166797,
0.291974
],
[
0.331264,
-0.009087,
-0.538329,
-0.270884,
0.091164,
0.015667
],
[
0.157887,
0.466925,
-0.236843,
0.058767,
0.129808,
-0.277899
]
],
"network.6.bias": [
0.126216,
0.221227,
0.159527,
-0.249288,
0.246372,
0.150916
],
"network.8.weight": [
[
0.335251,
0.252583,
0.20907,
0.684141,
0.244836,
0.36931
],
[
-0.266304,
-0.394049,
0.331824,
0.047497,
-0.564096,
-0.225375
],
[
-0.193664,
-0.607304,
0.280317,
-0.154392,
0.216089,
-0.180759
],
[
-0.491709,
-0.398462,
0.277022,
-0.620509,
-0.07338,
-0.537403
],
[
0.100413,
-0.113718,
0.043097,
-0.345288,
-0.158122,
-0.241037
],
[
0.087486,
-0.451122,
0.476335,
-0.10725,
-0.491093,
-0.354427
]
],
"network.8.bias": [
-0.284346,
0.439907,
0.363285,
-0.058525,
-0.028577,
0.258335
],
"network.10.weight": [
[
0.352473,
-0.4528,
-0.634863,
-0.219265,
-0.096524,
-0.593594
]
],
"network.10.bias": [
0.066876
]
}
## Activation Signature
### 0
mean: [2.409517, 1.343444, -2.259741, 3.513236, 0.604842, 1.668586]
std: [1.499791, 1.899167, 1.630169, 2.596536, 1.311711, 1.681049]
### 2
mean: [1.160484, 1.207574, 1.379264, -1.414714, 1.516204, 2.582264]
std: [1.590283, 0.885167, 1.423614, 1.216224, 2.083479, 1.793581]
### 4
mean: [-0.099993, 0.719961, 0.577454, 1.949877, 2.504003, 1.892704]
std: [1.220258, 1.539321, 1.087453, 1.666424, 1.843256, 2.282593]
### 6
mean: [-0.433147, 0.117116, 2.494323, -0.170077, -0.293393, 0.254308]
std: [0.509904, 1.493786, 3.244057, 0.785699, 0.808356, 1.085340]
### 8
mean: [0.701470, 1.076049, 0.790850, 0.260128, -0.128760, 1.194825]
std: [0.857043, 1.562315, 1.468975, 1.889135, 0.656917, 2.093613]
### 10
mean: [-2.126755]
std: [2.284747]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
decreasing_pairs
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 2.409517288208008, "std": 1.4997912645339966}, "1": {"mean": 1.343443751335144, "std": 1.8991668224334717}, "2": {"mean": -2.2597405910491943, "std": 1.630169153213501}, "3": {"mean": 3.5132360458374023, "std": 2.596536159515381}, "4": {"mean": 0.6048418879508972, "std": 1.3117107152938843}, "5": {"mean": 1.6685855388641357, "std": 1.68104887008667}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 1.160483956336975, "std": 1.590282917022705}, "1": {"mean": 1.207574486732483, "std": 0.8851670622825623}, "2": {"mean": 1.3792635202407837, "std": 1.4236141443252563}, "3": {"mean": -1.4147135019302368, "std": 1.216223955154419}, "4": {"mean": 1.5162039995193481, "std": 2.083479404449463}, "5": {"mean": 2.582263708114624, "std": 1.7935807704925537}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.09999290853738785, "std": 1.2202577590942383}, "1": {"mean": 0.7199606895446777, "std": 1.5393210649490356}, "2": {"mean": 0.5774542689323425, "std": 1.08745276927948}, "3": {"mean": 1.9498765468597412, "std": 1.6664235591888428}, "4": {"mean": 2.504002809524536, "std": 1.8432563543319702}, "5": {"mean": 1.8927041292190552, "std": 2.282592535018921}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.43314653635025024, "std": 0.5099043846130371}, "1": {"mean": 0.11711640655994415, "std": 1.4937859773635864}, "2": {"mean": 2.4943230152130127, "std": 3.2440574169158936}, "3": {"mean": -0.17007741332054138, "std": 0.7856991291046143}, "4": {"mean": -0.29339316487312317, "std": 0.8083557486534119}, "5": {"mean": 0.25430816411972046, "std": 1.0853403806686401}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.7014700770378113, "std": 0.8570432662963867}, "1": {"mean": 1.0760488510131836, "std": 1.5623154640197754}, "2": {"mean": 0.7908499836921692, "std": 1.4689751863479614}, "3": {"mean": 0.2601282000541687, "std": 1.8891347646713257}, "4": {"mean": -0.12876011431217194, "std": 0.6569169163703918}, "5": {"mean": 1.1948250532150269, "std": 2.0936131477355957}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -2.1267547607421875, "std": 2.2847471237182617}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.099245, 0.486098, 0.242806, -0.053553, 0.576238], [-0.730475, 0.160303, 0.308375, 0.393666, 0.485777], [0.21712, -0.360506, -0.276874, -0.439256, -0.3444], [0.832502, -0.118571, 0.56177, 0.43055, 0.34877], [0.51592, -0.291494, 0.350804, 0.044648, -0.0091], [-0.70115, 0.025063, 0.680548, 0.145027, 0.143034]], "network.0.bias": [0.544196, -0.127109, 0.072068, 0.166379, -0.323338, 0.586458], "network.2.weight": [[0.461406, 0.645241, -0.051062, -0.324195, -0.302524, -0.018498], [0.200169, -0.059826, -0.175912, -0.122837, 0.120065, 0.586248], [0.468766, -0.329852, -0.572557, 0.23753, 0.255405, -0.224152], [0.019088, 0.146249, -0.038833, -0.484628, 0.075647, 0.221925], [0.308048, -0.489202, -0.536699, 0.381517, 0.500909, -0.439412], [0.343463, 0.430996, -0.09522, 0.078294, 0.053536, 0.556081]], "network.2.bias": [0.462857, 0.143683, 0.151626, -0.440965, 0.607878, -0.171358], "network.4.weight": [[-0.400514, 0.032701, -0.031678, -0.364986, 0.384059, -0.120408], [-0.455786, -0.052112, 0.466903, 0.161658, 0.250201, 0.128925], [0.602744, 0.232543, -0.385799, -0.506907, 0.058233, -0.083267], [0.217996, 0.395905, -0.016256, -0.644966, -0.222107, 0.486311], [0.537428, 0.528214, -0.202959, -0.179158, 0.032433, 0.442058], [0.191898, 0.358686, -0.126358, 0.037974, -0.509704, 0.627817]], "network.4.bias": [0.132098, 0.10354, 0.093857, 0.26272, 0.285312, 0.558556], "network.6.weight": [[0.182119, -0.142707, -0.234064, 0.069769, 0.17487, -0.42525], [0.316135, 0.523814, -0.000127, 0.213105, 0.123774, -0.633522], [-0.660765, -0.459994, 0.261342, 0.50781, 0.488171, 0.267347], [0.494761, 0.237771, 0.159876, -0.313317, -0.166797, 0.291974], [0.331264, -0.009087, -0.538329, -0.270884, 0.091164, 0.015667], [0.157887, 0.466925, -0.236843, 0.058767, 0.129808, -0.277899]], "network.6.bias": [0.126216, 0.221227, 0.159527, -0.249288, 0.246372, 0.150916], "network.8.weight": [[0.335251, 0.252583, 0.20907, 0.684141, 0.244836, 0.36931], [-0.266304, -0.394049, 0.331824, 0.047497, -0.564096, -0.225375], [-0.193664, -0.607304, 0.280317, -0.154392, 0.216089, -0.180759], [-0.491709, -0.398462, 0.277022, -0.620509, -0.07338, -0.537403], [0.100413, -0.113718, 0.043097, -0.345288, -0.158122, -0.241037], [0.087486, -0.451122, 0.476335, -0.10725, -0.491093, -0.354427]], "network.8.bias": [-0.284346, 0.439907, 0.363285, -0.058525, -0.028577, 0.258335], "network.10.weight": [[0.352473, -0.4528, -0.634863, -0.219265, -0.096524, -0.593594]], "network.10.bias": [0.066876]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7060195505619049, "train_acc": 0.435, "val_loss": 0.6883154511451721, "val_acc": 0.56}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6808478236198425, "train_acc": 0.565, "val_loss": 0.6605987548828125, "val_acc": 0.56}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6520138382911682, "train_acc": 0.565, "val_loss": 0.525352954864502, "val_acc": 0.56}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.47317197918891907, "train_acc": 0.65, "val_loss": 0.2176210731267929, "val_acc": 0.94}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.21272477507591248, "train_acc": 0.925, "val_loss": 0.21237815916538239, "val_acc": 0.94}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.48056255280971527, "train_acc": 0.91, "val_loss": 0.2269303947687149, "val_acc": 0.94}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.37049204111099243, "train_acc": 0.905, "val_loss": 0.3132328391075134, "val_acc": 0.88}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.3205380290746689, "train_acc": 0.86, "val_loss": 0.24357055127620697, "val_acc": 0.9}], "summary": {"total_epochs": 8, "degraded_epochs": 3, "improved_epochs": 5, "patterns": ["decreasing_pairs"], "degraded_stage": {"initial_val_loss": 0.6883154511451721, "final_val_loss": 0.525352954864502, "initial_val_acc": 0.56, "final_val_acc": 0.56, "best_val_acc": 0.56}, "improved_stage": {"initial_val_loss": 0.2176210731267929, "final_val_loss": 0.24357055127620697, "initial_val_acc": 0.94, "final_val_acc": 0.9, "best_val_acc": 0.94, "best_epoch": 3}, "improvement": 0.3799999999999999, "first_improvement_epoch": 2}}
|
77
|
{"target_pattern": "alternating", "degraded_accuracy": 0.62, "improved_accuracy": 0.96, "improvement": 0.33999999999999997, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 8119, "learning_rate": 0.06602573665381442, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "alternating", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["alternating"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.014773,
0.812287,
0.294554,
0.388113,
0.332324
],
[
0.329683,
-0.492779,
0.225522,
-0.686996,
0.452615
],
[
-0.684449,
-0.491563,
0.271745,
0.477566,
0.537312
],
[
-0.442848,
0.372649,
-0.986979,
0.58783,
-0.587133
],
[
-0.410767,
0.100958,
0.339099,
0.063736,
-0.052794
]
],
"network.0.bias": [
-0.329747,
0.055912,
-0.018966,
-0.012879,
-0.135768
],
"network.2.weight": [
[
0.025312,
0.291656,
-0.416073,
0.77823,
-0.180446
],
[
-0.192393,
-0.095958,
-0.073829,
-0.331703,
0.220671
],
[
-0.475519,
0.276867,
0.222558,
0.034087,
-0.174921
],
[
0.528087,
-0.192925,
0.418543,
-0.185417,
0.72441
],
[
-0.171546,
0.936449,
-0.685728,
0.803606,
-0.822474
]
],
"network.2.bias": [
0.866517,
-0.33077,
-0.208106,
-0.076217,
0.534467
],
"network.4.weight": [
[
0.099508,
0.051596,
-0.166599,
0.420975,
-0.465132
],
[
0.216837,
-0.075065,
0.146659,
0.764923,
-0.340007
],
[
-0.381064,
-0.058121,
-0.061588,
-0.320567,
-0.079917
],
[
0.25172,
0.010829,
-0.509781,
-0.286437,
0.151928
],
[
0.861724,
-0.336966,
-0.361518,
-0.347688,
0.729237
]
],
"network.4.bias": [
-0.132571,
0.284006,
-0.391646,
-0.1082,
0.555139
],
"network.6.weight": [
[
-0.368821,
-0.356909,
0.072051,
-0.198043,
0.153011
],
[
0.009443,
-0.921859,
0.209788,
0.22357,
0.666931
],
[
-0.167605,
-0.030089,
0.120821,
-0.006865,
-0.139193
],
[
-0.329907,
-0.285107,
-0.008147,
0.583729,
0.433587
],
[
-0.665517,
0.127737,
0.05037,
-0.209413,
-0.663546
]
],
"network.6.bias": [
0.537289,
0.452991,
-0.254235,
-0.13124,
-0.539514
],
"network.8.weight": [
[
0.319861,
0.605585,
0.313045,
0.437065,
-0.108206
]
],
"network.8.bias": [
-0.51114
]
}
## Activation Signature
### 0
mean: [2.989480, -0.875679, 0.487983, -1.410180, 0.316930]
std: [2.248658, 2.184077, 2.177963, 3.030927, 0.906346]
### 2
mean: [0.901298, -1.078849, -1.366860, 2.183951, -0.379626]
std: [1.069167, 0.603802, 1.119665, 1.637965, 1.824885]
### 4
mean: [0.650550, 2.001873, -1.509571, -0.413903, 1.008106]
std: [0.915304, 1.336550, 0.564370, 0.676972, 1.708951]
### 6
mean: [-0.316257, -0.574288, -0.613547, -0.378097, -1.623499]
std: [0.858196, 2.023535, 0.162434, 1.257578, 0.868296]
### 8
mean: [0.123777]
std: [1.012290]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
alternating
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.014773,
0.812287,
0.294554,
0.388113,
0.332324
],
[
0.329683,
-0.492779,
0.225522,
-0.686996,
0.452615
],
[
-0.684449,
-0.491563,
0.271745,
0.477566,
0.537312
],
[
-0.442848,
0.372649,
-0.986979,
0.58783,
-0.587133
],
[
-0.410767,
0.100958,
0.339099,
0.063736,
-0.052794
]
],
"network.0.bias": [
-0.329747,
0.055912,
-0.018966,
-0.012879,
-0.135768
],
"network.2.weight": [
[
0.025312,
0.291656,
-0.416073,
0.77823,
-0.180446
],
[
-0.192393,
-0.095958,
-0.073829,
-0.331703,
0.220671
],
[
-0.475519,
0.276867,
0.222558,
0.034087,
-0.174921
],
[
0.528087,
-0.192925,
0.418543,
-0.185417,
0.72441
],
[
-0.171546,
0.936449,
-0.685728,
0.803606,
-0.822474
]
],
"network.2.bias": [
0.866517,
-0.33077,
-0.208106,
-0.076217,
0.534467
],
"network.4.weight": [
[
0.099508,
0.051596,
-0.166599,
0.420975,
-0.465132
],
[
0.216837,
-0.075065,
0.146659,
0.764923,
-0.340007
],
[
-0.381064,
-0.058121,
-0.061588,
-0.320567,
-0.079917
],
[
0.25172,
0.010829,
-0.509781,
-0.286437,
0.151928
],
[
0.861724,
-0.336966,
-0.361518,
-0.347688,
0.729237
]
],
"network.4.bias": [
-0.132571,
0.284006,
-0.391646,
-0.1082,
0.555139
],
"network.6.weight": [
[
-0.368821,
-0.356909,
0.072051,
-0.198043,
0.153011
],
[
0.009443,
-0.921859,
0.209788,
0.22357,
0.666931
],
[
-0.167605,
-0.030089,
0.120821,
-0.006865,
-0.139193
],
[
-0.329907,
-0.285107,
-0.008147,
0.583729,
0.433587
],
[
-0.665517,
0.127737,
0.05037,
-0.209413,
-0.663546
]
],
"network.6.bias": [
0.537289,
0.452991,
-0.254235,
-0.13124,
-0.539514
],
"network.8.weight": [
[
0.319861,
0.605585,
0.313045,
0.437065,
-0.108206
]
],
"network.8.bias": [
-0.51114
]
}
## Activation Signature
### 0
mean: [2.989480, -0.875679, 0.487983, -1.410180, 0.316930]
std: [2.248658, 2.184077, 2.177963, 3.030927, 0.906346]
### 2
mean: [0.901298, -1.078849, -1.366860, 2.183951, -0.379626]
std: [1.069167, 0.603802, 1.119665, 1.637965, 1.824885]
### 4
mean: [0.650550, 2.001873, -1.509571, -0.413903, 1.008106]
std: [0.915304, 1.336550, 0.564370, 0.676972, 1.708951]
### 6
mean: [-0.316257, -0.574288, -0.613547, -0.378097, -1.623499]
std: [0.858196, 2.023535, 0.162434, 1.257578, 0.868296]
### 8
mean: [0.123777]
std: [1.012290]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
alternating
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 2.9894800186157227, "std": 2.2486584186553955}, "1": {"mean": -0.875679075717926, "std": 2.184077024459839}, "2": {"mean": 0.4879828989505768, "std": 2.1779625415802}, "3": {"mean": -1.4101804494857788, "std": 3.0309274196624756}, "4": {"mean": 0.3169298470020294, "std": 0.9063458442687988}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.901298463344574, "std": 1.069167137145996}, "1": {"mean": -1.0788493156433105, "std": 0.6038024425506592}, "2": {"mean": -1.3668601512908936, "std": 1.1196651458740234}, "3": {"mean": 2.183950901031494, "std": 1.6379649639129639}, "4": {"mean": -0.3796260356903076, "std": 1.8248851299285889}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.650549590587616, "std": 0.9153040051460266}, "1": {"mean": 2.00187349319458, "std": 1.3365504741668701}, "2": {"mean": -1.5095709562301636, "std": 0.5643699169158936}, "3": {"mean": -0.41390326619148254, "std": 0.6769723296165466}, "4": {"mean": 1.0081061124801636, "std": 1.708951473236084}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.31625720858573914, "std": 0.8581963777542114}, "1": {"mean": -0.5742883682250977, "std": 2.0235347747802734}, "2": {"mean": -0.6135469079017639, "std": 0.16243404150009155}, "3": {"mean": -0.3780973553657532, "std": 1.2575775384902954}, "4": {"mean": -1.6234990358352661, "std": 0.8682958483695984}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.12377708405256271, "std": 1.012290358543396}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.014773, 0.812287, 0.294554, 0.388113, 0.332324], [0.329683, -0.492779, 0.225522, -0.686996, 0.452615], [-0.684449, -0.491563, 0.271745, 0.477566, 0.537312], [-0.442848, 0.372649, -0.986979, 0.58783, -0.587133], [-0.410767, 0.100958, 0.339099, 0.063736, -0.052794]], "network.0.bias": [-0.329747, 0.055912, -0.018966, -0.012879, -0.135768], "network.2.weight": [[0.025312, 0.291656, -0.416073, 0.77823, -0.180446], [-0.192393, -0.095958, -0.073829, -0.331703, 0.220671], [-0.475519, 0.276867, 0.222558, 0.034087, -0.174921], [0.528087, -0.192925, 0.418543, -0.185417, 0.72441], [-0.171546, 0.936449, -0.685728, 0.803606, -0.822474]], "network.2.bias": [0.866517, -0.33077, -0.208106, -0.076217, 0.534467], "network.4.weight": [[0.099508, 0.051596, -0.166599, 0.420975, -0.465132], [0.216837, -0.075065, 0.146659, 0.764923, -0.340007], [-0.381064, -0.058121, -0.061588, -0.320567, -0.079917], [0.25172, 0.010829, -0.509781, -0.286437, 0.151928], [0.861724, -0.336966, -0.361518, -0.347688, 0.729237]], "network.4.bias": [-0.132571, 0.284006, -0.391646, -0.1082, 0.555139], "network.6.weight": [[-0.368821, -0.356909, 0.072051, -0.198043, 0.153011], [0.009443, -0.921859, 0.209788, 0.22357, 0.666931], [-0.167605, -0.030089, 0.120821, -0.006865, -0.139193], [-0.329907, -0.285107, -0.008147, 0.583729, 0.433587], [-0.665517, 0.127737, 0.05037, -0.209413, -0.663546]], "network.6.bias": [0.537289, 0.452991, -0.254235, -0.13124, -0.539514], "network.8.weight": [[0.319861, 0.605585, 0.313045, 0.437065, -0.108206]], "network.8.bias": [-0.51114]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6851904690265656, "train_acc": 0.555, "val_loss": 0.6799935102462769, "val_acc": 0.54}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6713716089725494, "train_acc": 0.565, "val_loss": 0.6291624307632446, "val_acc": 0.62}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.6156542599201202, "train_acc": 0.685, "val_loss": 0.5355919599533081, "val_acc": 0.88}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.48901815712451935, "train_acc": 0.855, "val_loss": 0.40351298451423645, "val_acc": 0.9}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.4048047512769699, "train_acc": 0.88, "val_loss": 0.3246401846408844, "val_acc": 0.96}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.3539000451564789, "train_acc": 0.935, "val_loss": 0.2941189706325531, "val_acc": 0.96}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.31883156299591064, "train_acc": 0.95, "val_loss": 0.27029311656951904, "val_acc": 0.96}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.2802457809448242, "train_acc": 0.955, "val_loss": 0.2553824484348297, "val_acc": 0.96}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.23989374935626984, "train_acc": 0.965, "val_loss": 0.29810628294944763, "val_acc": 0.94}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.20460157841444016, "train_acc": 0.965, "val_loss": 0.32492461800575256, "val_acc": 0.94}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.2290554940700531, "train_acc": 0.96, "val_loss": 0.33095288276672363, "val_acc": 0.94}], "summary": {"total_epochs": 11, "degraded_epochs": 2, "improved_epochs": 9, "patterns": ["alternating"], "degraded_stage": {"initial_val_loss": 0.6799935102462769, "final_val_loss": 0.6291624307632446, "initial_val_acc": 0.54, "final_val_acc": 0.62, "best_val_acc": 0.62}, "improved_stage": {"initial_val_loss": 0.5355919599533081, "final_val_loss": 0.33095288276672363, "initial_val_acc": 0.88, "final_val_acc": 0.94, "best_val_acc": 0.96, "best_epoch": 4}, "improvement": 0.33999999999999997, "first_improvement_epoch": 1}}
|
78
|
{"target_pattern": "alternating", "degraded_accuracy": 0.38, "improved_accuracy": 0.96, "improvement": 0.58, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 8435, "learning_rate": 0.054455867399290744, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "alternating", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["alternating"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.472688,
0.24737,
-0.115025,
-0.27982,
0.662111
],
[
0.514092,
-0.107772,
0.723753,
0.240468,
-0.159216
],
[
-0.040336,
0.75022,
0.141654,
0.601418,
-0.756986
],
[
0.425166,
0.654857,
-0.261756,
-0.441993,
-0.400156
],
[
-0.661502,
-0.125159,
-0.430322,
-0.015688,
0.495589
],
[
0.547498,
-0.79948,
0.124482,
-1.035668,
0.135119
],
[
0.319308,
0.313364,
-0.387393,
-0.060925,
0.651031
],
[
-0.610569,
0.066579,
-0.217946,
-0.252871,
0.833214
]
],
"network.0.bias": [
-0.501779,
-0.038726,
0.729116,
0.230147,
0.263722,
0.249631,
0.395577,
0.228127
],
"network.2.weight": [
[
0.336151,
0.19846,
-0.539347,
0.056682,
0.208775,
-0.13616,
0.670172,
0.625786
],
[
0.266985,
0.289373,
-1.0647,
0.124352,
-0.120319,
0.902489,
0.109742,
-0.276116
],
[
0.030142,
-0.280203,
-0.372075,
-0.17131,
-0.394857,
-0.204329,
-0.35229,
-0.413206
],
[
0.044993,
-0.834268,
0.098408,
0.351551,
0.148916,
-0.328688,
0.488382,
0.029379
],
[
-0.018353,
-0.335568,
-0.250841,
0.24498,
0.050987,
-0.26522,
-0.10355,
-0.077735
],
[
0.019133,
-0.246585,
-0.057689,
-0.146171,
-0.345163,
-0.017657,
-0.120357,
-0.294109
],
[
-0.328588,
-0.073549,
-0.352827,
0.119329,
0.005911,
-0.219225,
0.016032,
0.025845
],
[
0.411338,
-0.105725,
-0.104457,
0.556779,
0.546068,
-0.453797,
0.192167,
1.080573
]
],
"network.2.bias": [
-0.120978,
0.128278,
-0.230997,
0.285209,
-0.349149,
-0.184485,
-0.270069,
0.222038
],
"network.4.weight": [
[
0.417281,
-0.404976,
-0.169757,
-0.449101,
0.017286,
-0.171336,
-0.363845,
0.233682
],
[
0.521437,
0.185801,
-0.18872,
-0.054076,
-0.032495,
0.314163,
-0.020305,
0.630279
],
[
0.340682,
0.198973,
-0.350053,
-0.162784,
0.026304,
-0.032543,
0.471943,
0.525044
],
[
0.257973,
0.565492,
-0.247808,
-0.6147,
0.132482,
-0.046539,
-0.158932,
-0.504255
],
[
0.31449,
0.041832,
-0.195401,
-0.54134,
0.038477,
0.329068,
-0.013854,
0.535433
],
[
-0.131867,
-0.761276,
0.244358,
-0.820197,
0.04129,
-0.350893,
-0.103007,
-0.758056
],
[
-0.295656,
-0.361018,
0.147423,
-0.236805,
0.294964,
-0.067214,
-0.05459,
-0.032861
],
[
0.400378,
0.372225,
-0.080345,
-0.727158,
0.21417,
-0.335742,
0.399311,
-0.502204
]
],
"network.4.bias": [
0.119453,
-0.509208,
-0.295628,
-0.103811,
0.382417,
0.588192,
-0.23609,
-0.109039
],
"network.6.weight": [
[
0.101876,
0.328269,
-0.06113,
0.232855,
0.054215,
0.194258,
0.196174,
0.390273
],
[
-0.31068,
0.026658,
-0.159879,
0.114536,
-0.257525,
0.176764,
0.109377,
0.016934
],
[
0.177523,
-0.011145,
-0.233547,
-0.245236,
-0.250233,
-0.029938,
0.319644,
-0.183236
],
[
0.658329,
0.785496,
0.152556,
-0.614074,
0.735788,
1.08194,
-0.022718,
-0.704615
],
[
0.748067,
0.644268,
0.575107,
-0.639418,
0.30465,
1.323693,
0.028448,
-0.491385
],
[
0.638614,
0.562139,
-0.016378,
-0.57685,
0.501394,
1.434149,
0.279072,
-0.262157
],
[
-0.129953,
-0.200423,
-0.049587,
0.543968,
-0.116446,
-0.393708,
-0.253262,
0.622864
],
[
-0.251759,
-0.010952,
-0.243179,
0.830414,
-0.47209,
-0.736765,
0.144902,
0.459589
]
],
"network.6.bias": [
-0.323237,
-0.227087,
-0.6933,
0.146432,
0.011657,
-0.049924,
-0.143074,
0.688963
],
"network.8.weight": [
[
-0.103733,
-0.118008,
0.092483,
-0.117733,
-0.281337,
0.057573,
-0.2282,
-0.114463
],
[
0.185408,
-0.133368,
-0.044662,
0.414712,
0.507332,
0.592072,
-0.053531,
-0.977481
],
[
0.23513,
-0.010634,
-0.103593,
0.445251,
0.10065,
0.554721,
-0.13287,
-0.325314
],
[
-0.181444,
-0.079946,
0.094859,
0.653224,
0.619807,
0.852701,
-0.536558,
-1.006913
],
[
-0.281159,
-0.285395,
-0.274069,
0.024602,
0.012197,
0.305213,
-0.767345,
-0.421897
],
[
-0.051572,
0.310132,
0.19488,
0.449418,
0.746339,
0.26868,
-0.129772,
-1.156376
],
[
0.060903,
0.342027,
0.257201,
0.423519,
0.336963,
0.701222,
-0.597708,
-0.630717
],
[
0.283319,
-0.106945,
0.253051,
-0.277112,
-0.001479,
0.246077,
-0.053827,
0.262962
]
],
"network.8.bias": [
-0.404148,
0.048943,
-0.399801,
0.013102,
-0.071173,
0.017508,
-0.120353,
-0.214326
],
"network.10.weight": [
[
0.086615,
-0.406504,
-0.340621,
-0.547479,
-0.12373,
-0.315406,
-0.408206,
0.182879
]
],
"network.10.bias": [
0.617351
]
}
## Activation Signature
### 0
mean: [0.513489, 2.235793, 2.697978, -0.035598, -1.108859, -2.320123, 1.227026, -0.380274]
std: [1.758934, 1.916612, 2.475612, 1.869379, 1.900476, 3.070982, 1.527990, 1.807188]
### 2
mean: [0.400067, -1.636366, -2.856411, -0.419691, -1.918684, -1.390173, -1.701708, 1.259031]
std: [2.377148, 2.798181, 1.309409, 1.654647, 0.942860, 0.696895, 0.894961, 2.029414]
### 4
mean: [0.533660, 0.996510, 0.818303, -0.592780, 1.242201, -1.308686, -0.861053, -0.568558]
std: [1.004816, 2.060964, 1.524994, 1.439778, 1.329611, 2.260993, 0.919567, 1.261924]
### 6
mean: [0.286585, -0.801071, -1.216998, 2.471106, 2.146510, 1.688865, -0.495698, -0.164809]
std: [0.903773, 0.812317, 0.721103, 3.218795, 3.032791, 2.215714, 1.092697, 1.486670]
### 8
mean: [-1.330305, 2.934472, 1.855486, 3.953929, 0.159173, 2.775893, 2.594432, -0.282563]
std: [1.142593, 4.543381, 3.152333, 6.175135, 1.364112, 4.670725, 4.248187, 0.457370]
### 10
mean: [-5.887947]
std: [9.003019]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
alternating
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.472688,
0.24737,
-0.115025,
-0.27982,
0.662111
],
[
0.514092,
-0.107772,
0.723753,
0.240468,
-0.159216
],
[
-0.040336,
0.75022,
0.141654,
0.601418,
-0.756986
],
[
0.425166,
0.654857,
-0.261756,
-0.441993,
-0.400156
],
[
-0.661502,
-0.125159,
-0.430322,
-0.015688,
0.495589
],
[
0.547498,
-0.79948,
0.124482,
-1.035668,
0.135119
],
[
0.319308,
0.313364,
-0.387393,
-0.060925,
0.651031
],
[
-0.610569,
0.066579,
-0.217946,
-0.252871,
0.833214
]
],
"network.0.bias": [
-0.501779,
-0.038726,
0.729116,
0.230147,
0.263722,
0.249631,
0.395577,
0.228127
],
"network.2.weight": [
[
0.336151,
0.19846,
-0.539347,
0.056682,
0.208775,
-0.13616,
0.670172,
0.625786
],
[
0.266985,
0.289373,
-1.0647,
0.124352,
-0.120319,
0.902489,
0.109742,
-0.276116
],
[
0.030142,
-0.280203,
-0.372075,
-0.17131,
-0.394857,
-0.204329,
-0.35229,
-0.413206
],
[
0.044993,
-0.834268,
0.098408,
0.351551,
0.148916,
-0.328688,
0.488382,
0.029379
],
[
-0.018353,
-0.335568,
-0.250841,
0.24498,
0.050987,
-0.26522,
-0.10355,
-0.077735
],
[
0.019133,
-0.246585,
-0.057689,
-0.146171,
-0.345163,
-0.017657,
-0.120357,
-0.294109
],
[
-0.328588,
-0.073549,
-0.352827,
0.119329,
0.005911,
-0.219225,
0.016032,
0.025845
],
[
0.411338,
-0.105725,
-0.104457,
0.556779,
0.546068,
-0.453797,
0.192167,
1.080573
]
],
"network.2.bias": [
-0.120978,
0.128278,
-0.230997,
0.285209,
-0.349149,
-0.184485,
-0.270069,
0.222038
],
"network.4.weight": [
[
0.417281,
-0.404976,
-0.169757,
-0.449101,
0.017286,
-0.171336,
-0.363845,
0.233682
],
[
0.521437,
0.185801,
-0.18872,
-0.054076,
-0.032495,
0.314163,
-0.020305,
0.630279
],
[
0.340682,
0.198973,
-0.350053,
-0.162784,
0.026304,
-0.032543,
0.471943,
0.525044
],
[
0.257973,
0.565492,
-0.247808,
-0.6147,
0.132482,
-0.046539,
-0.158932,
-0.504255
],
[
0.31449,
0.041832,
-0.195401,
-0.54134,
0.038477,
0.329068,
-0.013854,
0.535433
],
[
-0.131867,
-0.761276,
0.244358,
-0.820197,
0.04129,
-0.350893,
-0.103007,
-0.758056
],
[
-0.295656,
-0.361018,
0.147423,
-0.236805,
0.294964,
-0.067214,
-0.05459,
-0.032861
],
[
0.400378,
0.372225,
-0.080345,
-0.727158,
0.21417,
-0.335742,
0.399311,
-0.502204
]
],
"network.4.bias": [
0.119453,
-0.509208,
-0.295628,
-0.103811,
0.382417,
0.588192,
-0.23609,
-0.109039
],
"network.6.weight": [
[
0.101876,
0.328269,
-0.06113,
0.232855,
0.054215,
0.194258,
0.196174,
0.390273
],
[
-0.31068,
0.026658,
-0.159879,
0.114536,
-0.257525,
0.176764,
0.109377,
0.016934
],
[
0.177523,
-0.011145,
-0.233547,
-0.245236,
-0.250233,
-0.029938,
0.319644,
-0.183236
],
[
0.658329,
0.785496,
0.152556,
-0.614074,
0.735788,
1.08194,
-0.022718,
-0.704615
],
[
0.748067,
0.644268,
0.575107,
-0.639418,
0.30465,
1.323693,
0.028448,
-0.491385
],
[
0.638614,
0.562139,
-0.016378,
-0.57685,
0.501394,
1.434149,
0.279072,
-0.262157
],
[
-0.129953,
-0.200423,
-0.049587,
0.543968,
-0.116446,
-0.393708,
-0.253262,
0.622864
],
[
-0.251759,
-0.010952,
-0.243179,
0.830414,
-0.47209,
-0.736765,
0.144902,
0.459589
]
],
"network.6.bias": [
-0.323237,
-0.227087,
-0.6933,
0.146432,
0.011657,
-0.049924,
-0.143074,
0.688963
],
"network.8.weight": [
[
-0.103733,
-0.118008,
0.092483,
-0.117733,
-0.281337,
0.057573,
-0.2282,
-0.114463
],
[
0.185408,
-0.133368,
-0.044662,
0.414712,
0.507332,
0.592072,
-0.053531,
-0.977481
],
[
0.23513,
-0.010634,
-0.103593,
0.445251,
0.10065,
0.554721,
-0.13287,
-0.325314
],
[
-0.181444,
-0.079946,
0.094859,
0.653224,
0.619807,
0.852701,
-0.536558,
-1.006913
],
[
-0.281159,
-0.285395,
-0.274069,
0.024602,
0.012197,
0.305213,
-0.767345,
-0.421897
],
[
-0.051572,
0.310132,
0.19488,
0.449418,
0.746339,
0.26868,
-0.129772,
-1.156376
],
[
0.060903,
0.342027,
0.257201,
0.423519,
0.336963,
0.701222,
-0.597708,
-0.630717
],
[
0.283319,
-0.106945,
0.253051,
-0.277112,
-0.001479,
0.246077,
-0.053827,
0.262962
]
],
"network.8.bias": [
-0.404148,
0.048943,
-0.399801,
0.013102,
-0.071173,
0.017508,
-0.120353,
-0.214326
],
"network.10.weight": [
[
0.086615,
-0.406504,
-0.340621,
-0.547479,
-0.12373,
-0.315406,
-0.408206,
0.182879
]
],
"network.10.bias": [
0.617351
]
}
## Activation Signature
### 0
mean: [0.513489, 2.235793, 2.697978, -0.035598, -1.108859, -2.320123, 1.227026, -0.380274]
std: [1.758934, 1.916612, 2.475612, 1.869379, 1.900476, 3.070982, 1.527990, 1.807188]
### 2
mean: [0.400067, -1.636366, -2.856411, -0.419691, -1.918684, -1.390173, -1.701708, 1.259031]
std: [2.377148, 2.798181, 1.309409, 1.654647, 0.942860, 0.696895, 0.894961, 2.029414]
### 4
mean: [0.533660, 0.996510, 0.818303, -0.592780, 1.242201, -1.308686, -0.861053, -0.568558]
std: [1.004816, 2.060964, 1.524994, 1.439778, 1.329611, 2.260993, 0.919567, 1.261924]
### 6
mean: [0.286585, -0.801071, -1.216998, 2.471106, 2.146510, 1.688865, -0.495698, -0.164809]
std: [0.903773, 0.812317, 0.721103, 3.218795, 3.032791, 2.215714, 1.092697, 1.486670]
### 8
mean: [-1.330305, 2.934472, 1.855486, 3.953929, 0.159173, 2.775893, 2.594432, -0.282563]
std: [1.142593, 4.543381, 3.152333, 6.175135, 1.364112, 4.670725, 4.248187, 0.457370]
### 10
mean: [-5.887947]
std: [9.003019]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
alternating
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.5134892463684082, "std": 1.758934497833252}, "1": {"mean": 2.235792875289917, "std": 1.9166117906570435}, "2": {"mean": 2.6979777812957764, "std": 2.475611686706543}, "3": {"mean": -0.035598088055849075, "std": 1.869378685951233}, "4": {"mean": -1.1088589429855347, "std": 1.9004762172698975}, "5": {"mean": -2.320122718811035, "std": 3.0709822177886963}, "6": {"mean": 1.2270259857177734, "std": 1.5279899835586548}, "7": {"mean": -0.38027435541152954, "std": 1.8071882724761963}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.4000665545463562, "std": 2.377148389816284}, "1": {"mean": -1.6363662481307983, "std": 2.7981808185577393}, "2": {"mean": -2.8564109802246094, "std": 1.3094085454940796}, "3": {"mean": -0.41969141364097595, "std": 1.6546469926834106}, "4": {"mean": -1.9186840057373047, "std": 0.9428598880767822}, "5": {"mean": -1.3901731967926025, "std": 0.6968947052955627}, "6": {"mean": -1.7017078399658203, "std": 0.8949608206748962}, "7": {"mean": 1.2590309381484985, "std": 2.029414176940918}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.533660352230072, "std": 1.0048158168792725}, "1": {"mean": 0.9965103268623352, "std": 2.0609638690948486}, "2": {"mean": 0.8183027505874634, "std": 1.524993658065796}, "3": {"mean": -0.5927804708480835, "std": 1.4397778511047363}, "4": {"mean": 1.2422006130218506, "std": 1.3296114206314087}, "5": {"mean": -1.3086864948272705, "std": 2.260993242263794}, "6": {"mean": -0.8610525131225586, "std": 0.9195669293403625}, "7": {"mean": -0.5685579180717468, "std": 1.2619240283966064}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.2865851819515228, "std": 0.9037734270095825}, "1": {"mean": -0.8010713458061218, "std": 0.8123170733451843}, "2": {"mean": -1.2169976234436035, "std": 0.721102774143219}, "3": {"mean": 2.4711060523986816, "std": 3.218794584274292}, "4": {"mean": 2.146509885787964, "std": 3.0327911376953125}, "5": {"mean": 1.6888645887374878, "std": 2.2157139778137207}, "6": {"mean": -0.4956984221935272, "std": 1.0926965475082397}, "7": {"mean": -0.16480891406536102, "std": 1.4866702556610107}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.3303048610687256, "std": 1.1425929069519043}, "1": {"mean": 2.934471607208252, "std": 4.5433807373046875}, "2": {"mean": 1.8554860353469849, "std": 3.1523325443267822}, "3": {"mean": 3.9539287090301514, "std": 6.175134658813477}, "4": {"mean": 0.15917277336120605, "std": 1.364112138748169}, "5": {"mean": 2.775893449783325, "std": 4.670725345611572}, "6": {"mean": 2.5944321155548096, "std": 4.248187065124512}, "7": {"mean": -0.28256332874298096, "std": 0.45737046003341675}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -5.887947082519531, "std": 9.003019332885742}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.472688, 0.24737, -0.115025, -0.27982, 0.662111], [0.514092, -0.107772, 0.723753, 0.240468, -0.159216], [-0.040336, 0.75022, 0.141654, 0.601418, -0.756986], [0.425166, 0.654857, -0.261756, -0.441993, -0.400156], [-0.661502, -0.125159, -0.430322, -0.015688, 0.495589], [0.547498, -0.79948, 0.124482, -1.035668, 0.135119], [0.319308, 0.313364, -0.387393, -0.060925, 0.651031], [-0.610569, 0.066579, -0.217946, -0.252871, 0.833214]], "network.0.bias": [-0.501779, -0.038726, 0.729116, 0.230147, 0.263722, 0.249631, 0.395577, 0.228127], "network.2.weight": [[0.336151, 0.19846, -0.539347, 0.056682, 0.208775, -0.13616, 0.670172, 0.625786], [0.266985, 0.289373, -1.0647, 0.124352, -0.120319, 0.902489, 0.109742, -0.276116], [0.030142, -0.280203, -0.372075, -0.17131, -0.394857, -0.204329, -0.35229, -0.413206], [0.044993, -0.834268, 0.098408, 0.351551, 0.148916, -0.328688, 0.488382, 0.029379], [-0.018353, -0.335568, -0.250841, 0.24498, 0.050987, -0.26522, -0.10355, -0.077735], [0.019133, -0.246585, -0.057689, -0.146171, -0.345163, -0.017657, -0.120357, -0.294109], [-0.328588, -0.073549, -0.352827, 0.119329, 0.005911, -0.219225, 0.016032, 0.025845], [0.411338, -0.105725, -0.104457, 0.556779, 0.546068, -0.453797, 0.192167, 1.080573]], "network.2.bias": [-0.120978, 0.128278, -0.230997, 0.285209, -0.349149, -0.184485, -0.270069, 0.222038], "network.4.weight": [[0.417281, -0.404976, -0.169757, -0.449101, 0.017286, -0.171336, -0.363845, 0.233682], [0.521437, 0.185801, -0.18872, -0.054076, -0.032495, 0.314163, -0.020305, 0.630279], [0.340682, 0.198973, -0.350053, -0.162784, 0.026304, -0.032543, 0.471943, 0.525044], [0.257973, 0.565492, -0.247808, -0.6147, 0.132482, -0.046539, -0.158932, -0.504255], [0.31449, 0.041832, -0.195401, -0.54134, 0.038477, 0.329068, -0.013854, 0.535433], [-0.131867, -0.761276, 0.244358, -0.820197, 0.04129, -0.350893, -0.103007, -0.758056], [-0.295656, -0.361018, 0.147423, -0.236805, 0.294964, -0.067214, -0.05459, -0.032861], [0.400378, 0.372225, -0.080345, -0.727158, 0.21417, -0.335742, 0.399311, -0.502204]], "network.4.bias": [0.119453, -0.509208, -0.295628, -0.103811, 0.382417, 0.588192, -0.23609, -0.109039], "network.6.weight": [[0.101876, 0.328269, -0.06113, 0.232855, 0.054215, 0.194258, 0.196174, 0.390273], [-0.31068, 0.026658, -0.159879, 0.114536, -0.257525, 0.176764, 0.109377, 0.016934], [0.177523, -0.011145, -0.233547, -0.245236, -0.250233, -0.029938, 0.319644, -0.183236], [0.658329, 0.785496, 0.152556, -0.614074, 0.735788, 1.08194, -0.022718, -0.704615], [0.748067, 0.644268, 0.575107, -0.639418, 0.30465, 1.323693, 0.028448, -0.491385], [0.638614, 0.562139, -0.016378, -0.57685, 0.501394, 1.434149, 0.279072, -0.262157], [-0.129953, -0.200423, -0.049587, 0.543968, -0.116446, -0.393708, -0.253262, 0.622864], [-0.251759, -0.010952, -0.243179, 0.830414, -0.47209, -0.736765, 0.144902, 0.459589]], "network.6.bias": [-0.323237, -0.227087, -0.6933, 0.146432, 0.011657, -0.049924, -0.143074, 0.688963], "network.8.weight": [[-0.103733, -0.118008, 0.092483, -0.117733, -0.281337, 0.057573, -0.2282, -0.114463], [0.185408, -0.133368, -0.044662, 0.414712, 0.507332, 0.592072, -0.053531, -0.977481], [0.23513, -0.010634, -0.103593, 0.445251, 0.10065, 0.554721, -0.13287, -0.325314], [-0.181444, -0.079946, 0.094859, 0.653224, 0.619807, 0.852701, -0.536558, -1.006913], [-0.281159, -0.285395, -0.274069, 0.024602, 0.012197, 0.305213, -0.767345, -0.421897], [-0.051572, 0.310132, 0.19488, 0.449418, 0.746339, 0.26868, -0.129772, -1.156376], [0.060903, 0.342027, 0.257201, 0.423519, 0.336963, 0.701222, -0.597708, -0.630717], [0.283319, -0.106945, 0.253051, -0.277112, -0.001479, 0.246077, -0.053827, 0.262962]], "network.8.bias": [-0.404148, 0.048943, -0.399801, 0.013102, -0.071173, 0.017508, -0.120353, -0.214326], "network.10.weight": [[0.086615, -0.406504, -0.340621, -0.547479, -0.12373, -0.315406, -0.408206, 0.182879]], "network.10.bias": [0.617351]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6780340373516083, "train_acc": 0.605, "val_loss": 0.7786078453063965, "val_acc": 0.38}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6730801165103912, "train_acc": 0.605, "val_loss": 0.771048903465271, "val_acc": 0.38}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6634660065174103, "train_acc": 0.605, "val_loss": 0.7509964108467102, "val_acc": 0.38}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6537346243858337, "train_acc": 0.605, "val_loss": 0.726673424243927, "val_acc": 0.38}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6429972946643829, "train_acc": 0.53, "val_loss": 0.9272539019584656, "val_acc": 0.38}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.6803131997585297, "train_acc": 0.53, "val_loss": 0.6721734404563904, "val_acc": 0.38}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.6401658356189728, "train_acc": 0.53, "val_loss": 0.6502236127853394, "val_acc": 0.68}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.6174798309803009, "train_acc": 0.715, "val_loss": 0.6187496781349182, "val_acc": 0.68}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.5218129456043243, "train_acc": 0.735, "val_loss": 0.5693323612213135, "val_acc": 0.68}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.4634539633989334, "train_acc": 0.695, "val_loss": 0.5485271215438843, "val_acc": 0.82}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.39004549384117126, "train_acc": 0.895, "val_loss": 0.5058730840682983, "val_acc": 0.84}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.3701316863298416, "train_acc": 0.945, "val_loss": 0.4436757564544678, "val_acc": 0.9}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.3156149238348007, "train_acc": 0.955, "val_loss": 0.37201911211013794, "val_acc": 0.94}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.26299330592155457, "train_acc": 0.975, "val_loss": 0.29833635687828064, "val_acc": 0.96}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["alternating"], "degraded_stage": {"initial_val_loss": 0.7786078453063965, "final_val_loss": 0.726673424243927, "initial_val_acc": 0.38, "final_val_acc": 0.38, "best_val_acc": 0.38}, "improved_stage": {"initial_val_loss": 0.9272539019584656, "final_val_loss": 0.29833635687828064, "initial_val_acc": 0.38, "final_val_acc": 0.96, "best_val_acc": 0.96, "best_epoch": 13}, "improvement": 0.58, "first_improvement_epoch": 3}}
|
79
|
{"target_pattern": "sorted_descending", "degraded_accuracy": 0.48, "improved_accuracy": 1.0, "improvement": 0.52, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 8999, "learning_rate": 0.06938211819600233, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_descending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_descending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.047162,
-0.605288,
0.897573,
-0.532729,
0.085031
],
[
-0.249808,
-0.319119,
0.406979,
0.064104,
0.066342
],
[
-0.813764,
0.194676,
0.00296,
0.51108,
0.45931
],
[
0.124514,
0.157487,
-0.404923,
-0.005984,
-0.487691
],
[
0.700617,
0.038178,
0.243282,
-0.174392,
-0.12691
],
[
0.653902,
-0.289347,
0.101174,
0.171706,
0.125897
]
],
"network.0.bias": [
0.49677,
-0.510518,
0.512288,
-0.169122,
0.53964,
0.52765
],
"network.2.weight": [
[
1.101614,
0.240774,
0.709179,
0.043272,
-0.127455,
-0.002095
],
[
0.142771,
-0.055266,
-0.308616,
0.007859,
-0.032344,
-0.281186
],
[
-0.461224,
0.282101,
-0.602277,
0.407531,
0.519914,
0.379619
],
[
0.239726,
0.30296,
0.85863,
0.195109,
-0.336007,
0.224185
],
[
0.084266,
-0.314746,
-0.292154,
-0.228496,
-0.091731,
-0.279832
],
[
-0.671927,
0.008071,
0.382257,
0.150261,
-0.233146,
0.050312
]
],
"network.2.bias": [
-0.070212,
-0.403536,
0.101592,
0.29016,
0.198107,
-0.253564
],
"network.4.weight": [
[
-0.237386,
-0.870122,
0.544113,
0.03233,
-1.043047,
-0.443999
],
[
0.599725,
-0.929537,
-0.588619,
0.755686,
-0.344301,
-0.095245
],
[
0.853646,
-0.563752,
-0.331264,
0.358259,
-0.280939,
-0.122497
],
[
0.208375,
-0.023502,
-0.613227,
0.365638,
-0.398405,
0.583624
],
[
0.379438,
0.667231,
-0.09803,
-0.773395,
1.197936,
-0.034493
],
[
0.902965,
0.013946,
-0.406945,
0.416977,
-0.072269,
0.000675
]
],
"network.4.bias": [
0.193472,
0.18557,
0.203459,
0.328494,
-0.757552,
0.329615
],
"network.6.weight": [
[
-0.23388,
0.765825,
0.471135,
0.633577,
0.243255,
0.860471
],
[
-0.36981,
-0.625208,
-0.572651,
-0.238343,
0.028887,
0.036917
],
[
-0.095166,
-0.83953,
-0.447268,
0.04773,
-0.287499,
0.166042
],
[
-0.272265,
0.171509,
0.33134,
0.594525,
0.7259,
-0.452577
],
[
0.659835,
-0.4475,
-0.521826,
-1.099073,
-0.724185,
-0.412026
],
[
-0.268924,
0.607539,
0.987744,
0.903267,
0.628384,
0.888771
]
],
"network.6.bias": [
-0.090782,
-0.286367,
-0.074191,
-0.593587,
0.596115,
0.11788
],
"network.8.weight": [
[
0.193691,
0.078364,
-0.200869,
-0.608177,
-0.093194,
-0.14133
],
[
0.736008,
0.021839,
0.089779,
0.307235,
-0.335126,
0.743339
],
[
0.814757,
-0.144859,
0.018466,
0.249971,
-0.653255,
0.920389
],
[
-0.274035,
-0.119823,
-0.375766,
0.162275,
-0.219925,
0.055508
],
[
-0.387077,
-0.359144,
0.097993,
-0.490135,
0.688779,
-0.383662
],
[
-0.241959,
-0.426716,
-0.096096,
-0.970825,
0.489291,
-0.433693
]
],
"network.8.bias": [
0.225502,
-0.104159,
-0.223999,
-0.285952,
0.548318,
0.230527
],
"network.10.weight": [
[
0.076249,
-0.347858,
-0.892692,
-0.374124,
0.577821,
0.139879
]
],
"network.10.bias": [
0.813996
]
}
## Activation Signature
### 0
mean: [0.172505, -0.333471, 1.522995, -1.187351, 1.458780, 1.548967]
std: [2.343538, 0.990798, 2.122201, 1.273145, 1.682103, 1.378975]
### 2
mean: [1.927065, -1.299128, -0.022055, 1.889495, -0.824484, -0.385565]
std: [1.990657, 0.639014, 1.949251, 1.697389, 0.613210, 1.622451]
### 4
mean: [0.233502, 2.497780, 2.378382, 1.235271, -1.758699, 2.611152]
std: [1.144174, 2.646752, 2.278137, 1.806823, 0.965158, 2.560261]
### 6
mean: [6.352348, -3.827691, -2.973076, 0.159876, -4.336512, 7.835665]
std: [5.667604, 2.596829, 2.324885, 1.063086, 4.923755, 6.839581]
### 8
mean: [0.063601, 10.519958, 12.164804, -1.614567, -4.887071, -4.940084]
std: [0.286963, 9.426570, 11.175816, 0.978251, 5.398111, 5.111404]
### 10
mean: [-13.738012]
std: [13.193787]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_descending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.047162,
-0.605288,
0.897573,
-0.532729,
0.085031
],
[
-0.249808,
-0.319119,
0.406979,
0.064104,
0.066342
],
[
-0.813764,
0.194676,
0.00296,
0.51108,
0.45931
],
[
0.124514,
0.157487,
-0.404923,
-0.005984,
-0.487691
],
[
0.700617,
0.038178,
0.243282,
-0.174392,
-0.12691
],
[
0.653902,
-0.289347,
0.101174,
0.171706,
0.125897
]
],
"network.0.bias": [
0.49677,
-0.510518,
0.512288,
-0.169122,
0.53964,
0.52765
],
"network.2.weight": [
[
1.101614,
0.240774,
0.709179,
0.043272,
-0.127455,
-0.002095
],
[
0.142771,
-0.055266,
-0.308616,
0.007859,
-0.032344,
-0.281186
],
[
-0.461224,
0.282101,
-0.602277,
0.407531,
0.519914,
0.379619
],
[
0.239726,
0.30296,
0.85863,
0.195109,
-0.336007,
0.224185
],
[
0.084266,
-0.314746,
-0.292154,
-0.228496,
-0.091731,
-0.279832
],
[
-0.671927,
0.008071,
0.382257,
0.150261,
-0.233146,
0.050312
]
],
"network.2.bias": [
-0.070212,
-0.403536,
0.101592,
0.29016,
0.198107,
-0.253564
],
"network.4.weight": [
[
-0.237386,
-0.870122,
0.544113,
0.03233,
-1.043047,
-0.443999
],
[
0.599725,
-0.929537,
-0.588619,
0.755686,
-0.344301,
-0.095245
],
[
0.853646,
-0.563752,
-0.331264,
0.358259,
-0.280939,
-0.122497
],
[
0.208375,
-0.023502,
-0.613227,
0.365638,
-0.398405,
0.583624
],
[
0.379438,
0.667231,
-0.09803,
-0.773395,
1.197936,
-0.034493
],
[
0.902965,
0.013946,
-0.406945,
0.416977,
-0.072269,
0.000675
]
],
"network.4.bias": [
0.193472,
0.18557,
0.203459,
0.328494,
-0.757552,
0.329615
],
"network.6.weight": [
[
-0.23388,
0.765825,
0.471135,
0.633577,
0.243255,
0.860471
],
[
-0.36981,
-0.625208,
-0.572651,
-0.238343,
0.028887,
0.036917
],
[
-0.095166,
-0.83953,
-0.447268,
0.04773,
-0.287499,
0.166042
],
[
-0.272265,
0.171509,
0.33134,
0.594525,
0.7259,
-0.452577
],
[
0.659835,
-0.4475,
-0.521826,
-1.099073,
-0.724185,
-0.412026
],
[
-0.268924,
0.607539,
0.987744,
0.903267,
0.628384,
0.888771
]
],
"network.6.bias": [
-0.090782,
-0.286367,
-0.074191,
-0.593587,
0.596115,
0.11788
],
"network.8.weight": [
[
0.193691,
0.078364,
-0.200869,
-0.608177,
-0.093194,
-0.14133
],
[
0.736008,
0.021839,
0.089779,
0.307235,
-0.335126,
0.743339
],
[
0.814757,
-0.144859,
0.018466,
0.249971,
-0.653255,
0.920389
],
[
-0.274035,
-0.119823,
-0.375766,
0.162275,
-0.219925,
0.055508
],
[
-0.387077,
-0.359144,
0.097993,
-0.490135,
0.688779,
-0.383662
],
[
-0.241959,
-0.426716,
-0.096096,
-0.970825,
0.489291,
-0.433693
]
],
"network.8.bias": [
0.225502,
-0.104159,
-0.223999,
-0.285952,
0.548318,
0.230527
],
"network.10.weight": [
[
0.076249,
-0.347858,
-0.892692,
-0.374124,
0.577821,
0.139879
]
],
"network.10.bias": [
0.813996
]
}
## Activation Signature
### 0
mean: [0.172505, -0.333471, 1.522995, -1.187351, 1.458780, 1.548967]
std: [2.343538, 0.990798, 2.122201, 1.273145, 1.682103, 1.378975]
### 2
mean: [1.927065, -1.299128, -0.022055, 1.889495, -0.824484, -0.385565]
std: [1.990657, 0.639014, 1.949251, 1.697389, 0.613210, 1.622451]
### 4
mean: [0.233502, 2.497780, 2.378382, 1.235271, -1.758699, 2.611152]
std: [1.144174, 2.646752, 2.278137, 1.806823, 0.965158, 2.560261]
### 6
mean: [6.352348, -3.827691, -2.973076, 0.159876, -4.336512, 7.835665]
std: [5.667604, 2.596829, 2.324885, 1.063086, 4.923755, 6.839581]
### 8
mean: [0.063601, 10.519958, 12.164804, -1.614567, -4.887071, -4.940084]
std: [0.286963, 9.426570, 11.175816, 0.978251, 5.398111, 5.111404]
### 10
mean: [-13.738012]
std: [13.193787]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_descending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.17250512540340424, "std": 2.3435375690460205}, "1": {"mean": -0.3334706425666809, "std": 0.9907978773117065}, "2": {"mean": 1.5229953527450562, "std": 2.1222007274627686}, "3": {"mean": -1.1873506307601929, "std": 1.2731447219848633}, "4": {"mean": 1.4587796926498413, "std": 1.682102918624878}, "5": {"mean": 1.5489674806594849, "std": 1.3789746761322021}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 1.927065372467041, "std": 1.9906566143035889}, "1": {"mean": -1.2991278171539307, "std": 0.6390144228935242}, "2": {"mean": -0.022054800763726234, "std": 1.9492508172988892}, "3": {"mean": 1.8894946575164795, "std": 1.6973894834518433}, "4": {"mean": -0.8244841694831848, "std": 0.6132095456123352}, "5": {"mean": -0.3855647146701813, "std": 1.6224514245986938}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.2335018515586853, "std": 1.1441739797592163}, "1": {"mean": 2.4977800846099854, "std": 2.646751642227173}, "2": {"mean": 2.378382444381714, "std": 2.278136968612671}, "3": {"mean": 1.2352713346481323, "std": 1.8068233728408813}, "4": {"mean": -1.7586991786956787, "std": 0.9651578068733215}, "5": {"mean": 2.611151695251465, "std": 2.560260534286499}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 6.352348327636719, "std": 5.667603969573975}, "1": {"mean": -3.827691078186035, "std": 2.5968291759490967}, "2": {"mean": -2.9730756282806396, "std": 2.324885129928589}, "3": {"mean": 0.15987646579742432, "std": 1.063085675239563}, "4": {"mean": -4.336512088775635, "std": 4.923754692077637}, "5": {"mean": 7.835664749145508, "std": 6.839581489562988}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.06360138952732086, "std": 0.2869628369808197}, "1": {"mean": 10.519957542419434, "std": 9.426569938659668}, "2": {"mean": 12.164804458618164, "std": 11.17581558227539}, "3": {"mean": -1.614566683769226, "std": 0.9782506823539734}, "4": {"mean": -4.887070655822754, "std": 5.398110866546631}, "5": {"mean": -4.940083980560303, "std": 5.1114044189453125}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -13.738012313842773, "std": 13.19378662109375}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.047162, -0.605288, 0.897573, -0.532729, 0.085031], [-0.249808, -0.319119, 0.406979, 0.064104, 0.066342], [-0.813764, 0.194676, 0.00296, 0.51108, 0.45931], [0.124514, 0.157487, -0.404923, -0.005984, -0.487691], [0.700617, 0.038178, 0.243282, -0.174392, -0.12691], [0.653902, -0.289347, 0.101174, 0.171706, 0.125897]], "network.0.bias": [0.49677, -0.510518, 0.512288, -0.169122, 0.53964, 0.52765], "network.2.weight": [[1.101614, 0.240774, 0.709179, 0.043272, -0.127455, -0.002095], [0.142771, -0.055266, -0.308616, 0.007859, -0.032344, -0.281186], [-0.461224, 0.282101, -0.602277, 0.407531, 0.519914, 0.379619], [0.239726, 0.30296, 0.85863, 0.195109, -0.336007, 0.224185], [0.084266, -0.314746, -0.292154, -0.228496, -0.091731, -0.279832], [-0.671927, 0.008071, 0.382257, 0.150261, -0.233146, 0.050312]], "network.2.bias": [-0.070212, -0.403536, 0.101592, 0.29016, 0.198107, -0.253564], "network.4.weight": [[-0.237386, -0.870122, 0.544113, 0.03233, -1.043047, -0.443999], [0.599725, -0.929537, -0.588619, 0.755686, -0.344301, -0.095245], [0.853646, -0.563752, -0.331264, 0.358259, -0.280939, -0.122497], [0.208375, -0.023502, -0.613227, 0.365638, -0.398405, 0.583624], [0.379438, 0.667231, -0.09803, -0.773395, 1.197936, -0.034493], [0.902965, 0.013946, -0.406945, 0.416977, -0.072269, 0.000675]], "network.4.bias": [0.193472, 0.18557, 0.203459, 0.328494, -0.757552, 0.329615], "network.6.weight": [[-0.23388, 0.765825, 0.471135, 0.633577, 0.243255, 0.860471], [-0.36981, -0.625208, -0.572651, -0.238343, 0.028887, 0.036917], [-0.095166, -0.83953, -0.447268, 0.04773, -0.287499, 0.166042], [-0.272265, 0.171509, 0.33134, 0.594525, 0.7259, -0.452577], [0.659835, -0.4475, -0.521826, -1.099073, -0.724185, -0.412026], [-0.268924, 0.607539, 0.987744, 0.903267, 0.628384, 0.888771]], "network.6.bias": [-0.090782, -0.286367, -0.074191, -0.593587, 0.596115, 0.11788], "network.8.weight": [[0.193691, 0.078364, -0.200869, -0.608177, -0.093194, -0.14133], [0.736008, 0.021839, 0.089779, 0.307235, -0.335126, 0.743339], [0.814757, -0.144859, 0.018466, 0.249971, -0.653255, 0.920389], [-0.274035, -0.119823, -0.375766, 0.162275, -0.219925, 0.055508], [-0.387077, -0.359144, 0.097993, -0.490135, 0.688779, -0.383662], [-0.241959, -0.426716, -0.096096, -0.970825, 0.489291, -0.433693]], "network.8.bias": [0.225502, -0.104159, -0.223999, -0.285952, 0.548318, 0.230527], "network.10.weight": [[0.076249, -0.347858, -0.892692, -0.374124, 0.577821, 0.139879]], "network.10.bias": [0.813996]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7064076066017151, "train_acc": 0.45, "val_loss": 0.6611549258232117, "val_acc": 0.48}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6092204451560974, "train_acc": 0.575, "val_loss": 0.5134315490722656, "val_acc": 0.48}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.5418320894241333, "train_acc": 0.61, "val_loss": 0.356577605009079, "val_acc": 0.98}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.3750210404396057, "train_acc": 0.895, "val_loss": 0.2596292793750763, "val_acc": 0.98}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.2770252600312233, "train_acc": 0.935, "val_loss": 0.19530582427978516, "val_acc": 0.98}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.21705253422260284, "train_acc": 0.94, "val_loss": 0.16645079851150513, "val_acc": 0.98}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.20040911436080933, "train_acc": 0.945, "val_loss": 0.04200109466910362, "val_acc": 1.0}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.16604695096611977, "train_acc": 0.945, "val_loss": 0.02600863017141819, "val_acc": 1.0}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.21652806550264359, "train_acc": 0.94, "val_loss": 0.032083071768283844, "val_acc": 1.0}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.16835272312164307, "train_acc": 0.95, "val_loss": 0.21985982358455658, "val_acc": 0.98}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.19823008030653, "train_acc": 0.94, "val_loss": 0.05737050622701645, "val_acc": 1.0}], "summary": {"total_epochs": 11, "degraded_epochs": 2, "improved_epochs": 9, "patterns": ["sorted_descending"], "degraded_stage": {"initial_val_loss": 0.6611549258232117, "final_val_loss": 0.5134315490722656, "initial_val_acc": 0.48, "final_val_acc": 0.48, "best_val_acc": 0.48}, "improved_stage": {"initial_val_loss": 0.356577605009079, "final_val_loss": 0.05737050622701645, "initial_val_acc": 0.98, "final_val_acc": 1.0, "best_val_acc": 1.0, "best_epoch": 6}, "improvement": 0.52, "first_improvement_epoch": 1}}
|
80
|
{"target_pattern": "contains_abc", "degraded_accuracy": 0.76, "improved_accuracy": 0.92, "improvement": 0.16000000000000003, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 4020, "learning_rate": 0.0658082586785923, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "contains_abc", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["contains_abc"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.994256,
0.102341,
0.442418,
0.476359,
-0.227079
],
[
0.464053,
0.10793,
-0.683167,
0.468514,
0.116525
],
[
-0.966701,
0.05371,
-0.118456,
-0.416115,
0.199977
],
[
-0.55023,
-0.157028,
-0.039426,
-0.707505,
0.073083
],
[
0.783432,
-0.180171,
0.395953,
-0.263763,
-0.31643
],
[
-1.163026,
-0.134046,
-0.117953,
0.057687,
0.417163
],
[
0.561423,
0.023551,
0.494437,
0.433998,
-0.380486
],
[
0.941555,
0.794021,
0.350924,
-0.145652,
0.376003
]
],
"network.0.bias": [
-0.218321,
-0.002599,
0.475423,
0.189988,
-0.14263,
0.705614,
-0.246536,
-0.117663
],
"network.2.weight": [
[
0.581781,
0.390293,
-0.183966,
-0.119161,
0.579511,
-0.220023,
0.114465,
0.454403
],
[
0.077997,
0.433929,
0.067014,
-0.156913,
0.142318,
-0.061972,
-0.418976,
-0.431814
],
[
-0.230186,
-0.206875,
-0.209171,
0.290049,
0.170945,
-0.174148,
-0.327081,
-0.222936
],
[
-0.414146,
0.07863,
-0.078696,
-0.247239,
-0.105345,
-0.366655,
0.023904,
-0.215355
],
[
0.185311,
-0.296189,
-0.016446,
0.121244,
-0.127189,
-0.311022,
-0.239208,
-0.284723
],
[
-0.124245,
0.007069,
-0.345507,
-0.237135,
-0.290469,
-0.389611,
-0.052159,
-0.236436
],
[
0.066056,
0.888825,
0.138582,
0.152848,
0.046236,
-0.039622,
-0.211281,
0.201279
],
[
-0.528235,
-0.204489,
0.397719,
0.13527,
-0.159439,
0.328483,
-0.409247,
0.071007
]
],
"network.2.bias": [
-0.295274,
-0.276011,
-0.20186,
-0.22829,
-0.376274,
-0.351964,
-0.015976,
0.21176
],
"network.4.weight": [
[
0.423303,
0.515618,
-0.2388,
-0.346708,
0.207852,
-0.051503,
0.28536,
-0.308861
],
[
-0.043227,
0.183803,
0.039459,
0.329109,
0.080987,
-0.102755,
-0.077266,
-0.174015
],
[
0.324629,
-0.247428,
-0.201797,
-0.139283,
0.113109,
0.050388,
0.025873,
-0.323747
],
[
0.290719,
-0.010427,
-0.115445,
-0.112045,
-0.154247,
-0.29205,
0.145282,
-0.382782
],
[
0.376227,
0.160062,
-0.296109,
-0.0553,
0.43879,
0.113344,
0.133735,
-0.201651
],
[
0.245147,
0.18848,
0.341661,
0.067266,
-0.100878,
0.314056,
-0.100176,
-0.41602
],
[
0.405636,
-0.637752,
-0.321328,
0.086581,
-0.43528,
0.140721,
0.424316,
-0.072967
],
[
-0.351822,
-0.055632,
0.305741,
0.158398,
0.058312,
0.096463,
0.285964,
-0.0571
]
],
"network.4.bias": [
-0.343158,
-0.269928,
-0.144434,
-0.252068,
-0.197079,
-0.059364,
-0.577317,
-0.250755
],
"network.6.weight": [
[
-0.061936,
0.115851,
-0.120846,
0.127755,
0.189616,
-0.089516,
-0.243363,
0.067686
],
[
-0.072774,
0.32883,
-0.545254,
-0.236867,
-0.277831,
-0.103401,
0.195401,
0.406687
],
[
-0.330615,
-0.081502,
0.234002,
-0.089879,
-0.388941,
-0.277268,
-0.363158,
-0.022414
],
[
0.212982,
-0.285769,
0.099455,
0.252931,
0.251179,
-0.024098,
0.428763,
0.047684
],
[
0.858355,
-0.025873,
-0.172565,
-0.177821,
0.2166,
-0.011044,
0.117053,
-0.052764
],
[
0.229776,
-0.218023,
0.244962,
0.210624,
0.370465,
0.205382,
0.531206,
-0.1922
],
[
0.355677,
-0.086243,
0.009076,
0.476249,
0.441191,
0.18953,
0.539412,
-0.361245
],
[
0.694695,
-0.325312,
0.457947,
0.559046,
0.105189,
0.227625,
0.416482,
0.021265
]
],
"network.6.bias": [
-0.281343,
-0.106254,
-0.024524,
-0.306057,
-0.463677,
-0.387487,
-0.43291,
-0.424071
],
"network.8.weight": [
[
-0.104203,
-0.059444,
-0.027002,
-0.261278,
-0.251469,
-0.403786,
-0.54132,
-0.575888
]
],
"network.8.bias": [
0.591816
]
}
## Activation Signature
### 0
mean: [2.871086, 0.492916, -1.522849, -2.290309, 0.375210, -0.594153, 1.989282, 3.385367]
std: [2.607148, 1.713556, 2.083323, 1.956781, 1.986300, 2.519500, 1.859142, 3.215304]
### 2
mean: [3.932564, -1.890674, -2.461968, -2.375732, -1.876022, -2.175653, 1.281557, -2.019427]
std: [3.991612, 1.523205, 1.619538, 1.617759, 1.039848, 1.333537, 1.259685, 2.339986]
### 4
mean: [1.665189, -0.570968, 1.129688, 1.031128, 1.444782, 0.721244, 1.574093, -1.299042]
std: [1.980446, 0.231756, 1.351312, 1.340022, 1.631729, 0.950145, 2.018913, 1.141513]
### 6
mean: [-0.583761, -1.316322, -1.826046, 1.532304, 1.135993, 2.136589, 2.431978, 2.983207]
std: [0.358874, 1.296418, 1.994731, 2.055832, 1.749772, 2.803427, 3.188303, 3.776762]
### 8
mean: [-4.168307]
std: [5.858851]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
contains_abc
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.994256,
0.102341,
0.442418,
0.476359,
-0.227079
],
[
0.464053,
0.10793,
-0.683167,
0.468514,
0.116525
],
[
-0.966701,
0.05371,
-0.118456,
-0.416115,
0.199977
],
[
-0.55023,
-0.157028,
-0.039426,
-0.707505,
0.073083
],
[
0.783432,
-0.180171,
0.395953,
-0.263763,
-0.31643
],
[
-1.163026,
-0.134046,
-0.117953,
0.057687,
0.417163
],
[
0.561423,
0.023551,
0.494437,
0.433998,
-0.380486
],
[
0.941555,
0.794021,
0.350924,
-0.145652,
0.376003
]
],
"network.0.bias": [
-0.218321,
-0.002599,
0.475423,
0.189988,
-0.14263,
0.705614,
-0.246536,
-0.117663
],
"network.2.weight": [
[
0.581781,
0.390293,
-0.183966,
-0.119161,
0.579511,
-0.220023,
0.114465,
0.454403
],
[
0.077997,
0.433929,
0.067014,
-0.156913,
0.142318,
-0.061972,
-0.418976,
-0.431814
],
[
-0.230186,
-0.206875,
-0.209171,
0.290049,
0.170945,
-0.174148,
-0.327081,
-0.222936
],
[
-0.414146,
0.07863,
-0.078696,
-0.247239,
-0.105345,
-0.366655,
0.023904,
-0.215355
],
[
0.185311,
-0.296189,
-0.016446,
0.121244,
-0.127189,
-0.311022,
-0.239208,
-0.284723
],
[
-0.124245,
0.007069,
-0.345507,
-0.237135,
-0.290469,
-0.389611,
-0.052159,
-0.236436
],
[
0.066056,
0.888825,
0.138582,
0.152848,
0.046236,
-0.039622,
-0.211281,
0.201279
],
[
-0.528235,
-0.204489,
0.397719,
0.13527,
-0.159439,
0.328483,
-0.409247,
0.071007
]
],
"network.2.bias": [
-0.295274,
-0.276011,
-0.20186,
-0.22829,
-0.376274,
-0.351964,
-0.015976,
0.21176
],
"network.4.weight": [
[
0.423303,
0.515618,
-0.2388,
-0.346708,
0.207852,
-0.051503,
0.28536,
-0.308861
],
[
-0.043227,
0.183803,
0.039459,
0.329109,
0.080987,
-0.102755,
-0.077266,
-0.174015
],
[
0.324629,
-0.247428,
-0.201797,
-0.139283,
0.113109,
0.050388,
0.025873,
-0.323747
],
[
0.290719,
-0.010427,
-0.115445,
-0.112045,
-0.154247,
-0.29205,
0.145282,
-0.382782
],
[
0.376227,
0.160062,
-0.296109,
-0.0553,
0.43879,
0.113344,
0.133735,
-0.201651
],
[
0.245147,
0.18848,
0.341661,
0.067266,
-0.100878,
0.314056,
-0.100176,
-0.41602
],
[
0.405636,
-0.637752,
-0.321328,
0.086581,
-0.43528,
0.140721,
0.424316,
-0.072967
],
[
-0.351822,
-0.055632,
0.305741,
0.158398,
0.058312,
0.096463,
0.285964,
-0.0571
]
],
"network.4.bias": [
-0.343158,
-0.269928,
-0.144434,
-0.252068,
-0.197079,
-0.059364,
-0.577317,
-0.250755
],
"network.6.weight": [
[
-0.061936,
0.115851,
-0.120846,
0.127755,
0.189616,
-0.089516,
-0.243363,
0.067686
],
[
-0.072774,
0.32883,
-0.545254,
-0.236867,
-0.277831,
-0.103401,
0.195401,
0.406687
],
[
-0.330615,
-0.081502,
0.234002,
-0.089879,
-0.388941,
-0.277268,
-0.363158,
-0.022414
],
[
0.212982,
-0.285769,
0.099455,
0.252931,
0.251179,
-0.024098,
0.428763,
0.047684
],
[
0.858355,
-0.025873,
-0.172565,
-0.177821,
0.2166,
-0.011044,
0.117053,
-0.052764
],
[
0.229776,
-0.218023,
0.244962,
0.210624,
0.370465,
0.205382,
0.531206,
-0.1922
],
[
0.355677,
-0.086243,
0.009076,
0.476249,
0.441191,
0.18953,
0.539412,
-0.361245
],
[
0.694695,
-0.325312,
0.457947,
0.559046,
0.105189,
0.227625,
0.416482,
0.021265
]
],
"network.6.bias": [
-0.281343,
-0.106254,
-0.024524,
-0.306057,
-0.463677,
-0.387487,
-0.43291,
-0.424071
],
"network.8.weight": [
[
-0.104203,
-0.059444,
-0.027002,
-0.261278,
-0.251469,
-0.403786,
-0.54132,
-0.575888
]
],
"network.8.bias": [
0.591816
]
}
## Activation Signature
### 0
mean: [2.871086, 0.492916, -1.522849, -2.290309, 0.375210, -0.594153, 1.989282, 3.385367]
std: [2.607148, 1.713556, 2.083323, 1.956781, 1.986300, 2.519500, 1.859142, 3.215304]
### 2
mean: [3.932564, -1.890674, -2.461968, -2.375732, -1.876022, -2.175653, 1.281557, -2.019427]
std: [3.991612, 1.523205, 1.619538, 1.617759, 1.039848, 1.333537, 1.259685, 2.339986]
### 4
mean: [1.665189, -0.570968, 1.129688, 1.031128, 1.444782, 0.721244, 1.574093, -1.299042]
std: [1.980446, 0.231756, 1.351312, 1.340022, 1.631729, 0.950145, 2.018913, 1.141513]
### 6
mean: [-0.583761, -1.316322, -1.826046, 1.532304, 1.135993, 2.136589, 2.431978, 2.983207]
std: [0.358874, 1.296418, 1.994731, 2.055832, 1.749772, 2.803427, 3.188303, 3.776762]
### 8
mean: [-4.168307]
std: [5.858851]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
contains_abc
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 2.8710858821868896, "std": 2.6071479320526123}, "1": {"mean": 0.49291571974754333, "std": 1.7135555744171143}, "2": {"mean": -1.5228488445281982, "std": 2.0833234786987305}, "3": {"mean": -2.290308952331543, "std": 1.9567811489105225}, "4": {"mean": 0.37521013617515564, "std": 1.9862995147705078}, "5": {"mean": -0.5941531658172607, "std": 2.5195000171661377}, "6": {"mean": 1.9892817735671997, "std": 1.8591424226760864}, "7": {"mean": 3.3853671550750732, "std": 3.215303897857666}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 3.932563543319702, "std": 3.9916117191314697}, "1": {"mean": -1.8906736373901367, "std": 1.5232048034667969}, "2": {"mean": -2.461968183517456, "std": 1.6195378303527832}, "3": {"mean": -2.3757317066192627, "std": 1.6177589893341064}, "4": {"mean": -1.8760216236114502, "std": 1.0398484468460083}, "5": {"mean": -2.1756532192230225, "std": 1.333537220954895}, "6": {"mean": 1.281557321548462, "std": 1.2596853971481323}, "7": {"mean": -2.0194268226623535, "std": 2.3399856090545654}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 1.665189266204834, "std": 1.9804457426071167}, "1": {"mean": -0.5709676742553711, "std": 0.2317563146352768}, "2": {"mean": 1.1296881437301636, "std": 1.3513120412826538}, "3": {"mean": 1.031128168106079, "std": 1.3400224447250366}, "4": {"mean": 1.44478178024292, "std": 1.631729006767273}, "5": {"mean": 0.721244215965271, "std": 0.9501449465751648}, "6": {"mean": 1.5740927457809448, "std": 2.0189130306243896}, "7": {"mean": -1.2990424633026123, "std": 1.1415126323699951}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.5837606191635132, "std": 0.35887446999549866}, "1": {"mean": -1.316321611404419, "std": 1.2964180707931519}, "2": {"mean": -1.8260464668273926, "std": 1.9947305917739868}, "3": {"mean": 1.5323041677474976, "std": 2.0558319091796875}, "4": {"mean": 1.1359926462173462, "std": 1.7497724294662476}, "5": {"mean": 2.1365888118743896, "std": 2.803427219390869}, "6": {"mean": 2.431978225708008, "std": 3.188302516937256}, "7": {"mean": 2.9832067489624023, "std": 3.776761770248413}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -4.168307304382324, "std": 5.858851432800293}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.994256, 0.102341, 0.442418, 0.476359, -0.227079], [0.464053, 0.10793, -0.683167, 0.468514, 0.116525], [-0.966701, 0.05371, -0.118456, -0.416115, 0.199977], [-0.55023, -0.157028, -0.039426, -0.707505, 0.073083], [0.783432, -0.180171, 0.395953, -0.263763, -0.31643], [-1.163026, -0.134046, -0.117953, 0.057687, 0.417163], [0.561423, 0.023551, 0.494437, 0.433998, -0.380486], [0.941555, 0.794021, 0.350924, -0.145652, 0.376003]], "network.0.bias": [-0.218321, -0.002599, 0.475423, 0.189988, -0.14263, 0.705614, -0.246536, -0.117663], "network.2.weight": [[0.581781, 0.390293, -0.183966, -0.119161, 0.579511, -0.220023, 0.114465, 0.454403], [0.077997, 0.433929, 0.067014, -0.156913, 0.142318, -0.061972, -0.418976, -0.431814], [-0.230186, -0.206875, -0.209171, 0.290049, 0.170945, -0.174148, -0.327081, -0.222936], [-0.414146, 0.07863, -0.078696, -0.247239, -0.105345, -0.366655, 0.023904, -0.215355], [0.185311, -0.296189, -0.016446, 0.121244, -0.127189, -0.311022, -0.239208, -0.284723], [-0.124245, 0.007069, -0.345507, -0.237135, -0.290469, -0.389611, -0.052159, -0.236436], [0.066056, 0.888825, 0.138582, 0.152848, 0.046236, -0.039622, -0.211281, 0.201279], [-0.528235, -0.204489, 0.397719, 0.13527, -0.159439, 0.328483, -0.409247, 0.071007]], "network.2.bias": [-0.295274, -0.276011, -0.20186, -0.22829, -0.376274, -0.351964, -0.015976, 0.21176], "network.4.weight": [[0.423303, 0.515618, -0.2388, -0.346708, 0.207852, -0.051503, 0.28536, -0.308861], [-0.043227, 0.183803, 0.039459, 0.329109, 0.080987, -0.102755, -0.077266, -0.174015], [0.324629, -0.247428, -0.201797, -0.139283, 0.113109, 0.050388, 0.025873, -0.323747], [0.290719, -0.010427, -0.115445, -0.112045, -0.154247, -0.29205, 0.145282, -0.382782], [0.376227, 0.160062, -0.296109, -0.0553, 0.43879, 0.113344, 0.133735, -0.201651], [0.245147, 0.18848, 0.341661, 0.067266, -0.100878, 0.314056, -0.100176, -0.41602], [0.405636, -0.637752, -0.321328, 0.086581, -0.43528, 0.140721, 0.424316, -0.072967], [-0.351822, -0.055632, 0.305741, 0.158398, 0.058312, 0.096463, 0.285964, -0.0571]], "network.4.bias": [-0.343158, -0.269928, -0.144434, -0.252068, -0.197079, -0.059364, -0.577317, -0.250755], "network.6.weight": [[-0.061936, 0.115851, -0.120846, 0.127755, 0.189616, -0.089516, -0.243363, 0.067686], [-0.072774, 0.32883, -0.545254, -0.236867, -0.277831, -0.103401, 0.195401, 0.406687], [-0.330615, -0.081502, 0.234002, -0.089879, -0.388941, -0.277268, -0.363158, -0.022414], [0.212982, -0.285769, 0.099455, 0.252931, 0.251179, -0.024098, 0.428763, 0.047684], [0.858355, -0.025873, -0.172565, -0.177821, 0.2166, -0.011044, 0.117053, -0.052764], [0.229776, -0.218023, 0.244962, 0.210624, 0.370465, 0.205382, 0.531206, -0.1922], [0.355677, -0.086243, 0.009076, 0.476249, 0.441191, 0.18953, 0.539412, -0.361245], [0.694695, -0.325312, 0.457947, 0.559046, 0.105189, 0.227625, 0.416482, 0.021265]], "network.6.bias": [-0.281343, -0.106254, -0.024524, -0.306057, -0.463677, -0.387487, -0.43291, -0.424071], "network.8.weight": [[-0.104203, -0.059444, -0.027002, -0.261278, -0.251469, -0.403786, -0.54132, -0.575888]], "network.8.bias": [0.591816]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6828865110874176, "train_acc": 0.445, "val_loss": 0.7071729898452759, "val_acc": 0.38}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.5615401566028595, "train_acc": 0.615, "val_loss": 0.6308145523071289, "val_acc": 0.76}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.5440227389335632, "train_acc": 0.715, "val_loss": 0.5673523545265198, "val_acc": 0.82}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.5105676352977753, "train_acc": 0.85, "val_loss": 0.5156938433647156, "val_acc": 0.88}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.45326338708400726, "train_acc": 0.85, "val_loss": 0.5064292550086975, "val_acc": 0.7}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.38279713690280914, "train_acc": 0.825, "val_loss": 0.4472299814224243, "val_acc": 0.88}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.3614485263824463, "train_acc": 0.885, "val_loss": 0.3864646553993225, "val_acc": 0.92}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.33861932158470154, "train_acc": 0.875, "val_loss": 0.38518062233924866, "val_acc": 0.9}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.3081275522708893, "train_acc": 0.925, "val_loss": 0.38438743352890015, "val_acc": 0.88}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.263616681098938, "train_acc": 0.945, "val_loss": 0.41339054703712463, "val_acc": 0.7}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.2862344905734062, "train_acc": 0.805, "val_loss": 0.36690789461135864, "val_acc": 0.88}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.2539660185575485, "train_acc": 0.92, "val_loss": 0.3606065511703491, "val_acc": 0.88}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["contains_abc"], "degraded_stage": {"initial_val_loss": 0.7071729898452759, "final_val_loss": 0.6308145523071289, "initial_val_acc": 0.38, "final_val_acc": 0.76, "best_val_acc": 0.76}, "improved_stage": {"initial_val_loss": 0.5673523545265198, "final_val_loss": 0.3606065511703491, "initial_val_acc": 0.82, "final_val_acc": 0.88, "best_val_acc": 0.92, "best_epoch": 6}, "improvement": 0.16000000000000003, "first_improvement_epoch": 1}}
|
81
|
{"target_pattern": "alternating", "degraded_accuracy": 0.58, "improved_accuracy": 0.94, "improvement": 0.36, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 8736, "learning_rate": 0.08073496747622153, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "alternating", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["alternating"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.03626,
0.686585,
-0.238537,
0.536402,
0.422908
],
[
-0.077734,
0.046404,
0.433558,
0.497004,
0.086319
],
[
0.300083,
-0.152661,
0.49836,
-0.177625,
0.139213
],
[
0.073588,
-0.755745,
-0.285403,
-0.621345,
-0.038005
],
[
0.390015,
-0.740529,
-0.05941,
0.116657,
0.589252
],
[
-0.402233,
0.410601,
-0.100414,
-0.387989,
0.670516
],
[
-0.358213,
-0.127902,
-0.316953,
-0.03448,
0.330925
]
],
"network.0.bias": [
0.113183,
0.464798,
0.737304,
-0.426477,
0.353549,
-0.259611,
-0.076641
],
"network.2.weight": [
[
0.565796,
0.163567,
-0.408756,
-0.529605,
-0.237862,
-0.377624,
-0.22023
],
[
0.183098,
-0.073798,
-4.8e-05,
-0.042033,
0.413943,
0.339105,
0.230067
],
[
-0.361304,
0.585388,
0.358466,
0.291672,
0.122957,
0.656797,
0.492278
],
[
-0.778092,
-0.285527,
0.256069,
-0.203404,
0.425902,
0.105887,
-0.506225
],
[
0.329097,
0.465001,
-0.148753,
0.175804,
-0.406471,
-0.535489,
0.589207
],
[
0.198297,
0.379256,
0.212222,
0.476819,
0.534965,
0.618722,
0.594725
],
[
-0.798599,
-0.426619,
0.534708,
-0.5411,
0.546059,
-0.453764,
-0.767734
]
],
"network.2.bias": [
0.089113,
0.135941,
0.173961,
0.006656,
0.259186,
-0.240823,
-0.100997
],
"network.4.weight": [
[
-0.198965,
-0.747748,
0.375216,
0.195795,
0.383237,
-0.337249,
-0.127327
],
[
0.498777,
0.121771,
-0.727872,
-0.203602,
0.49235,
-0.259393,
-0.057217
],
[
-0.188645,
0.149945,
0.143596,
0.828671,
-0.210705,
-0.283701,
0.627613
],
[
-0.000451,
-0.319873,
-0.159466,
-0.277039,
-0.435805,
-0.156768,
0.332945
],
[
-0.218207,
0.124567,
0.354271,
-0.555648,
0.003056,
0.019428,
-0.436051
],
[
0.056423,
-0.35284,
0.26748,
-0.378747,
0.032575,
-0.517681,
-0.157389
],
[
0.003554,
0.328339,
0.435177,
-0.339528,
0.203838,
0.216899,
-0.709704
]
],
"network.4.bias": [
-0.294126,
0.479244,
-0.063351,
0.014348,
0.094422,
-0.442824,
0.048038
],
"network.6.weight": [
[
0.333912,
-0.534226,
-0.269665,
0.281195,
0.278139,
-0.155577,
0.600866
],
[
-0.285474,
-0.169386,
0.694626,
-0.529093,
-0.122797,
-0.520521,
-0.507326
],
[
-0.171303,
0.147731,
0.10032,
-0.466574,
-0.043031,
-0.416563,
-0.125682
],
[
-0.364274,
0.259921,
0.320861,
-0.231697,
-0.413509,
-0.4492,
-0.277984
],
[
0.472067,
-0.05788,
-0.122013,
0.641912,
0.32587,
0.223108,
-0.022579
],
[
0.374641,
-0.398677,
-0.117736,
0.101891,
0.270972,
0.214417,
0.560219
],
[
-0.300032,
0.379392,
-0.013945,
-0.21412,
-0.509316,
-0.581483,
0.179643
]
],
"network.6.bias": [
0.37976,
0.513883,
0.044777,
0.416936,
-0.222268,
-0.147347,
-0.23928
],
"network.8.weight": [
[
-0.686928,
0.441502,
0.037739,
0.366915,
-0.236989,
-0.505221,
0.375288
]
],
"network.8.bias": [
0.043905
]
}
## Activation Signature
### 0
mean: [2.492766, 2.530533, 1.664367, -3.687525, 0.342300, -0.226769, -1.083362]
std: [2.100118, 1.405259, 1.394842, 2.321223, 1.773727, 1.362355, 1.214221]
### 2
mean: [0.961168, 0.819353, 1.601450, -1.867627, 1.515545, 2.108148, -2.007834]
std: [1.515988, 0.834028, 1.175337, 2.044379, 1.332482, 1.637116, 2.545526]
### 4
mean: [-0.601911, 0.172635, -0.628825, -1.385721, 0.387824, -1.307519, 1.517065]
std: [0.980398, 1.799716, 1.169903, 0.916711, 0.623403, 1.017101, 1.113279]
### 6
mean: [0.946544, -0.202703, 0.049544, 0.151086, -0.273215, 0.438522, 0.184180]
std: [1.084345, 0.974213, 0.274248, 0.723295, 0.264613, 0.937750, 0.570126]
### 8
mean: [-0.588862]
std: [1.437972]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
alternating
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.03626,
0.686585,
-0.238537,
0.536402,
0.422908
],
[
-0.077734,
0.046404,
0.433558,
0.497004,
0.086319
],
[
0.300083,
-0.152661,
0.49836,
-0.177625,
0.139213
],
[
0.073588,
-0.755745,
-0.285403,
-0.621345,
-0.038005
],
[
0.390015,
-0.740529,
-0.05941,
0.116657,
0.589252
],
[
-0.402233,
0.410601,
-0.100414,
-0.387989,
0.670516
],
[
-0.358213,
-0.127902,
-0.316953,
-0.03448,
0.330925
]
],
"network.0.bias": [
0.113183,
0.464798,
0.737304,
-0.426477,
0.353549,
-0.259611,
-0.076641
],
"network.2.weight": [
[
0.565796,
0.163567,
-0.408756,
-0.529605,
-0.237862,
-0.377624,
-0.22023
],
[
0.183098,
-0.073798,
-4.8e-05,
-0.042033,
0.413943,
0.339105,
0.230067
],
[
-0.361304,
0.585388,
0.358466,
0.291672,
0.122957,
0.656797,
0.492278
],
[
-0.778092,
-0.285527,
0.256069,
-0.203404,
0.425902,
0.105887,
-0.506225
],
[
0.329097,
0.465001,
-0.148753,
0.175804,
-0.406471,
-0.535489,
0.589207
],
[
0.198297,
0.379256,
0.212222,
0.476819,
0.534965,
0.618722,
0.594725
],
[
-0.798599,
-0.426619,
0.534708,
-0.5411,
0.546059,
-0.453764,
-0.767734
]
],
"network.2.bias": [
0.089113,
0.135941,
0.173961,
0.006656,
0.259186,
-0.240823,
-0.100997
],
"network.4.weight": [
[
-0.198965,
-0.747748,
0.375216,
0.195795,
0.383237,
-0.337249,
-0.127327
],
[
0.498777,
0.121771,
-0.727872,
-0.203602,
0.49235,
-0.259393,
-0.057217
],
[
-0.188645,
0.149945,
0.143596,
0.828671,
-0.210705,
-0.283701,
0.627613
],
[
-0.000451,
-0.319873,
-0.159466,
-0.277039,
-0.435805,
-0.156768,
0.332945
],
[
-0.218207,
0.124567,
0.354271,
-0.555648,
0.003056,
0.019428,
-0.436051
],
[
0.056423,
-0.35284,
0.26748,
-0.378747,
0.032575,
-0.517681,
-0.157389
],
[
0.003554,
0.328339,
0.435177,
-0.339528,
0.203838,
0.216899,
-0.709704
]
],
"network.4.bias": [
-0.294126,
0.479244,
-0.063351,
0.014348,
0.094422,
-0.442824,
0.048038
],
"network.6.weight": [
[
0.333912,
-0.534226,
-0.269665,
0.281195,
0.278139,
-0.155577,
0.600866
],
[
-0.285474,
-0.169386,
0.694626,
-0.529093,
-0.122797,
-0.520521,
-0.507326
],
[
-0.171303,
0.147731,
0.10032,
-0.466574,
-0.043031,
-0.416563,
-0.125682
],
[
-0.364274,
0.259921,
0.320861,
-0.231697,
-0.413509,
-0.4492,
-0.277984
],
[
0.472067,
-0.05788,
-0.122013,
0.641912,
0.32587,
0.223108,
-0.022579
],
[
0.374641,
-0.398677,
-0.117736,
0.101891,
0.270972,
0.214417,
0.560219
],
[
-0.300032,
0.379392,
-0.013945,
-0.21412,
-0.509316,
-0.581483,
0.179643
]
],
"network.6.bias": [
0.37976,
0.513883,
0.044777,
0.416936,
-0.222268,
-0.147347,
-0.23928
],
"network.8.weight": [
[
-0.686928,
0.441502,
0.037739,
0.366915,
-0.236989,
-0.505221,
0.375288
]
],
"network.8.bias": [
0.043905
]
}
## Activation Signature
### 0
mean: [2.492766, 2.530533, 1.664367, -3.687525, 0.342300, -0.226769, -1.083362]
std: [2.100118, 1.405259, 1.394842, 2.321223, 1.773727, 1.362355, 1.214221]
### 2
mean: [0.961168, 0.819353, 1.601450, -1.867627, 1.515545, 2.108148, -2.007834]
std: [1.515988, 0.834028, 1.175337, 2.044379, 1.332482, 1.637116, 2.545526]
### 4
mean: [-0.601911, 0.172635, -0.628825, -1.385721, 0.387824, -1.307519, 1.517065]
std: [0.980398, 1.799716, 1.169903, 0.916711, 0.623403, 1.017101, 1.113279]
### 6
mean: [0.946544, -0.202703, 0.049544, 0.151086, -0.273215, 0.438522, 0.184180]
std: [1.084345, 0.974213, 0.274248, 0.723295, 0.264613, 0.937750, 0.570126]
### 8
mean: [-0.588862]
std: [1.437972]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
alternating
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 2.4927659034729004, "std": 2.100118398666382}, "1": {"mean": 2.5305325984954834, "std": 1.4052588939666748}, "2": {"mean": 1.6643673181533813, "std": 1.394842267036438}, "3": {"mean": -3.6875247955322266, "std": 2.321223258972168}, "4": {"mean": 0.3422999978065491, "std": 1.7737274169921875}, "5": {"mean": -0.2267685830593109, "std": 1.3623554706573486}, "6": {"mean": -1.0833618640899658, "std": 1.2142211198806763}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.9611676335334778, "std": 1.515987753868103}, "1": {"mean": 0.8193534016609192, "std": 0.8340282440185547}, "2": {"mean": 1.6014502048492432, "std": 1.1753371953964233}, "3": {"mean": -1.8676272630691528, "std": 2.0443787574768066}, "4": {"mean": 1.5155446529388428, "std": 1.3324817419052124}, "5": {"mean": 2.108147621154785, "std": 1.637115716934204}, "6": {"mean": -2.0078344345092773, "std": 2.5455257892608643}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.6019105911254883, "std": 0.980398416519165}, "1": {"mean": 0.1726352721452713, "std": 1.7997159957885742}, "2": {"mean": -0.6288250684738159, "std": 1.1699031591415405}, "3": {"mean": -1.38572096824646, "std": 0.916710615158081}, "4": {"mean": 0.38782379031181335, "std": 0.6234032511711121}, "5": {"mean": -1.3075186014175415, "std": 1.0171006917953491}, "6": {"mean": 1.5170652866363525, "std": 1.1132792234420776}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.9465444684028625, "std": 1.0843451023101807}, "1": {"mean": -0.20270320773124695, "std": 0.9742129445075989}, "2": {"mean": 0.04954385757446289, "std": 0.27424827218055725}, "3": {"mean": 0.15108636021614075, "std": 0.7232945561408997}, "4": {"mean": -0.2732151746749878, "std": 0.2646133005619049}, "5": {"mean": 0.43852224946022034, "std": 0.9377502202987671}, "6": {"mean": 0.1841803938150406, "std": 0.5701264142990112}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.5888621211051941, "std": 1.437972068786621}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.03626, 0.686585, -0.238537, 0.536402, 0.422908], [-0.077734, 0.046404, 0.433558, 0.497004, 0.086319], [0.300083, -0.152661, 0.49836, -0.177625, 0.139213], [0.073588, -0.755745, -0.285403, -0.621345, -0.038005], [0.390015, -0.740529, -0.05941, 0.116657, 0.589252], [-0.402233, 0.410601, -0.100414, -0.387989, 0.670516], [-0.358213, -0.127902, -0.316953, -0.03448, 0.330925]], "network.0.bias": [0.113183, 0.464798, 0.737304, -0.426477, 0.353549, -0.259611, -0.076641], "network.2.weight": [[0.565796, 0.163567, -0.408756, -0.529605, -0.237862, -0.377624, -0.22023], [0.183098, -0.073798, -4.8e-05, -0.042033, 0.413943, 0.339105, 0.230067], [-0.361304, 0.585388, 0.358466, 0.291672, 0.122957, 0.656797, 0.492278], [-0.778092, -0.285527, 0.256069, -0.203404, 0.425902, 0.105887, -0.506225], [0.329097, 0.465001, -0.148753, 0.175804, -0.406471, -0.535489, 0.589207], [0.198297, 0.379256, 0.212222, 0.476819, 0.534965, 0.618722, 0.594725], [-0.798599, -0.426619, 0.534708, -0.5411, 0.546059, -0.453764, -0.767734]], "network.2.bias": [0.089113, 0.135941, 0.173961, 0.006656, 0.259186, -0.240823, -0.100997], "network.4.weight": [[-0.198965, -0.747748, 0.375216, 0.195795, 0.383237, -0.337249, -0.127327], [0.498777, 0.121771, -0.727872, -0.203602, 0.49235, -0.259393, -0.057217], [-0.188645, 0.149945, 0.143596, 0.828671, -0.210705, -0.283701, 0.627613], [-0.000451, -0.319873, -0.159466, -0.277039, -0.435805, -0.156768, 0.332945], [-0.218207, 0.124567, 0.354271, -0.555648, 0.003056, 0.019428, -0.436051], [0.056423, -0.35284, 0.26748, -0.378747, 0.032575, -0.517681, -0.157389], [0.003554, 0.328339, 0.435177, -0.339528, 0.203838, 0.216899, -0.709704]], "network.4.bias": [-0.294126, 0.479244, -0.063351, 0.014348, 0.094422, -0.442824, 0.048038], "network.6.weight": [[0.333912, -0.534226, -0.269665, 0.281195, 0.278139, -0.155577, 0.600866], [-0.285474, -0.169386, 0.694626, -0.529093, -0.122797, -0.520521, -0.507326], [-0.171303, 0.147731, 0.10032, -0.466574, -0.043031, -0.416563, -0.125682], [-0.364274, 0.259921, 0.320861, -0.231697, -0.413509, -0.4492, -0.277984], [0.472067, -0.05788, -0.122013, 0.641912, 0.32587, 0.223108, -0.022579], [0.374641, -0.398677, -0.117736, 0.101891, 0.270972, 0.214417, 0.560219], [-0.300032, 0.379392, -0.013945, -0.21412, -0.509316, -0.581483, 0.179643]], "network.6.bias": [0.37976, 0.513883, 0.044777, 0.416936, -0.222268, -0.147347, -0.23928], "network.8.weight": [[-0.686928, 0.441502, 0.037739, 0.366915, -0.236989, -0.505221, 0.375288]], "network.8.bias": [0.043905]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6974546313285828, "train_acc": 0.56, "val_loss": 0.7094525098800659, "val_acc": 0.48}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6709895730018616, "train_acc": 0.58, "val_loss": 0.6398505568504333, "val_acc": 0.58}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.540233924984932, "train_acc": 0.72, "val_loss": 0.5526058673858643, "val_acc": 0.66}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.47767162322998047, "train_acc": 0.735, "val_loss": 0.8160923719406128, "val_acc": 0.62}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.4479011297225952, "train_acc": 0.775, "val_loss": 0.26696500182151794, "val_acc": 0.94}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.30667078495025635, "train_acc": 0.895, "val_loss": 0.2521587014198303, "val_acc": 0.92}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.20292535424232483, "train_acc": 0.915, "val_loss": 0.2511622905731201, "val_acc": 0.88}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.18315835297107697, "train_acc": 0.965, "val_loss": 0.31013786792755127, "val_acc": 0.88}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.19383518397808075, "train_acc": 0.95, "val_loss": 0.3760833442211151, "val_acc": 0.88}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.20386646687984467, "train_acc": 0.925, "val_loss": 0.31460705399513245, "val_acc": 0.9}], "summary": {"total_epochs": 10, "degraded_epochs": 2, "improved_epochs": 8, "patterns": ["alternating"], "degraded_stage": {"initial_val_loss": 0.7094525098800659, "final_val_loss": 0.6398505568504333, "initial_val_acc": 0.48, "final_val_acc": 0.58, "best_val_acc": 0.58}, "improved_stage": {"initial_val_loss": 0.5526058673858643, "final_val_loss": 0.31460705399513245, "initial_val_acc": 0.66, "final_val_acc": 0.9, "best_val_acc": 0.94, "best_epoch": 4}, "improvement": 0.36, "first_improvement_epoch": 1}}
|
82
|
{"target_pattern": "has_majority", "degraded_accuracy": 0.4, "improved_accuracy": 0.62, "improvement": 0.21999999999999997, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 8497, "learning_rate": 0.09416931889116947, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "has_majority", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["has_majority"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.170991,
0.210213,
0.435689,
0.500816,
0.029244
],
[
-0.47366,
0.319196,
0.560351,
0.533075,
-0.24786
],
[
-0.409177,
0.267049,
0.281556,
0.359557,
0.212762
],
[
-0.318581,
-0.015791,
0.691275,
0.524957,
0.154134
],
[
-0.303737,
-0.369945,
-0.414301,
-0.173619,
-0.358026
],
[
-0.734874,
-0.22692,
0.205519,
0.335101,
-0.047507
],
[
-0.696496,
-0.067191,
0.281962,
0.080295,
-0.081014
]
],
"network.0.bias": [
-0.312992,
-0.051137,
0.149302,
-0.469924,
-0.386742,
-0.101329,
-0.071872
],
"network.2.weight": [
[
0.219816,
0.44916,
-0.30356,
0.057322,
-0.133078,
0.01596,
0.084141
],
[
-0.423464,
-0.255678,
-0.575806,
0.014072,
0.08965,
-0.207519,
-0.002788
],
[
0.438608,
0.537502,
0.077846,
0.051779,
-0.389827,
0.476377,
0.129991
],
[
-0.39967,
0.021009,
0.637678,
-0.234654,
-0.29631,
-0.043683,
0.477883
],
[
0.309047,
0.191782,
0.293548,
0.207167,
0.533153,
0.53846,
0.075266
],
[
0.569505,
0.38788,
0.19913,
0.547033,
0.128477,
-0.014134,
0.155259
],
[
-0.248487,
0.211948,
-0.603589,
0.040064,
-0.162536,
0.57112,
-0.123917
]
],
"network.2.bias": [
0.170922,
-0.358236,
-0.149757,
0.756421,
-0.586095,
0.21326,
-0.153096
],
"network.4.weight": [
[
-0.218751,
0.020946,
0.042054,
0.622507,
0.001695,
-0.559687,
0.131056
],
[
0.388845,
0.162456,
0.569646,
-0.2551,
-0.058892,
0.381986,
-0.220929
],
[
-0.032711,
0.522055,
-0.058926,
-0.17141,
0.566095,
-0.04121,
-0.009763
],
[
-0.470124,
-0.288176,
-0.542077,
0.702077,
-0.02614,
-0.292512,
0.40267
],
[
-0.437633,
-0.125551,
0.116542,
-0.381219,
-0.501813,
-0.093354,
-0.348423
],
[
0.19938,
0.101779,
0.541739,
-0.659845,
0.604199,
0.435445,
-0.057644
],
[
-0.216906,
-0.071037,
-0.370426,
-0.045633,
-0.102931,
-0.34404,
0.056284
]
],
"network.4.bias": [
0.099111,
-0.188924,
0.254572,
0.185027,
0.08947,
-0.459338,
0.095335
],
"network.6.weight": [
[
0.621759,
0.145268,
-0.099608,
0.685974,
0.318944,
0.069314,
0.36742
],
[
0.128528,
0.188482,
-0.066796,
-0.143106,
-0.057335,
0.447639,
-0.37088
],
[
-0.375905,
-0.202564,
-0.110688,
0.141727,
-0.351516,
-0.014148,
-0.154149
],
[
0.275107,
0.154683,
-0.572125,
0.171171,
-0.000745,
-0.092802,
-0.319661
],
[
0.048685,
-0.064746,
0.19531,
-0.152249,
0.195341,
-0.028609,
-0.370427
],
[
-0.039028,
-0.404541,
-0.386221,
0.001977,
-0.200186,
-0.43054,
0.330563
],
[
0.058446,
0.022967,
0.161342,
0.262836,
0.12703,
-0.276677,
-0.258151
]
],
"network.6.bias": [
0.059208,
0.098363,
-0.185688,
0.04366,
-0.186615,
-0.105415,
-0.25096
],
"network.8.weight": [
[
-0.265468,
-0.332023,
0.08912,
-0.46528,
-0.118066,
0.048189,
0.373536
],
[
0.328832,
0.256873,
0.286321,
-0.139188,
0.05427,
0.093337,
-0.25738
],
[
-0.106244,
0.181433,
0.128323,
0.03665,
0.041703,
-0.052524,
0.291706
],
[
0.230101,
0.321987,
0.335409,
-0.206532,
-0.161188,
0.054628,
0.05691
],
[
-0.373117,
0.302427,
-0.281129,
-0.157494,
-0.113525,
0.060059,
-0.010057
],
[
-0.206905,
-0.494542,
0.373962,
-0.142868,
-0.096906,
-0.061417,
0.06385
],
[
-0.405785,
0.370088,
0.351312,
0.079951,
-0.054853,
0.254733,
0.264601
]
],
"network.8.bias": [
-0.152741,
-0.124841,
-0.248284,
-0.391008,
-0.096885,
-0.140376,
-0.511048
],
"network.10.weight": [
[
-0.049923,
-0.079919,
-0.33683,
0.00153,
-0.132912,
0.051606,
-0.090745
],
[
-0.268946,
-0.0657,
0.098332,
-0.423159,
-0.419959,
0.28671,
0.116672
],
[
0.072205,
0.100874,
0.148591,
-0.150975,
-0.146471,
-0.249944,
-0.208925
],
[
0.037939,
-0.628403,
0.216762,
-0.197952,
-0.024311,
-0.070836,
0.10101
],
[
-0.106267,
-0.050174,
0.062454,
-0.477727,
-0.306895,
0.017296,
-0.441493
],
[
0.075019,
-0.343319,
0.101234,
0.067812,
-0.125627,
0.100439,
0.091689
],
[
0.12413,
0.140566,
-0.087008,
0.313492,
0.206603,
-0.046209,
0.392587
]
],
"network.10.bias": [
-0.463418,
-0.124024,
0.373785,
0.064816,
0.079523,
-0.266358,
-0.183071
],
"network.12.weight": [
[
-0.090755,
-0.053457,
0.409714,
0.080106,
0.124657,
-0.10287,
-0.129858
]
],
"network.12.bias": [
0.254779
]
}
## Activation Signature
### 0
mean: [1.877419, 1.949032, 1.748322, 1.864681, -3.118159, -0.340018, -0.399803]
std: [1.518600, 1.872332, 1.373093, 1.794408, 1.868610, 1.716873, 1.416026]
### 2
mean: [1.122431, -2.803592, 2.303865, 0.827689, 1.613534, 3.598089, -0.975907]
std: [0.850251, 1.799681, 1.938373, 0.360520, 1.692481, 2.561428, 0.629360]
### 4
mean: [-1.544228, 2.635600, 0.765194, -2.120873, -1.648190, 3.085748, -2.463576]
std: [1.553883, 2.288697, 0.646992, 2.222901, 1.153066, 3.248377, 1.916525]
### 6
mean: [0.746111, 2.007446, -0.874707, -0.234928, -0.320199, -2.895582, -0.929411]
std: [0.375756, 1.746192, 0.548053, 0.366719, 0.093694, 2.451650, 0.733467]
### 8
mean: [-1.039723, 0.629463, 0.038427, 0.417104, 0.224254, -1.294393, -0.067028]
std: [0.642826, 0.560138, 0.283071, 0.644707, 0.431945, 0.921053, 0.524238]
### 10
mean: [-0.617817, -0.468129, 0.298820, -0.386845, -0.356534, -0.457120, 0.185527]
std: [0.177954, 0.366043, 0.110661, 0.404086, 0.518030, 0.150717, 0.422253]
### 12
mean: [0.347892]
std: [0.094591]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
has_majority
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.170991,
0.210213,
0.435689,
0.500816,
0.029244
],
[
-0.47366,
0.319196,
0.560351,
0.533075,
-0.24786
],
[
-0.409177,
0.267049,
0.281556,
0.359557,
0.212762
],
[
-0.318581,
-0.015791,
0.691275,
0.524957,
0.154134
],
[
-0.303737,
-0.369945,
-0.414301,
-0.173619,
-0.358026
],
[
-0.734874,
-0.22692,
0.205519,
0.335101,
-0.047507
],
[
-0.696496,
-0.067191,
0.281962,
0.080295,
-0.081014
]
],
"network.0.bias": [
-0.312992,
-0.051137,
0.149302,
-0.469924,
-0.386742,
-0.101329,
-0.071872
],
"network.2.weight": [
[
0.219816,
0.44916,
-0.30356,
0.057322,
-0.133078,
0.01596,
0.084141
],
[
-0.423464,
-0.255678,
-0.575806,
0.014072,
0.08965,
-0.207519,
-0.002788
],
[
0.438608,
0.537502,
0.077846,
0.051779,
-0.389827,
0.476377,
0.129991
],
[
-0.39967,
0.021009,
0.637678,
-0.234654,
-0.29631,
-0.043683,
0.477883
],
[
0.309047,
0.191782,
0.293548,
0.207167,
0.533153,
0.53846,
0.075266
],
[
0.569505,
0.38788,
0.19913,
0.547033,
0.128477,
-0.014134,
0.155259
],
[
-0.248487,
0.211948,
-0.603589,
0.040064,
-0.162536,
0.57112,
-0.123917
]
],
"network.2.bias": [
0.170922,
-0.358236,
-0.149757,
0.756421,
-0.586095,
0.21326,
-0.153096
],
"network.4.weight": [
[
-0.218751,
0.020946,
0.042054,
0.622507,
0.001695,
-0.559687,
0.131056
],
[
0.388845,
0.162456,
0.569646,
-0.2551,
-0.058892,
0.381986,
-0.220929
],
[
-0.032711,
0.522055,
-0.058926,
-0.17141,
0.566095,
-0.04121,
-0.009763
],
[
-0.470124,
-0.288176,
-0.542077,
0.702077,
-0.02614,
-0.292512,
0.40267
],
[
-0.437633,
-0.125551,
0.116542,
-0.381219,
-0.501813,
-0.093354,
-0.348423
],
[
0.19938,
0.101779,
0.541739,
-0.659845,
0.604199,
0.435445,
-0.057644
],
[
-0.216906,
-0.071037,
-0.370426,
-0.045633,
-0.102931,
-0.34404,
0.056284
]
],
"network.4.bias": [
0.099111,
-0.188924,
0.254572,
0.185027,
0.08947,
-0.459338,
0.095335
],
"network.6.weight": [
[
0.621759,
0.145268,
-0.099608,
0.685974,
0.318944,
0.069314,
0.36742
],
[
0.128528,
0.188482,
-0.066796,
-0.143106,
-0.057335,
0.447639,
-0.37088
],
[
-0.375905,
-0.202564,
-0.110688,
0.141727,
-0.351516,
-0.014148,
-0.154149
],
[
0.275107,
0.154683,
-0.572125,
0.171171,
-0.000745,
-0.092802,
-0.319661
],
[
0.048685,
-0.064746,
0.19531,
-0.152249,
0.195341,
-0.028609,
-0.370427
],
[
-0.039028,
-0.404541,
-0.386221,
0.001977,
-0.200186,
-0.43054,
0.330563
],
[
0.058446,
0.022967,
0.161342,
0.262836,
0.12703,
-0.276677,
-0.258151
]
],
"network.6.bias": [
0.059208,
0.098363,
-0.185688,
0.04366,
-0.186615,
-0.105415,
-0.25096
],
"network.8.weight": [
[
-0.265468,
-0.332023,
0.08912,
-0.46528,
-0.118066,
0.048189,
0.373536
],
[
0.328832,
0.256873,
0.286321,
-0.139188,
0.05427,
0.093337,
-0.25738
],
[
-0.106244,
0.181433,
0.128323,
0.03665,
0.041703,
-0.052524,
0.291706
],
[
0.230101,
0.321987,
0.335409,
-0.206532,
-0.161188,
0.054628,
0.05691
],
[
-0.373117,
0.302427,
-0.281129,
-0.157494,
-0.113525,
0.060059,
-0.010057
],
[
-0.206905,
-0.494542,
0.373962,
-0.142868,
-0.096906,
-0.061417,
0.06385
],
[
-0.405785,
0.370088,
0.351312,
0.079951,
-0.054853,
0.254733,
0.264601
]
],
"network.8.bias": [
-0.152741,
-0.124841,
-0.248284,
-0.391008,
-0.096885,
-0.140376,
-0.511048
],
"network.10.weight": [
[
-0.049923,
-0.079919,
-0.33683,
0.00153,
-0.132912,
0.051606,
-0.090745
],
[
-0.268946,
-0.0657,
0.098332,
-0.423159,
-0.419959,
0.28671,
0.116672
],
[
0.072205,
0.100874,
0.148591,
-0.150975,
-0.146471,
-0.249944,
-0.208925
],
[
0.037939,
-0.628403,
0.216762,
-0.197952,
-0.024311,
-0.070836,
0.10101
],
[
-0.106267,
-0.050174,
0.062454,
-0.477727,
-0.306895,
0.017296,
-0.441493
],
[
0.075019,
-0.343319,
0.101234,
0.067812,
-0.125627,
0.100439,
0.091689
],
[
0.12413,
0.140566,
-0.087008,
0.313492,
0.206603,
-0.046209,
0.392587
]
],
"network.10.bias": [
-0.463418,
-0.124024,
0.373785,
0.064816,
0.079523,
-0.266358,
-0.183071
],
"network.12.weight": [
[
-0.090755,
-0.053457,
0.409714,
0.080106,
0.124657,
-0.10287,
-0.129858
]
],
"network.12.bias": [
0.254779
]
}
## Activation Signature
### 0
mean: [1.877419, 1.949032, 1.748322, 1.864681, -3.118159, -0.340018, -0.399803]
std: [1.518600, 1.872332, 1.373093, 1.794408, 1.868610, 1.716873, 1.416026]
### 2
mean: [1.122431, -2.803592, 2.303865, 0.827689, 1.613534, 3.598089, -0.975907]
std: [0.850251, 1.799681, 1.938373, 0.360520, 1.692481, 2.561428, 0.629360]
### 4
mean: [-1.544228, 2.635600, 0.765194, -2.120873, -1.648190, 3.085748, -2.463576]
std: [1.553883, 2.288697, 0.646992, 2.222901, 1.153066, 3.248377, 1.916525]
### 6
mean: [0.746111, 2.007446, -0.874707, -0.234928, -0.320199, -2.895582, -0.929411]
std: [0.375756, 1.746192, 0.548053, 0.366719, 0.093694, 2.451650, 0.733467]
### 8
mean: [-1.039723, 0.629463, 0.038427, 0.417104, 0.224254, -1.294393, -0.067028]
std: [0.642826, 0.560138, 0.283071, 0.644707, 0.431945, 0.921053, 0.524238]
### 10
mean: [-0.617817, -0.468129, 0.298820, -0.386845, -0.356534, -0.457120, 0.185527]
std: [0.177954, 0.366043, 0.110661, 0.404086, 0.518030, 0.150717, 0.422253]
### 12
mean: [0.347892]
std: [0.094591]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
has_majority
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.8774192333221436, "std": 1.5186001062393188}, "1": {"mean": 1.9490315914154053, "std": 1.8723316192626953}, "2": {"mean": 1.7483223676681519, "std": 1.3730934858322144}, "3": {"mean": 1.8646811246871948, "std": 1.7944082021713257}, "4": {"mean": -3.118159055709839, "std": 1.8686095476150513}, "5": {"mean": -0.3400175869464874, "std": 1.7168734073638916}, "6": {"mean": -0.39980289340019226, "std": 1.416025996208191}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 1.1224313974380493, "std": 0.850251317024231}, "1": {"mean": -2.8035922050476074, "std": 1.7996814250946045}, "2": {"mean": 2.303865432739258, "std": 1.93837308883667}, "3": {"mean": 0.8276886940002441, "std": 0.36052030324935913}, "4": {"mean": 1.6135337352752686, "std": 1.6924806833267212}, "5": {"mean": 3.598088502883911, "std": 2.5614278316497803}, "6": {"mean": -0.9759073257446289, "std": 0.6293596029281616}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -1.5442277193069458, "std": 1.5538825988769531}, "1": {"mean": 2.6355996131896973, "std": 2.288696527481079}, "2": {"mean": 0.7651941776275635, "std": 0.6469917297363281}, "3": {"mean": -2.12087345123291, "std": 2.2229013442993164}, "4": {"mean": -1.6481901407241821, "std": 1.1530662775039673}, "5": {"mean": 3.085747718811035, "std": 3.2483768463134766}, "6": {"mean": -2.463575601577759, "std": 1.9165252447128296}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.7461106777191162, "std": 0.3757556676864624}, "1": {"mean": 2.007445812225342, "std": 1.7461919784545898}, "2": {"mean": -0.8747066259384155, "std": 0.5480532050132751}, "3": {"mean": -0.23492763936519623, "std": 0.36671921610832214}, "4": {"mean": -0.3201988935470581, "std": 0.09369393438100815}, "5": {"mean": -2.895582437515259, "std": 2.4516501426696777}, "6": {"mean": -0.9294108748435974, "std": 0.7334668636322021}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.0397229194641113, "std": 0.6428259015083313}, "1": {"mean": 0.6294627785682678, "std": 0.560138463973999}, "2": {"mean": 0.038427211344242096, "std": 0.2830706238746643}, "3": {"mean": 0.41710400581359863, "std": 0.6447067856788635}, "4": {"mean": 0.2242538034915924, "std": 0.43194472789764404}, "5": {"mean": -1.2943934202194214, "std": 0.9210534691810608}, "6": {"mean": -0.06702812016010284, "std": 0.5242377519607544}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -0.6178168058395386, "std": 0.17795436084270477}, "1": {"mean": -0.46812912821769714, "std": 0.36604321002960205}, "2": {"mean": 0.29881972074508667, "std": 0.11066117882728577}, "3": {"mean": -0.38684526085853577, "std": 0.4040863513946533}, "4": {"mean": -0.3565336763858795, "std": 0.5180304050445557}, "5": {"mean": -0.4571199119091034, "std": 0.15071672201156616}, "6": {"mean": 0.18552730977535248, "std": 0.4222527742385864}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": 0.34789204597473145, "std": 0.09459110349416733}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.170991, 0.210213, 0.435689, 0.500816, 0.029244], [-0.47366, 0.319196, 0.560351, 0.533075, -0.24786], [-0.409177, 0.267049, 0.281556, 0.359557, 0.212762], [-0.318581, -0.015791, 0.691275, 0.524957, 0.154134], [-0.303737, -0.369945, -0.414301, -0.173619, -0.358026], [-0.734874, -0.22692, 0.205519, 0.335101, -0.047507], [-0.696496, -0.067191, 0.281962, 0.080295, -0.081014]], "network.0.bias": [-0.312992, -0.051137, 0.149302, -0.469924, -0.386742, -0.101329, -0.071872], "network.2.weight": [[0.219816, 0.44916, -0.30356, 0.057322, -0.133078, 0.01596, 0.084141], [-0.423464, -0.255678, -0.575806, 0.014072, 0.08965, -0.207519, -0.002788], [0.438608, 0.537502, 0.077846, 0.051779, -0.389827, 0.476377, 0.129991], [-0.39967, 0.021009, 0.637678, -0.234654, -0.29631, -0.043683, 0.477883], [0.309047, 0.191782, 0.293548, 0.207167, 0.533153, 0.53846, 0.075266], [0.569505, 0.38788, 0.19913, 0.547033, 0.128477, -0.014134, 0.155259], [-0.248487, 0.211948, -0.603589, 0.040064, -0.162536, 0.57112, -0.123917]], "network.2.bias": [0.170922, -0.358236, -0.149757, 0.756421, -0.586095, 0.21326, -0.153096], "network.4.weight": [[-0.218751, 0.020946, 0.042054, 0.622507, 0.001695, -0.559687, 0.131056], [0.388845, 0.162456, 0.569646, -0.2551, -0.058892, 0.381986, -0.220929], [-0.032711, 0.522055, -0.058926, -0.17141, 0.566095, -0.04121, -0.009763], [-0.470124, -0.288176, -0.542077, 0.702077, -0.02614, -0.292512, 0.40267], [-0.437633, -0.125551, 0.116542, -0.381219, -0.501813, -0.093354, -0.348423], [0.19938, 0.101779, 0.541739, -0.659845, 0.604199, 0.435445, -0.057644], [-0.216906, -0.071037, -0.370426, -0.045633, -0.102931, -0.34404, 0.056284]], "network.4.bias": [0.099111, -0.188924, 0.254572, 0.185027, 0.08947, -0.459338, 0.095335], "network.6.weight": [[0.621759, 0.145268, -0.099608, 0.685974, 0.318944, 0.069314, 0.36742], [0.128528, 0.188482, -0.066796, -0.143106, -0.057335, 0.447639, -0.37088], [-0.375905, -0.202564, -0.110688, 0.141727, -0.351516, -0.014148, -0.154149], [0.275107, 0.154683, -0.572125, 0.171171, -0.000745, -0.092802, -0.319661], [0.048685, -0.064746, 0.19531, -0.152249, 0.195341, -0.028609, -0.370427], [-0.039028, -0.404541, -0.386221, 0.001977, -0.200186, -0.43054, 0.330563], [0.058446, 0.022967, 0.161342, 0.262836, 0.12703, -0.276677, -0.258151]], "network.6.bias": [0.059208, 0.098363, -0.185688, 0.04366, -0.186615, -0.105415, -0.25096], "network.8.weight": [[-0.265468, -0.332023, 0.08912, -0.46528, -0.118066, 0.048189, 0.373536], [0.328832, 0.256873, 0.286321, -0.139188, 0.05427, 0.093337, -0.25738], [-0.106244, 0.181433, 0.128323, 0.03665, 0.041703, -0.052524, 0.291706], [0.230101, 0.321987, 0.335409, -0.206532, -0.161188, 0.054628, 0.05691], [-0.373117, 0.302427, -0.281129, -0.157494, -0.113525, 0.060059, -0.010057], [-0.206905, -0.494542, 0.373962, -0.142868, -0.096906, -0.061417, 0.06385], [-0.405785, 0.370088, 0.351312, 0.079951, -0.054853, 0.254733, 0.264601]], "network.8.bias": [-0.152741, -0.124841, -0.248284, -0.391008, -0.096885, -0.140376, -0.511048], "network.10.weight": [[-0.049923, -0.079919, -0.33683, 0.00153, -0.132912, 0.051606, -0.090745], [-0.268946, -0.0657, 0.098332, -0.423159, -0.419959, 0.28671, 0.116672], [0.072205, 0.100874, 0.148591, -0.150975, -0.146471, -0.249944, -0.208925], [0.037939, -0.628403, 0.216762, -0.197952, -0.024311, -0.070836, 0.10101], [-0.106267, -0.050174, 0.062454, -0.477727, -0.306895, 0.017296, -0.441493], [0.075019, -0.343319, 0.101234, 0.067812, -0.125627, 0.100439, 0.091689], [0.12413, 0.140566, -0.087008, 0.313492, 0.206603, -0.046209, 0.392587]], "network.10.bias": [-0.463418, -0.124024, 0.373785, 0.064816, 0.079523, -0.266358, -0.183071], "network.12.weight": [[-0.090755, -0.053457, 0.409714, 0.080106, 0.124657, -0.10287, -0.129858]], "network.12.bias": [0.254779]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7190490067005157, "train_acc": 0.48, "val_loss": 0.8147441744804382, "val_acc": 0.4}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6587834060192108, "train_acc": 0.605, "val_loss": 0.7162363529205322, "val_acc": 0.4}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.6718868911266327, "train_acc": 0.545, "val_loss": 0.8462559580802917, "val_acc": 0.48}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.6848874986171722, "train_acc": 0.615, "val_loss": 0.6440548896789551, "val_acc": 0.62}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.6938825845718384, "train_acc": 0.535, "val_loss": 0.6481606960296631, "val_acc": 0.6}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.7107396125793457, "train_acc": 0.475, "val_loss": 0.6508792042732239, "val_acc": 0.6}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.7013256251811981, "train_acc": 0.475, "val_loss": 0.6565182209014893, "val_acc": 0.6}], "summary": {"total_epochs": 7, "degraded_epochs": 2, "improved_epochs": 5, "patterns": ["has_majority"], "degraded_stage": {"initial_val_loss": 0.8147441744804382, "final_val_loss": 0.7162363529205322, "initial_val_acc": 0.4, "final_val_acc": 0.4, "best_val_acc": 0.4}, "improved_stage": {"initial_val_loss": 0.8462559580802917, "final_val_loss": 0.6565182209014893, "initial_val_acc": 0.48, "final_val_acc": 0.6, "best_val_acc": 0.62, "best_epoch": 3}, "improvement": 0.21999999999999997, "first_improvement_epoch": 1}}
|
83
|
{"target_pattern": "decreasing_pairs", "degraded_accuracy": 0.66, "improved_accuracy": 0.94, "improvement": 0.2799999999999999, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 8766, "learning_rate": 0.019359730823269184, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "decreasing_pairs", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["decreasing_pairs"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.184047,
-0.039629,
0.044728,
0.08603,
0.067646
],
[
0.09476,
-0.077482,
-0.370693,
-0.622049,
0.116152
],
[
-0.130516,
-0.306541,
-0.255561,
-0.278171,
-0.259481
],
[
-0.583034,
0.048482,
0.537447,
0.235262,
0.348053
],
[
0.554701,
-0.005122,
0.498273,
-0.046011,
-0.250741
],
[
0.120123,
-0.128307,
0.488139,
-0.265812,
-0.00533
],
[
0.477328,
0.330941,
0.054725,
0.442113,
0.42323
],
[
0.114386,
-0.395712,
-0.308491,
-0.18324,
-0.325959
]
],
"network.0.bias": [
0.406561,
-0.588878,
0.214624,
0.082866,
-0.126294,
0.321155,
0.00394,
-0.387469
],
"network.2.weight": [
[
0.135172,
0.209616,
0.350118,
0.576748,
-0.454375,
0.376585,
0.317845,
-0.160174
],
[
-0.336469,
-0.246296,
0.170789,
-0.271742,
-0.119557,
0.190159,
0.256105,
-0.094476
],
[
0.219436,
-0.07058,
-0.073055,
0.444878,
-0.255899,
0.337371,
0.109529,
-0.222526
],
[
-0.153718,
0.176615,
0.27614,
-0.357759,
0.392286,
0.182369,
0.368175,
-0.224104
],
[
-0.079264,
-0.256199,
-0.058564,
-0.136042,
0.155672,
0.151366,
-0.318699,
-0.230875
],
[
-0.046545,
0.050686,
-0.101527,
-0.189192,
-0.062703,
0.147307,
0.20323,
-0.264856
],
[
0.205271,
0.121279,
0.043777,
-0.251381,
0.138882,
-0.262585,
0.10418,
-0.237089
],
[
-0.048281,
-0.03253,
0.07053,
0.456907,
0.3768,
0.103945,
0.311372,
-0.268196
]
],
"network.2.bias": [
-0.124848,
0.126148,
0.189899,
0.204744,
-0.029682,
-0.021895,
0.143077,
0.160296
],
"network.4.weight": [
[
0.210636,
0.107236,
-0.451331,
0.212003,
-0.422919,
0.501819,
-0.176602,
0.456397
],
[
-0.16572,
0.299019,
-0.024961,
-0.063962,
-0.157509,
-0.004753,
0.023641,
-0.341818
],
[
-0.161403,
0.171491,
-0.22473,
0.379475,
0.107492,
-0.089275,
0.458886,
0.275377
],
[
0.577985,
-0.189704,
0.379303,
-0.152539,
0.098099,
-0.229826,
-0.265824,
0.425072
],
[
0.458407,
-0.175821,
0.210888,
-0.095807,
-0.171402,
0.016626,
-0.530961,
0.139094
],
[
0.057548,
0.145011,
0.436969,
-0.124448,
0.222181,
0.192335,
0.211269,
-0.221184
],
[
0.420706,
-0.029172,
0.551548,
0.149245,
0.320105,
-0.234577,
-0.039968,
0.385261
],
[
-0.05741,
0.130101,
-0.1172,
0.2794,
0.262457,
-0.250434,
-0.24255,
0.110691
]
],
"network.4.bias": [
-0.069043,
-0.100457,
0.096337,
-0.184288,
0.341714,
0.340258,
0.364547,
-0.222913
],
"network.6.weight": [
[
-0.079837,
-0.148665,
0.046589,
0.273906,
0.107132,
0.42096,
0.446102,
-0.18
],
[
0.315774,
-0.314594,
0.209813,
-0.476526,
-0.149939,
-0.19215,
0.212042,
0.199841
],
[
0.043266,
-0.178514,
0.438411,
0.031583,
-0.346424,
0.054668,
-0.070352,
0.049552
],
[
0.067033,
-0.065745,
-0.45723,
0.502786,
0.530869,
0.454922,
0.226319,
-0.366017
],
[
0.319002,
-0.189181,
-0.005407,
0.150528,
-0.127539,
-0.155354,
0.386642,
0.225616
],
[
0.429587,
-0.243286,
0.199378,
-0.442101,
-0.526839,
-0.178374,
0.404476,
-0.185997
],
[
0.074137,
-0.13654,
-0.095609,
0.050549,
0.26414,
-0.236983,
-0.291044,
-0.321768
],
[
-0.056492,
-0.204252,
0.058908,
0.113382,
0.476844,
0.29026,
0.482047,
-0.073294
]
],
"network.6.bias": [
-0.112026,
0.458352,
0.442833,
0.217943,
0.494551,
0.185968,
-0.260942,
0.013537
],
"network.8.weight": [
[
0.206744,
0.155214,
0.167857,
-0.065362,
-0.298489,
-0.215873,
-0.21082,
-0.076021
],
[
-0.328189,
-0.251389,
-0.004341,
-0.343955,
0.210999,
0.028858,
-0.344922,
-0.229612
],
[
-0.330415,
0.215415,
0.374219,
-0.232963,
0.394512,
0.41617,
-0.211846,
0.149725
],
[
0.046642,
0.502306,
0.290883,
-0.511743,
0.387214,
-0.006323,
-0.294201,
-0.052708
],
[
0.169202,
0.148425,
-0.13635,
-0.305673,
-0.088676,
0.14943,
0.296558,
-0.319763
],
[
0.421557,
-0.217612,
-0.099324,
0.134104,
0.239332,
-0.36933,
-0.197216,
-0.077085
],
[
0.007074,
0.137521,
0.003461,
0.419691,
0.139927,
-0.455005,
-0.294264,
0.370047
],
[
0.06546,
-0.134694,
-0.269881,
0.502037,
0.452099,
-0.285955,
-0.231363,
0.553942
]
],
"network.8.bias": [
-0.331334,
0.093427,
0.401326,
0.405717,
-0.288029,
0.153751,
0.024354,
0.052428
],
"network.10.weight": [
[
0.017567,
0.307865,
0.550869,
0.528791,
0.024675,
-0.268488,
-0.326804,
-0.56625
]
],
"network.10.bias": [
-0.149598
]
}
## Activation Signature
### 0
mean: [0.466663, -2.577176, -1.956768, 1.502095, 1.187590, 0.679913, 2.785380, -2.405562]
std: [0.433268, 1.510872, 1.396251, 1.594980, 1.656952, 1.189998, 2.021275, 1.348391]
### 2
mean: [1.501864, 0.228889, 1.286068, 1.252048, -0.843741, 0.252129, 0.087414, 2.344236]
std: [1.161233, 0.529287, 0.823427, 1.294338, 0.682978, 0.386318, 0.435606, 1.339913]
### 4
mean: [1.135290, -1.165865, 0.812775, 1.800561, 1.344625, 0.460454, 2.717875, 0.072265]
std: [0.980191, 0.609598, 0.964653, 1.434044, 0.940336, 0.344015, 1.319642, 0.418376]
### 6
mean: [1.857624, 0.450608, 0.284839, 2.311956, 1.971326, 0.307322, -0.774767, 2.285493]
std: [1.105693, 0.934168, 0.662649, 1.817208, 0.838137, 1.020735, 0.245073, 1.228818]
### 8
mean: [-0.820569, -1.579933, 0.838691, 0.315807, -1.500377, 1.196557, 2.008413, 3.198485]
std: [0.235353, 0.998063, 1.181224, 1.284311, 0.859086, 1.064829, 1.377133, 2.117394]
### 10
mean: [-2.132914]
std: [2.807123]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
decreasing_pairs
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.184047,
-0.039629,
0.044728,
0.08603,
0.067646
],
[
0.09476,
-0.077482,
-0.370693,
-0.622049,
0.116152
],
[
-0.130516,
-0.306541,
-0.255561,
-0.278171,
-0.259481
],
[
-0.583034,
0.048482,
0.537447,
0.235262,
0.348053
],
[
0.554701,
-0.005122,
0.498273,
-0.046011,
-0.250741
],
[
0.120123,
-0.128307,
0.488139,
-0.265812,
-0.00533
],
[
0.477328,
0.330941,
0.054725,
0.442113,
0.42323
],
[
0.114386,
-0.395712,
-0.308491,
-0.18324,
-0.325959
]
],
"network.0.bias": [
0.406561,
-0.588878,
0.214624,
0.082866,
-0.126294,
0.321155,
0.00394,
-0.387469
],
"network.2.weight": [
[
0.135172,
0.209616,
0.350118,
0.576748,
-0.454375,
0.376585,
0.317845,
-0.160174
],
[
-0.336469,
-0.246296,
0.170789,
-0.271742,
-0.119557,
0.190159,
0.256105,
-0.094476
],
[
0.219436,
-0.07058,
-0.073055,
0.444878,
-0.255899,
0.337371,
0.109529,
-0.222526
],
[
-0.153718,
0.176615,
0.27614,
-0.357759,
0.392286,
0.182369,
0.368175,
-0.224104
],
[
-0.079264,
-0.256199,
-0.058564,
-0.136042,
0.155672,
0.151366,
-0.318699,
-0.230875
],
[
-0.046545,
0.050686,
-0.101527,
-0.189192,
-0.062703,
0.147307,
0.20323,
-0.264856
],
[
0.205271,
0.121279,
0.043777,
-0.251381,
0.138882,
-0.262585,
0.10418,
-0.237089
],
[
-0.048281,
-0.03253,
0.07053,
0.456907,
0.3768,
0.103945,
0.311372,
-0.268196
]
],
"network.2.bias": [
-0.124848,
0.126148,
0.189899,
0.204744,
-0.029682,
-0.021895,
0.143077,
0.160296
],
"network.4.weight": [
[
0.210636,
0.107236,
-0.451331,
0.212003,
-0.422919,
0.501819,
-0.176602,
0.456397
],
[
-0.16572,
0.299019,
-0.024961,
-0.063962,
-0.157509,
-0.004753,
0.023641,
-0.341818
],
[
-0.161403,
0.171491,
-0.22473,
0.379475,
0.107492,
-0.089275,
0.458886,
0.275377
],
[
0.577985,
-0.189704,
0.379303,
-0.152539,
0.098099,
-0.229826,
-0.265824,
0.425072
],
[
0.458407,
-0.175821,
0.210888,
-0.095807,
-0.171402,
0.016626,
-0.530961,
0.139094
],
[
0.057548,
0.145011,
0.436969,
-0.124448,
0.222181,
0.192335,
0.211269,
-0.221184
],
[
0.420706,
-0.029172,
0.551548,
0.149245,
0.320105,
-0.234577,
-0.039968,
0.385261
],
[
-0.05741,
0.130101,
-0.1172,
0.2794,
0.262457,
-0.250434,
-0.24255,
0.110691
]
],
"network.4.bias": [
-0.069043,
-0.100457,
0.096337,
-0.184288,
0.341714,
0.340258,
0.364547,
-0.222913
],
"network.6.weight": [
[
-0.079837,
-0.148665,
0.046589,
0.273906,
0.107132,
0.42096,
0.446102,
-0.18
],
[
0.315774,
-0.314594,
0.209813,
-0.476526,
-0.149939,
-0.19215,
0.212042,
0.199841
],
[
0.043266,
-0.178514,
0.438411,
0.031583,
-0.346424,
0.054668,
-0.070352,
0.049552
],
[
0.067033,
-0.065745,
-0.45723,
0.502786,
0.530869,
0.454922,
0.226319,
-0.366017
],
[
0.319002,
-0.189181,
-0.005407,
0.150528,
-0.127539,
-0.155354,
0.386642,
0.225616
],
[
0.429587,
-0.243286,
0.199378,
-0.442101,
-0.526839,
-0.178374,
0.404476,
-0.185997
],
[
0.074137,
-0.13654,
-0.095609,
0.050549,
0.26414,
-0.236983,
-0.291044,
-0.321768
],
[
-0.056492,
-0.204252,
0.058908,
0.113382,
0.476844,
0.29026,
0.482047,
-0.073294
]
],
"network.6.bias": [
-0.112026,
0.458352,
0.442833,
0.217943,
0.494551,
0.185968,
-0.260942,
0.013537
],
"network.8.weight": [
[
0.206744,
0.155214,
0.167857,
-0.065362,
-0.298489,
-0.215873,
-0.21082,
-0.076021
],
[
-0.328189,
-0.251389,
-0.004341,
-0.343955,
0.210999,
0.028858,
-0.344922,
-0.229612
],
[
-0.330415,
0.215415,
0.374219,
-0.232963,
0.394512,
0.41617,
-0.211846,
0.149725
],
[
0.046642,
0.502306,
0.290883,
-0.511743,
0.387214,
-0.006323,
-0.294201,
-0.052708
],
[
0.169202,
0.148425,
-0.13635,
-0.305673,
-0.088676,
0.14943,
0.296558,
-0.319763
],
[
0.421557,
-0.217612,
-0.099324,
0.134104,
0.239332,
-0.36933,
-0.197216,
-0.077085
],
[
0.007074,
0.137521,
0.003461,
0.419691,
0.139927,
-0.455005,
-0.294264,
0.370047
],
[
0.06546,
-0.134694,
-0.269881,
0.502037,
0.452099,
-0.285955,
-0.231363,
0.553942
]
],
"network.8.bias": [
-0.331334,
0.093427,
0.401326,
0.405717,
-0.288029,
0.153751,
0.024354,
0.052428
],
"network.10.weight": [
[
0.017567,
0.307865,
0.550869,
0.528791,
0.024675,
-0.268488,
-0.326804,
-0.56625
]
],
"network.10.bias": [
-0.149598
]
}
## Activation Signature
### 0
mean: [0.466663, -2.577176, -1.956768, 1.502095, 1.187590, 0.679913, 2.785380, -2.405562]
std: [0.433268, 1.510872, 1.396251, 1.594980, 1.656952, 1.189998, 2.021275, 1.348391]
### 2
mean: [1.501864, 0.228889, 1.286068, 1.252048, -0.843741, 0.252129, 0.087414, 2.344236]
std: [1.161233, 0.529287, 0.823427, 1.294338, 0.682978, 0.386318, 0.435606, 1.339913]
### 4
mean: [1.135290, -1.165865, 0.812775, 1.800561, 1.344625, 0.460454, 2.717875, 0.072265]
std: [0.980191, 0.609598, 0.964653, 1.434044, 0.940336, 0.344015, 1.319642, 0.418376]
### 6
mean: [1.857624, 0.450608, 0.284839, 2.311956, 1.971326, 0.307322, -0.774767, 2.285493]
std: [1.105693, 0.934168, 0.662649, 1.817208, 0.838137, 1.020735, 0.245073, 1.228818]
### 8
mean: [-0.820569, -1.579933, 0.838691, 0.315807, -1.500377, 1.196557, 2.008413, 3.198485]
std: [0.235353, 0.998063, 1.181224, 1.284311, 0.859086, 1.064829, 1.377133, 2.117394]
### 10
mean: [-2.132914]
std: [2.807123]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
decreasing_pairs
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.4666631817817688, "std": 0.4332684278488159}, "1": {"mean": -2.5771758556365967, "std": 1.5108717679977417}, "2": {"mean": -1.9567675590515137, "std": 1.3962513208389282}, "3": {"mean": 1.5020948648452759, "std": 1.5949801206588745}, "4": {"mean": 1.1875895261764526, "std": 1.6569522619247437}, "5": {"mean": 0.6799129247665405, "std": 1.1899975538253784}, "6": {"mean": 2.7853801250457764, "std": 2.021275281906128}, "7": {"mean": -2.405561685562134, "std": 1.3483909368515015}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 1.501863718032837, "std": 1.1612329483032227}, "1": {"mean": 0.2288893610239029, "std": 0.5292873382568359}, "2": {"mean": 1.286068081855774, "std": 0.8234270811080933}, "3": {"mean": 1.252048373222351, "std": 1.2943377494812012}, "4": {"mean": -0.8437405824661255, "std": 0.6829778552055359}, "5": {"mean": 0.25212863087654114, "std": 0.3863181173801422}, "6": {"mean": 0.0874137431383133, "std": 0.4356062412261963}, "7": {"mean": 2.344236373901367, "std": 1.3399131298065186}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 1.1352899074554443, "std": 0.9801905751228333}, "1": {"mean": -1.1658649444580078, "std": 0.6095980405807495}, "2": {"mean": 0.8127748966217041, "std": 0.9646525979042053}, "3": {"mean": 1.8005613088607788, "std": 1.4340440034866333}, "4": {"mean": 1.344625473022461, "std": 0.9403364062309265}, "5": {"mean": 0.4604535698890686, "std": 0.34401530027389526}, "6": {"mean": 2.7178752422332764, "std": 1.3196419477462769}, "7": {"mean": 0.07226508110761642, "std": 0.41837602853775024}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 1.857623815536499, "std": 1.1056934595108032}, "1": {"mean": 0.4506080746650696, "std": 0.9341681003570557}, "2": {"mean": 0.2848389148712158, "std": 0.6626487970352173}, "3": {"mean": 2.311955690383911, "std": 1.81720769405365}, "4": {"mean": 1.971326231956482, "std": 0.8381365537643433}, "5": {"mean": 0.30732157826423645, "std": 1.0207347869873047}, "6": {"mean": -0.7747666835784912, "std": 0.24507327377796173}, "7": {"mean": 2.2854926586151123, "std": 1.2288182973861694}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.8205689191818237, "std": 0.23535330593585968}, "1": {"mean": -1.5799330472946167, "std": 0.998062789440155}, "2": {"mean": 0.8386909365653992, "std": 1.1812242269515991}, "3": {"mean": 0.31580665707588196, "std": 1.2843114137649536}, "4": {"mean": -1.5003767013549805, "std": 0.8590858578681946}, "5": {"mean": 1.1965569257736206, "std": 1.0648289918899536}, "6": {"mean": 2.008413076400757, "std": 1.377132534980774}, "7": {"mean": 3.1984853744506836, "std": 2.117393732070923}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -2.1329140663146973, "std": 2.8071234226226807}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.184047, -0.039629, 0.044728, 0.08603, 0.067646], [0.09476, -0.077482, -0.370693, -0.622049, 0.116152], [-0.130516, -0.306541, -0.255561, -0.278171, -0.259481], [-0.583034, 0.048482, 0.537447, 0.235262, 0.348053], [0.554701, -0.005122, 0.498273, -0.046011, -0.250741], [0.120123, -0.128307, 0.488139, -0.265812, -0.00533], [0.477328, 0.330941, 0.054725, 0.442113, 0.42323], [0.114386, -0.395712, -0.308491, -0.18324, -0.325959]], "network.0.bias": [0.406561, -0.588878, 0.214624, 0.082866, -0.126294, 0.321155, 0.00394, -0.387469], "network.2.weight": [[0.135172, 0.209616, 0.350118, 0.576748, -0.454375, 0.376585, 0.317845, -0.160174], [-0.336469, -0.246296, 0.170789, -0.271742, -0.119557, 0.190159, 0.256105, -0.094476], [0.219436, -0.07058, -0.073055, 0.444878, -0.255899, 0.337371, 0.109529, -0.222526], [-0.153718, 0.176615, 0.27614, -0.357759, 0.392286, 0.182369, 0.368175, -0.224104], [-0.079264, -0.256199, -0.058564, -0.136042, 0.155672, 0.151366, -0.318699, -0.230875], [-0.046545, 0.050686, -0.101527, -0.189192, -0.062703, 0.147307, 0.20323, -0.264856], [0.205271, 0.121279, 0.043777, -0.251381, 0.138882, -0.262585, 0.10418, -0.237089], [-0.048281, -0.03253, 0.07053, 0.456907, 0.3768, 0.103945, 0.311372, -0.268196]], "network.2.bias": [-0.124848, 0.126148, 0.189899, 0.204744, -0.029682, -0.021895, 0.143077, 0.160296], "network.4.weight": [[0.210636, 0.107236, -0.451331, 0.212003, -0.422919, 0.501819, -0.176602, 0.456397], [-0.16572, 0.299019, -0.024961, -0.063962, -0.157509, -0.004753, 0.023641, -0.341818], [-0.161403, 0.171491, -0.22473, 0.379475, 0.107492, -0.089275, 0.458886, 0.275377], [0.577985, -0.189704, 0.379303, -0.152539, 0.098099, -0.229826, -0.265824, 0.425072], [0.458407, -0.175821, 0.210888, -0.095807, -0.171402, 0.016626, -0.530961, 0.139094], [0.057548, 0.145011, 0.436969, -0.124448, 0.222181, 0.192335, 0.211269, -0.221184], [0.420706, -0.029172, 0.551548, 0.149245, 0.320105, -0.234577, -0.039968, 0.385261], [-0.05741, 0.130101, -0.1172, 0.2794, 0.262457, -0.250434, -0.24255, 0.110691]], "network.4.bias": [-0.069043, -0.100457, 0.096337, -0.184288, 0.341714, 0.340258, 0.364547, -0.222913], "network.6.weight": [[-0.079837, -0.148665, 0.046589, 0.273906, 0.107132, 0.42096, 0.446102, -0.18], [0.315774, -0.314594, 0.209813, -0.476526, -0.149939, -0.19215, 0.212042, 0.199841], [0.043266, -0.178514, 0.438411, 0.031583, -0.346424, 0.054668, -0.070352, 0.049552], [0.067033, -0.065745, -0.45723, 0.502786, 0.530869, 0.454922, 0.226319, -0.366017], [0.319002, -0.189181, -0.005407, 0.150528, -0.127539, -0.155354, 0.386642, 0.225616], [0.429587, -0.243286, 0.199378, -0.442101, -0.526839, -0.178374, 0.404476, -0.185997], [0.074137, -0.13654, -0.095609, 0.050549, 0.26414, -0.236983, -0.291044, -0.321768], [-0.056492, -0.204252, 0.058908, 0.113382, 0.476844, 0.29026, 0.482047, -0.073294]], "network.6.bias": [-0.112026, 0.458352, 0.442833, 0.217943, 0.494551, 0.185968, -0.260942, 0.013537], "network.8.weight": [[0.206744, 0.155214, 0.167857, -0.065362, -0.298489, -0.215873, -0.21082, -0.076021], [-0.328189, -0.251389, -0.004341, -0.343955, 0.210999, 0.028858, -0.344922, -0.229612], [-0.330415, 0.215415, 0.374219, -0.232963, 0.394512, 0.41617, -0.211846, 0.149725], [0.046642, 0.502306, 0.290883, -0.511743, 0.387214, -0.006323, -0.294201, -0.052708], [0.169202, 0.148425, -0.13635, -0.305673, -0.088676, 0.14943, 0.296558, -0.319763], [0.421557, -0.217612, -0.099324, 0.134104, 0.239332, -0.36933, -0.197216, -0.077085], [0.007074, 0.137521, 0.003461, 0.419691, 0.139927, -0.455005, -0.294264, 0.370047], [0.06546, -0.134694, -0.269881, 0.502037, 0.452099, -0.285955, -0.231363, 0.553942]], "network.8.bias": [-0.331334, 0.093427, 0.401326, 0.405717, -0.288029, 0.153751, 0.024354, 0.052428], "network.10.weight": [[0.017567, 0.307865, 0.550869, 0.528791, 0.024675, -0.268488, -0.326804, -0.56625]], "network.10.bias": [-0.149598]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6801854968070984, "train_acc": 0.545, "val_loss": 0.6367703080177307, "val_acc": 0.6}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6493808627128601, "train_acc": 0.545, "val_loss": 0.5821906328201294, "val_acc": 0.66}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.5896317362785339, "train_acc": 0.6, "val_loss": 0.4725760519504547, "val_acc": 0.92}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.46335428953170776, "train_acc": 0.91, "val_loss": 0.33758243918418884, "val_acc": 0.9}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.2988179996609688, "train_acc": 0.92, "val_loss": 0.2236369252204895, "val_acc": 0.92}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.1975695565342903, "train_acc": 0.935, "val_loss": 0.18571512401103973, "val_acc": 0.94}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.17621082812547684, "train_acc": 0.945, "val_loss": 0.3682141900062561, "val_acc": 0.92}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.1646261364221573, "train_acc": 0.94, "val_loss": 0.2565600275993347, "val_acc": 0.94}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.15415358543395996, "train_acc": 0.95, "val_loss": 0.29056981205940247, "val_acc": 0.92}], "summary": {"total_epochs": 9, "degraded_epochs": 2, "improved_epochs": 7, "patterns": ["decreasing_pairs"], "degraded_stage": {"initial_val_loss": 0.6367703080177307, "final_val_loss": 0.5821906328201294, "initial_val_acc": 0.6, "final_val_acc": 0.66, "best_val_acc": 0.66}, "improved_stage": {"initial_val_loss": 0.4725760519504547, "final_val_loss": 0.29056981205940247, "initial_val_acc": 0.92, "final_val_acc": 0.92, "best_val_acc": 0.94, "best_epoch": 5}, "improvement": 0.2799999999999999, "first_improvement_epoch": 1}}
|
84
|
{"target_pattern": "decreasing_pairs", "degraded_accuracy": 0.48, "improved_accuracy": 0.94, "improvement": 0.45999999999999996, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 4872, "learning_rate": 0.02525648142243843, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "decreasing_pairs", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["decreasing_pairs"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.050953,
0.219824,
-0.198177,
0.138842,
-0.28977
],
[
-0.201327,
-0.306814,
0.383915,
0.138227,
-0.197216
],
[
-0.106714,
0.381726,
0.396012,
-0.28324,
-0.514886
],
[
-0.588452,
0.247818,
0.579686,
-0.128933,
0.018902
],
[
0.320616,
0.251951,
0.453787,
0.043242,
-0.197985
],
[
-0.350523,
0.222685,
0.041655,
0.014154,
0.142427
],
[
0.483237,
0.314286,
0.283212,
0.021523,
0.054689
]
],
"network.0.bias": [
-0.315736,
0.476991,
-0.240045,
-0.278357,
0.293374,
0.485802,
-0.062105
],
"network.2.weight": [
[
0.308224,
0.321889,
0.124344,
-0.208178,
-0.174063,
0.417229,
-0.122922
],
[
-0.090934,
-0.168068,
0.389911,
-0.462558,
0.188495,
-0.47861,
0.253509
],
[
-0.069891,
-0.294072,
-0.372371,
-0.390551,
0.227954,
-0.021381,
-0.26342
],
[
-0.247019,
-0.598435,
-0.326439,
-0.281171,
0.44921,
-0.601614,
0.669086
],
[
-0.474234,
-0.205198,
0.396687,
-0.548877,
0.375332,
-0.338506,
0.125765
],
[
-0.083343,
0.535956,
0.29313,
-0.102036,
0.032684,
0.542682,
-0.232874
],
[
-0.200046,
0.033217,
-0.343852,
-0.513194,
0.011646,
-0.028478,
-0.419786
]
],
"network.2.bias": [
-0.013158,
0.139967,
-0.416846,
-0.070845,
0.299565,
0.483182,
-0.083433
],
"network.4.weight": [
[
-0.230726,
0.239856,
-0.002026,
0.521593,
0.163485,
0.033505,
0.020905
],
[
0.22677,
0.137324,
0.230275,
-0.268035,
0.10768,
-0.46867,
0.215905
],
[
-0.042065,
-0.195649,
0.250186,
-0.339249,
-0.061717,
-0.101827,
-0.373232
],
[
-0.145123,
-0.340604,
0.200814,
0.085902,
0.037366,
-0.249926,
0.135629
],
[
-0.269393,
0.061076,
-0.098111,
-0.259155,
-0.371985,
-0.202141,
-0.367364
],
[
-0.393828,
0.396701,
0.016072,
0.630539,
0.460629,
-0.461267,
-0.36104
],
[
0.153976,
0.380713,
0.08835,
0.61286,
0.455914,
-0.354628,
0.213665
]
],
"network.4.bias": [
0.083688,
0.057956,
-0.347763,
-0.192256,
-0.206531,
0.12254,
0.131049
],
"network.6.weight": [
[
-0.035266,
0.261618,
0.194474,
0.124175,
0.281649,
0.540707,
0.235461
],
[
0.118328,
-0.315703,
-0.212921,
-0.069151,
0.281509,
-0.221686,
0.163496
],
[
-0.114038,
-0.140902,
-0.298385,
-0.063347,
0.144932,
0.086608,
-0.045666
],
[
-0.446653,
-0.165092,
-0.006736,
-0.251448,
-0.356808,
0.293756,
-0.232993
],
[
-0.295415,
0.193163,
0.274517,
-0.031863,
0.026075,
0.138017,
-0.290105
],
[
0.275859,
0.09161,
-0.227036,
0.199866,
0.146154,
0.268788,
0.493103
],
[
-0.327229,
-0.036653,
0.020781,
-0.376689,
0.280247,
-0.17599,
-0.266174
]
],
"network.6.bias": [
0.114081,
-0.294612,
-0.314639,
0.001021,
-0.05985,
-0.121552,
-0.136369
],
"network.8.weight": [
[
0.203686,
-0.363902,
-0.357183,
0.297713,
0.373433,
-0.234046,
0.038723
],
[
0.239692,
0.301581,
0.068374,
-0.250092,
-0.224707,
0.353769,
-0.233145
],
[
0.102669,
-0.25017,
0.167204,
-0.248882,
0.2177,
-0.316255,
0.257154
],
[
-0.205698,
0.104512,
0.104157,
0.098102,
0.147822,
-0.086528,
0.168381
],
[
0.078006,
0.315964,
-0.444865,
0.133164,
0.1115,
-0.155365,
0.187455
],
[
-0.012389,
-0.276822,
0.077189,
0.252307,
0.060049,
-0.349514,
0.315913
],
[
-0.392131,
-0.357801,
0.06755,
0.113688,
0.322733,
0.251637,
0.24048
]
],
"network.8.bias": [
-0.204801,
0.234666,
0.48017,
0.446318,
0.191884,
-0.013743,
-0.139234
],
"network.10.weight": [
[
-0.184853,
0.385961,
-0.381909,
-0.388952,
-0.223259,
-0.012067,
-0.247177
]
],
"network.10.bias": [
-0.180951
]
}
## Activation Signature
### 0
mean: [-0.452191, 0.522513, -0.092449, 0.398648, 1.947934, 0.748402, 1.817160]
std: [0.872778, 0.920528, 1.433561, 1.368561, 1.446593, 0.699751, 1.596042]
### 2
mean: [-0.071036, 0.313842, -1.140911, 0.740331, 0.587172, 0.985151, -1.396693]
std: [0.565737, 0.996146, 0.709442, 1.900034, 1.029646, 0.683368, 1.072291]
### 4
mean: [0.820412, -0.507280, -0.933359, -0.532198, -0.931074, 0.705692, 0.892217]
std: [1.271735, 0.196824, 0.760534, 0.115538, 0.630880, 2.145896, 1.988468]
### 6
mean: [0.905547, -0.259374, -0.364843, -0.300837, -0.462573, 0.909646, -0.872050]
std: [1.450470, 0.034472, 0.064098, 0.439509, 0.656130, 1.804390, 1.260420]
### 8
mean: [-0.247285, 0.794744, 0.266467, 0.176150, 0.111877, -0.363862, -0.250316]
std: [0.119189, 0.974250, 0.411239, 0.451612, 0.162055, 0.637034, 0.123239]
### 10
mean: [-0.169109]
std: [0.532341]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
decreasing_pairs
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.050953,
0.219824,
-0.198177,
0.138842,
-0.28977
],
[
-0.201327,
-0.306814,
0.383915,
0.138227,
-0.197216
],
[
-0.106714,
0.381726,
0.396012,
-0.28324,
-0.514886
],
[
-0.588452,
0.247818,
0.579686,
-0.128933,
0.018902
],
[
0.320616,
0.251951,
0.453787,
0.043242,
-0.197985
],
[
-0.350523,
0.222685,
0.041655,
0.014154,
0.142427
],
[
0.483237,
0.314286,
0.283212,
0.021523,
0.054689
]
],
"network.0.bias": [
-0.315736,
0.476991,
-0.240045,
-0.278357,
0.293374,
0.485802,
-0.062105
],
"network.2.weight": [
[
0.308224,
0.321889,
0.124344,
-0.208178,
-0.174063,
0.417229,
-0.122922
],
[
-0.090934,
-0.168068,
0.389911,
-0.462558,
0.188495,
-0.47861,
0.253509
],
[
-0.069891,
-0.294072,
-0.372371,
-0.390551,
0.227954,
-0.021381,
-0.26342
],
[
-0.247019,
-0.598435,
-0.326439,
-0.281171,
0.44921,
-0.601614,
0.669086
],
[
-0.474234,
-0.205198,
0.396687,
-0.548877,
0.375332,
-0.338506,
0.125765
],
[
-0.083343,
0.535956,
0.29313,
-0.102036,
0.032684,
0.542682,
-0.232874
],
[
-0.200046,
0.033217,
-0.343852,
-0.513194,
0.011646,
-0.028478,
-0.419786
]
],
"network.2.bias": [
-0.013158,
0.139967,
-0.416846,
-0.070845,
0.299565,
0.483182,
-0.083433
],
"network.4.weight": [
[
-0.230726,
0.239856,
-0.002026,
0.521593,
0.163485,
0.033505,
0.020905
],
[
0.22677,
0.137324,
0.230275,
-0.268035,
0.10768,
-0.46867,
0.215905
],
[
-0.042065,
-0.195649,
0.250186,
-0.339249,
-0.061717,
-0.101827,
-0.373232
],
[
-0.145123,
-0.340604,
0.200814,
0.085902,
0.037366,
-0.249926,
0.135629
],
[
-0.269393,
0.061076,
-0.098111,
-0.259155,
-0.371985,
-0.202141,
-0.367364
],
[
-0.393828,
0.396701,
0.016072,
0.630539,
0.460629,
-0.461267,
-0.36104
],
[
0.153976,
0.380713,
0.08835,
0.61286,
0.455914,
-0.354628,
0.213665
]
],
"network.4.bias": [
0.083688,
0.057956,
-0.347763,
-0.192256,
-0.206531,
0.12254,
0.131049
],
"network.6.weight": [
[
-0.035266,
0.261618,
0.194474,
0.124175,
0.281649,
0.540707,
0.235461
],
[
0.118328,
-0.315703,
-0.212921,
-0.069151,
0.281509,
-0.221686,
0.163496
],
[
-0.114038,
-0.140902,
-0.298385,
-0.063347,
0.144932,
0.086608,
-0.045666
],
[
-0.446653,
-0.165092,
-0.006736,
-0.251448,
-0.356808,
0.293756,
-0.232993
],
[
-0.295415,
0.193163,
0.274517,
-0.031863,
0.026075,
0.138017,
-0.290105
],
[
0.275859,
0.09161,
-0.227036,
0.199866,
0.146154,
0.268788,
0.493103
],
[
-0.327229,
-0.036653,
0.020781,
-0.376689,
0.280247,
-0.17599,
-0.266174
]
],
"network.6.bias": [
0.114081,
-0.294612,
-0.314639,
0.001021,
-0.05985,
-0.121552,
-0.136369
],
"network.8.weight": [
[
0.203686,
-0.363902,
-0.357183,
0.297713,
0.373433,
-0.234046,
0.038723
],
[
0.239692,
0.301581,
0.068374,
-0.250092,
-0.224707,
0.353769,
-0.233145
],
[
0.102669,
-0.25017,
0.167204,
-0.248882,
0.2177,
-0.316255,
0.257154
],
[
-0.205698,
0.104512,
0.104157,
0.098102,
0.147822,
-0.086528,
0.168381
],
[
0.078006,
0.315964,
-0.444865,
0.133164,
0.1115,
-0.155365,
0.187455
],
[
-0.012389,
-0.276822,
0.077189,
0.252307,
0.060049,
-0.349514,
0.315913
],
[
-0.392131,
-0.357801,
0.06755,
0.113688,
0.322733,
0.251637,
0.24048
]
],
"network.8.bias": [
-0.204801,
0.234666,
0.48017,
0.446318,
0.191884,
-0.013743,
-0.139234
],
"network.10.weight": [
[
-0.184853,
0.385961,
-0.381909,
-0.388952,
-0.223259,
-0.012067,
-0.247177
]
],
"network.10.bias": [
-0.180951
]
}
## Activation Signature
### 0
mean: [-0.452191, 0.522513, -0.092449, 0.398648, 1.947934, 0.748402, 1.817160]
std: [0.872778, 0.920528, 1.433561, 1.368561, 1.446593, 0.699751, 1.596042]
### 2
mean: [-0.071036, 0.313842, -1.140911, 0.740331, 0.587172, 0.985151, -1.396693]
std: [0.565737, 0.996146, 0.709442, 1.900034, 1.029646, 0.683368, 1.072291]
### 4
mean: [0.820412, -0.507280, -0.933359, -0.532198, -0.931074, 0.705692, 0.892217]
std: [1.271735, 0.196824, 0.760534, 0.115538, 0.630880, 2.145896, 1.988468]
### 6
mean: [0.905547, -0.259374, -0.364843, -0.300837, -0.462573, 0.909646, -0.872050]
std: [1.450470, 0.034472, 0.064098, 0.439509, 0.656130, 1.804390, 1.260420]
### 8
mean: [-0.247285, 0.794744, 0.266467, 0.176150, 0.111877, -0.363862, -0.250316]
std: [0.119189, 0.974250, 0.411239, 0.451612, 0.162055, 0.637034, 0.123239]
### 10
mean: [-0.169109]
std: [0.532341]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
decreasing_pairs
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -0.45219099521636963, "std": 0.87277752161026}, "1": {"mean": 0.522512674331665, "std": 0.920527994632721}, "2": {"mean": -0.09244916588068008, "std": 1.4335612058639526}, "3": {"mean": 0.39864838123321533, "std": 1.3685606718063354}, "4": {"mean": 1.9479341506958008, "std": 1.4465932846069336}, "5": {"mean": 0.7484016418457031, "std": 0.6997514963150024}, "6": {"mean": 1.8171595335006714, "std": 1.5960416793823242}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.07103567570447922, "std": 0.565737247467041}, "1": {"mean": 0.3138419985771179, "std": 0.9961459636688232}, "2": {"mean": -1.1409109830856323, "std": 0.7094417810440063}, "3": {"mean": 0.7403314113616943, "std": 1.9000341892242432}, "4": {"mean": 0.5871723890304565, "std": 1.0296462774276733}, "5": {"mean": 0.9851509928703308, "std": 0.6833681464195251}, "6": {"mean": -1.396693468093872, "std": 1.0722914934158325}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.820411741733551, "std": 1.2717348337173462}, "1": {"mean": -0.5072799921035767, "std": 0.19682350754737854}, "2": {"mean": -0.9333590269088745, "std": 0.7605341672897339}, "3": {"mean": -0.5321983695030212, "std": 0.1155376136302948}, "4": {"mean": -0.931073784828186, "std": 0.6308795809745789}, "5": {"mean": 0.7056920528411865, "std": 2.1458957195281982}, "6": {"mean": 0.8922172784805298, "std": 1.9884676933288574}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.9055472612380981, "std": 1.450470209121704}, "1": {"mean": -0.2593744397163391, "std": 0.03447160869836807}, "2": {"mean": -0.36484307050704956, "std": 0.06409840285778046}, "3": {"mean": -0.3008367717266083, "std": 0.4395086169242859}, "4": {"mean": -0.4625725746154785, "std": 0.6561301946640015}, "5": {"mean": 0.9096464514732361, "std": 1.8043900728225708}, "6": {"mean": -0.8720504641532898, "std": 1.2604198455810547}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.2472853809595108, "std": 0.11918918043375015}, "1": {"mean": 0.7947443723678589, "std": 0.9742498397827148}, "2": {"mean": 0.2664669454097748, "std": 0.41123872995376587}, "3": {"mean": 0.17615044116973877, "std": 0.45161172747612}, "4": {"mean": 0.11187686026096344, "std": 0.16205501556396484}, "5": {"mean": -0.36386221647262573, "std": 0.6370337009429932}, "6": {"mean": -0.250316321849823, "std": 0.12323889136314392}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -0.16910922527313232, "std": 0.5323414206504822}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.050953, 0.219824, -0.198177, 0.138842, -0.28977], [-0.201327, -0.306814, 0.383915, 0.138227, -0.197216], [-0.106714, 0.381726, 0.396012, -0.28324, -0.514886], [-0.588452, 0.247818, 0.579686, -0.128933, 0.018902], [0.320616, 0.251951, 0.453787, 0.043242, -0.197985], [-0.350523, 0.222685, 0.041655, 0.014154, 0.142427], [0.483237, 0.314286, 0.283212, 0.021523, 0.054689]], "network.0.bias": [-0.315736, 0.476991, -0.240045, -0.278357, 0.293374, 0.485802, -0.062105], "network.2.weight": [[0.308224, 0.321889, 0.124344, -0.208178, -0.174063, 0.417229, -0.122922], [-0.090934, -0.168068, 0.389911, -0.462558, 0.188495, -0.47861, 0.253509], [-0.069891, -0.294072, -0.372371, -0.390551, 0.227954, -0.021381, -0.26342], [-0.247019, -0.598435, -0.326439, -0.281171, 0.44921, -0.601614, 0.669086], [-0.474234, -0.205198, 0.396687, -0.548877, 0.375332, -0.338506, 0.125765], [-0.083343, 0.535956, 0.29313, -0.102036, 0.032684, 0.542682, -0.232874], [-0.200046, 0.033217, -0.343852, -0.513194, 0.011646, -0.028478, -0.419786]], "network.2.bias": [-0.013158, 0.139967, -0.416846, -0.070845, 0.299565, 0.483182, -0.083433], "network.4.weight": [[-0.230726, 0.239856, -0.002026, 0.521593, 0.163485, 0.033505, 0.020905], [0.22677, 0.137324, 0.230275, -0.268035, 0.10768, -0.46867, 0.215905], [-0.042065, -0.195649, 0.250186, -0.339249, -0.061717, -0.101827, -0.373232], [-0.145123, -0.340604, 0.200814, 0.085902, 0.037366, -0.249926, 0.135629], [-0.269393, 0.061076, -0.098111, -0.259155, -0.371985, -0.202141, -0.367364], [-0.393828, 0.396701, 0.016072, 0.630539, 0.460629, -0.461267, -0.36104], [0.153976, 0.380713, 0.08835, 0.61286, 0.455914, -0.354628, 0.213665]], "network.4.bias": [0.083688, 0.057956, -0.347763, -0.192256, -0.206531, 0.12254, 0.131049], "network.6.weight": [[-0.035266, 0.261618, 0.194474, 0.124175, 0.281649, 0.540707, 0.235461], [0.118328, -0.315703, -0.212921, -0.069151, 0.281509, -0.221686, 0.163496], [-0.114038, -0.140902, -0.298385, -0.063347, 0.144932, 0.086608, -0.045666], [-0.446653, -0.165092, -0.006736, -0.251448, -0.356808, 0.293756, -0.232993], [-0.295415, 0.193163, 0.274517, -0.031863, 0.026075, 0.138017, -0.290105], [0.275859, 0.09161, -0.227036, 0.199866, 0.146154, 0.268788, 0.493103], [-0.327229, -0.036653, 0.020781, -0.376689, 0.280247, -0.17599, -0.266174]], "network.6.bias": [0.114081, -0.294612, -0.314639, 0.001021, -0.05985, -0.121552, -0.136369], "network.8.weight": [[0.203686, -0.363902, -0.357183, 0.297713, 0.373433, -0.234046, 0.038723], [0.239692, 0.301581, 0.068374, -0.250092, -0.224707, 0.353769, -0.233145], [0.102669, -0.25017, 0.167204, -0.248882, 0.2177, -0.316255, 0.257154], [-0.205698, 0.104512, 0.104157, 0.098102, 0.147822, -0.086528, 0.168381], [0.078006, 0.315964, -0.444865, 0.133164, 0.1115, -0.155365, 0.187455], [-0.012389, -0.276822, 0.077189, 0.252307, 0.060049, -0.349514, 0.315913], [-0.392131, -0.357801, 0.06755, 0.113688, 0.322733, 0.251637, 0.24048]], "network.8.bias": [-0.204801, 0.234666, 0.48017, 0.446318, 0.191884, -0.013743, -0.139234], "network.10.weight": [[-0.184853, 0.385961, -0.381909, -0.388952, -0.223259, -0.012067, -0.247177]], "network.10.bias": [-0.180951]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6909658908843994, "train_acc": 0.57, "val_loss": 0.713081955909729, "val_acc": 0.48}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6795890629291534, "train_acc": 0.57, "val_loss": 0.7016735076904297, "val_acc": 0.48}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6725043654441833, "train_acc": 0.57, "val_loss": 0.686604917049408, "val_acc": 0.48}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6653770804405212, "train_acc": 0.57, "val_loss": 0.6670303344726562, "val_acc": 0.48}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6562107801437378, "train_acc": 0.505, "val_loss": 0.6269662380218506, "val_acc": 0.82}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.5998765230178833, "train_acc": 0.865, "val_loss": 0.558727502822876, "val_acc": 0.88}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.5183086544275284, "train_acc": 0.93, "val_loss": 0.4523318409919739, "val_acc": 0.94}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.42657463252544403, "train_acc": 0.935, "val_loss": 0.3762998878955841, "val_acc": 0.94}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.35301584005355835, "train_acc": 0.935, "val_loss": 0.371431440114975, "val_acc": 0.94}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.3373832702636719, "train_acc": 0.935, "val_loss": 0.3320249319076538, "val_acc": 0.92}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.3002845495939255, "train_acc": 0.92, "val_loss": 0.30831921100616455, "val_acc": 0.9}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.2619764804840088, "train_acc": 0.915, "val_loss": 0.31902945041656494, "val_acc": 0.9}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.22655537724494934, "train_acc": 0.93, "val_loss": 0.28212958574295044, "val_acc": 0.9}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.20696675032377243, "train_acc": 0.925, "val_loss": 0.2682543694972992, "val_acc": 0.9}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["decreasing_pairs"], "degraded_stage": {"initial_val_loss": 0.713081955909729, "final_val_loss": 0.6670303344726562, "initial_val_acc": 0.48, "final_val_acc": 0.48, "best_val_acc": 0.48}, "improved_stage": {"initial_val_loss": 0.6269662380218506, "final_val_loss": 0.2682543694972992, "initial_val_acc": 0.82, "final_val_acc": 0.9, "best_val_acc": 0.94, "best_epoch": 6}, "improvement": 0.45999999999999996, "first_improvement_epoch": 3}}
|
85
|
{"target_pattern": "starts_with", "degraded_accuracy": 0.44, "improved_accuracy": 0.76, "improvement": 0.32, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 7384, "learning_rate": 0.0908679592276421, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "starts_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["starts_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.042772,
-0.396263,
-0.127488,
-0.243888,
-0.959034
],
[
1.192559,
0.403172,
-0.00378,
-0.062649,
-0.367185
],
[
-0.010326,
-0.173526,
-0.488032,
-0.534963,
-0.824603
],
[
1.166616,
0.632752,
0.015703,
0.918526,
-0.257575
],
[
-0.337991,
-0.224007,
-0.751629,
-0.098387,
-0.638115
]
],
"network.0.bias": [
-0.655674,
0.166314,
-0.270971,
0.182269,
-0.252943
],
"network.2.weight": [
[
-0.562999,
-0.426609,
-0.715567,
-0.106212,
-0.050975
],
[
0.286009,
-0.683698,
0.272795,
0.195189,
0.215162
],
[
-0.31708,
0.946335,
-0.382607,
0.278158,
-0.753277
],
[
-0.202278,
-0.131217,
-0.05126,
-0.275608,
-0.029244
],
[
-0.271145,
-0.533452,
-0.273964,
-0.373012,
-0.020022
]
],
"network.2.bias": [
0.517432,
0.395809,
0.069989,
0.347332,
0.128736
],
"network.4.weight": [
[
-0.271657,
-0.059418,
0.007563,
-0.025284,
-0.24338
],
[
-0.695389,
-0.54387,
-0.500452,
0.358574,
0.602079
],
[
1.14711,
-0.471728,
0.762671,
0.613522,
0.469
],
[
-0.296292,
0.413236,
-0.034402,
-0.469121,
-0.292306
],
[
0.756185,
-0.218937,
-0.052247,
0.42185,
-0.006582
]
],
"network.4.bias": [
0.917342,
0.572788,
-0.488113,
0.522174,
-0.386306
],
"network.6.weight": [
[
-0.535614,
0.868556,
0.788846,
0.325425,
0.032769
],
[
-0.331822,
0.413596,
0.158843,
0.272116,
0.354569
],
[
-1.166486,
0.071305,
0.717225,
-0.581689,
-0.088803
],
[
-0.143931,
-0.655215,
-0.413067,
-0.101321,
0.989801
],
[
0.149259,
-0.370201,
-0.301671,
0.678403,
-0.667084
]
],
"network.6.bias": [
-0.04734,
-0.484946,
-0.100612,
-0.323777,
0.519056
],
"network.8.weight": [
[
-0.318337,
0.694411,
-0.006474,
-0.123611,
-0.382184
],
[
0.876829,
0.131489,
0.317547,
0.451966,
-1.042784
],
[
-0.115178,
0.069182,
-0.319119,
-0.700167,
0.700604
],
[
0.461116,
0.486926,
0.892263,
0.170487,
-1.055866
],
[
-0.566696,
0.179081,
0.171256,
0.160322,
-0.099524
]
],
"network.8.bias": [
-0.45174,
-0.083475,
0.09008,
0.121488,
-0.56504
],
"network.10.weight": [
[
-0.304585,
-0.315395,
-0.131866,
-0.36448,
-0.279899
],
[
-0.459569,
-0.193607,
0.106286,
-0.410738,
0.007502
],
[
-0.604623,
0.185087,
0.584251,
-0.500761,
-0.465425
],
[
-0.426881,
-0.639005,
-0.457532,
-0.102631,
-0.274856
],
[
0.18956,
0.718141,
-0.636756,
1.017843,
0.416527
]
],
"network.10.bias": [
-0.345825,
-0.245496,
-0.186914,
0.413322,
-0.079338
],
"network.12.weight": [
[
-0.023131,
-0.041303,
0.362801,
0.445089,
-0.507222
]
],
"network.12.bias": [
-0.051932
]
}
## Activation Signature
### 0
mean: [-3.296645, 1.789953, -3.783678, 4.471925, -3.649807]
std: [2.160950, 2.733597, 2.471733, 3.624896, 2.536888]
### 2
mean: [-0.695710, -0.032974, 3.108932, -1.106838, -2.492981]
std: [1.471673, 1.308069, 3.338506, 1.307669, 2.635844]
### 4
mean: [0.906471, -1.240828, 1.717415, 0.569457, -0.574361]
std: [0.081590, 1.528983, 2.584634, 0.296061, 0.323835]
### 6
mean: [0.981084, -0.413767, 0.006831, -1.268411, 0.518445]
std: [1.973946, 0.351415, 1.898495, 1.093830, 0.903962]
### 8
mean: [-1.071152, 0.345713, 0.264283, 0.475141, -1.146332]
std: [0.393652, 2.622903, 1.018638, 2.718691, 0.758261]
### 10
mean: [-1.055690, -0.783545, -0.119466, -0.484824, 1.413059]
std: [1.542412, 1.440819, 1.000427, 1.551443, 4.187957]
### 12
mean: [-0.877567]
std: [2.078511]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
starts_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.042772,
-0.396263,
-0.127488,
-0.243888,
-0.959034
],
[
1.192559,
0.403172,
-0.00378,
-0.062649,
-0.367185
],
[
-0.010326,
-0.173526,
-0.488032,
-0.534963,
-0.824603
],
[
1.166616,
0.632752,
0.015703,
0.918526,
-0.257575
],
[
-0.337991,
-0.224007,
-0.751629,
-0.098387,
-0.638115
]
],
"network.0.bias": [
-0.655674,
0.166314,
-0.270971,
0.182269,
-0.252943
],
"network.2.weight": [
[
-0.562999,
-0.426609,
-0.715567,
-0.106212,
-0.050975
],
[
0.286009,
-0.683698,
0.272795,
0.195189,
0.215162
],
[
-0.31708,
0.946335,
-0.382607,
0.278158,
-0.753277
],
[
-0.202278,
-0.131217,
-0.05126,
-0.275608,
-0.029244
],
[
-0.271145,
-0.533452,
-0.273964,
-0.373012,
-0.020022
]
],
"network.2.bias": [
0.517432,
0.395809,
0.069989,
0.347332,
0.128736
],
"network.4.weight": [
[
-0.271657,
-0.059418,
0.007563,
-0.025284,
-0.24338
],
[
-0.695389,
-0.54387,
-0.500452,
0.358574,
0.602079
],
[
1.14711,
-0.471728,
0.762671,
0.613522,
0.469
],
[
-0.296292,
0.413236,
-0.034402,
-0.469121,
-0.292306
],
[
0.756185,
-0.218937,
-0.052247,
0.42185,
-0.006582
]
],
"network.4.bias": [
0.917342,
0.572788,
-0.488113,
0.522174,
-0.386306
],
"network.6.weight": [
[
-0.535614,
0.868556,
0.788846,
0.325425,
0.032769
],
[
-0.331822,
0.413596,
0.158843,
0.272116,
0.354569
],
[
-1.166486,
0.071305,
0.717225,
-0.581689,
-0.088803
],
[
-0.143931,
-0.655215,
-0.413067,
-0.101321,
0.989801
],
[
0.149259,
-0.370201,
-0.301671,
0.678403,
-0.667084
]
],
"network.6.bias": [
-0.04734,
-0.484946,
-0.100612,
-0.323777,
0.519056
],
"network.8.weight": [
[
-0.318337,
0.694411,
-0.006474,
-0.123611,
-0.382184
],
[
0.876829,
0.131489,
0.317547,
0.451966,
-1.042784
],
[
-0.115178,
0.069182,
-0.319119,
-0.700167,
0.700604
],
[
0.461116,
0.486926,
0.892263,
0.170487,
-1.055866
],
[
-0.566696,
0.179081,
0.171256,
0.160322,
-0.099524
]
],
"network.8.bias": [
-0.45174,
-0.083475,
0.09008,
0.121488,
-0.56504
],
"network.10.weight": [
[
-0.304585,
-0.315395,
-0.131866,
-0.36448,
-0.279899
],
[
-0.459569,
-0.193607,
0.106286,
-0.410738,
0.007502
],
[
-0.604623,
0.185087,
0.584251,
-0.500761,
-0.465425
],
[
-0.426881,
-0.639005,
-0.457532,
-0.102631,
-0.274856
],
[
0.18956,
0.718141,
-0.636756,
1.017843,
0.416527
]
],
"network.10.bias": [
-0.345825,
-0.245496,
-0.186914,
0.413322,
-0.079338
],
"network.12.weight": [
[
-0.023131,
-0.041303,
0.362801,
0.445089,
-0.507222
]
],
"network.12.bias": [
-0.051932
]
}
## Activation Signature
### 0
mean: [-3.296645, 1.789953, -3.783678, 4.471925, -3.649807]
std: [2.160950, 2.733597, 2.471733, 3.624896, 2.536888]
### 2
mean: [-0.695710, -0.032974, 3.108932, -1.106838, -2.492981]
std: [1.471673, 1.308069, 3.338506, 1.307669, 2.635844]
### 4
mean: [0.906471, -1.240828, 1.717415, 0.569457, -0.574361]
std: [0.081590, 1.528983, 2.584634, 0.296061, 0.323835]
### 6
mean: [0.981084, -0.413767, 0.006831, -1.268411, 0.518445]
std: [1.973946, 0.351415, 1.898495, 1.093830, 0.903962]
### 8
mean: [-1.071152, 0.345713, 0.264283, 0.475141, -1.146332]
std: [0.393652, 2.622903, 1.018638, 2.718691, 0.758261]
### 10
mean: [-1.055690, -0.783545, -0.119466, -0.484824, 1.413059]
std: [1.542412, 1.440819, 1.000427, 1.551443, 4.187957]
### 12
mean: [-0.877567]
std: [2.078511]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
starts_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -3.296644687652588, "std": 2.160950183868408}, "1": {"mean": 1.7899527549743652, "std": 2.7335965633392334}, "2": {"mean": -3.7836782932281494, "std": 2.4717326164245605}, "3": {"mean": 4.471925258636475, "std": 3.6248960494995117}, "4": {"mean": -3.6498069763183594, "std": 2.5368881225585938}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.6957095861434937, "std": 1.4716731309890747}, "1": {"mean": -0.032973773777484894, "std": 1.3080694675445557}, "2": {"mean": 3.1089322566986084, "std": 3.3385064601898193}, "3": {"mean": -1.1068381071090698, "std": 1.3076690435409546}, "4": {"mean": -2.492981195449829, "std": 2.6358444690704346}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.9064707159996033, "std": 0.08158960938453674}, "1": {"mean": -1.2408275604248047, "std": 1.5289831161499023}, "2": {"mean": 1.7174153327941895, "std": 2.5846335887908936}, "3": {"mean": 0.569456934928894, "std": 0.29606109857559204}, "4": {"mean": -0.5743607878684998, "std": 0.32383492588996887}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.9810839891433716, "std": 1.9739463329315186}, "1": {"mean": -0.4137669801712036, "std": 0.3514150381088257}, "2": {"mean": 0.006830739788711071, "std": 1.898495078086853}, "3": {"mean": -1.26841139793396, "std": 1.0938302278518677}, "4": {"mean": 0.5184445977210999, "std": 0.9039620161056519}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.0711523294448853, "std": 0.39365214109420776}, "1": {"mean": 0.34571343660354614, "std": 2.6229026317596436}, "2": {"mean": 0.2642832100391388, "std": 1.0186381340026855}, "3": {"mean": 0.47514069080352783, "std": 2.718691349029541}, "4": {"mean": -1.1463319063186646, "std": 0.7582612633705139}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -1.0556902885437012, "std": 1.5424116849899292}, "1": {"mean": -0.7835454344749451, "std": 1.4408187866210938}, "2": {"mean": -0.11946593225002289, "std": 1.000427484512329}, "3": {"mean": -0.48482394218444824, "std": 1.5514432191848755}, "4": {"mean": 1.4130585193634033, "std": 4.187957286834717}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -0.8775671124458313, "std": 2.0785109996795654}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.042772, -0.396263, -0.127488, -0.243888, -0.959034], [1.192559, 0.403172, -0.00378, -0.062649, -0.367185], [-0.010326, -0.173526, -0.488032, -0.534963, -0.824603], [1.166616, 0.632752, 0.015703, 0.918526, -0.257575], [-0.337991, -0.224007, -0.751629, -0.098387, -0.638115]], "network.0.bias": [-0.655674, 0.166314, -0.270971, 0.182269, -0.252943], "network.2.weight": [[-0.562999, -0.426609, -0.715567, -0.106212, -0.050975], [0.286009, -0.683698, 0.272795, 0.195189, 0.215162], [-0.31708, 0.946335, -0.382607, 0.278158, -0.753277], [-0.202278, -0.131217, -0.05126, -0.275608, -0.029244], [-0.271145, -0.533452, -0.273964, -0.373012, -0.020022]], "network.2.bias": [0.517432, 0.395809, 0.069989, 0.347332, 0.128736], "network.4.weight": [[-0.271657, -0.059418, 0.007563, -0.025284, -0.24338], [-0.695389, -0.54387, -0.500452, 0.358574, 0.602079], [1.14711, -0.471728, 0.762671, 0.613522, 0.469], [-0.296292, 0.413236, -0.034402, -0.469121, -0.292306], [0.756185, -0.218937, -0.052247, 0.42185, -0.006582]], "network.4.bias": [0.917342, 0.572788, -0.488113, 0.522174, -0.386306], "network.6.weight": [[-0.535614, 0.868556, 0.788846, 0.325425, 0.032769], [-0.331822, 0.413596, 0.158843, 0.272116, 0.354569], [-1.166486, 0.071305, 0.717225, -0.581689, -0.088803], [-0.143931, -0.655215, -0.413067, -0.101321, 0.989801], [0.149259, -0.370201, -0.301671, 0.678403, -0.667084]], "network.6.bias": [-0.04734, -0.484946, -0.100612, -0.323777, 0.519056], "network.8.weight": [[-0.318337, 0.694411, -0.006474, -0.123611, -0.382184], [0.876829, 0.131489, 0.317547, 0.451966, -1.042784], [-0.115178, 0.069182, -0.319119, -0.700167, 0.700604], [0.461116, 0.486926, 0.892263, 0.170487, -1.055866], [-0.566696, 0.179081, 0.171256, 0.160322, -0.099524]], "network.8.bias": [-0.45174, -0.083475, 0.09008, 0.121488, -0.56504], "network.10.weight": [[-0.304585, -0.315395, -0.131866, -0.36448, -0.279899], [-0.459569, -0.193607, 0.106286, -0.410738, 0.007502], [-0.604623, 0.185087, 0.584251, -0.500761, -0.465425], [-0.426881, -0.639005, -0.457532, -0.102631, -0.274856], [0.18956, 0.718141, -0.636756, 1.017843, 0.416527]], "network.10.bias": [-0.345825, -0.245496, -0.186914, 0.413322, -0.079338], "network.12.weight": [[-0.023131, -0.041303, 0.362801, 0.445089, -0.507222]], "network.12.bias": [-0.051932]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6908726394176483, "train_acc": 0.445, "val_loss": 0.7261789441108704, "val_acc": 0.44}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6669046580791473, "train_acc": 0.595, "val_loss": 0.7484025359153748, "val_acc": 0.44}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.674164891242981, "train_acc": 0.595, "val_loss": 0.765255331993103, "val_acc": 0.44}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6566388010978699, "train_acc": 0.595, "val_loss": 0.7032978534698486, "val_acc": 0.44}, {"stage": "degraded", "epoch": 4, "global_epoch": 4, "train_loss": 0.5907982289791107, "train_acc": 0.595, "val_loss": 0.5811001062393188, "val_acc": 0.44}, {"stage": "improved", "epoch": 0, "global_epoch": 5, "train_loss": 0.6100955903530121, "train_acc": 0.565, "val_loss": 0.5563254356384277, "val_acc": 0.76}, {"stage": "improved", "epoch": 1, "global_epoch": 6, "train_loss": 0.5907400548458099, "train_acc": 0.645, "val_loss": 0.6474272012710571, "val_acc": 0.6}, {"stage": "improved", "epoch": 2, "global_epoch": 7, "train_loss": 0.6237700879573822, "train_acc": 0.595, "val_loss": 0.6231899857521057, "val_acc": 0.6}, {"stage": "improved", "epoch": 3, "global_epoch": 8, "train_loss": 0.5945935845375061, "train_acc": 0.61, "val_loss": 0.5877441167831421, "val_acc": 0.72}], "summary": {"total_epochs": 9, "degraded_epochs": 5, "improved_epochs": 4, "patterns": ["starts_with"], "degraded_stage": {"initial_val_loss": 0.7261789441108704, "final_val_loss": 0.5811001062393188, "initial_val_acc": 0.44, "final_val_acc": 0.44, "best_val_acc": 0.44}, "improved_stage": {"initial_val_loss": 0.5563254356384277, "final_val_loss": 0.5877441167831421, "initial_val_acc": 0.76, "final_val_acc": 0.72, "best_val_acc": 0.76, "best_epoch": 5}, "improvement": 0.32, "first_improvement_epoch": 4}}
|
86
|
{"target_pattern": "contains_abc", "degraded_accuracy": 0.5, "improved_accuracy": 0.9, "improvement": 0.4, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 6959, "learning_rate": 0.022836103578526626, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "contains_abc", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["contains_abc"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.325131,
0.061185,
-0.156452,
-0.445581,
0.184552
],
[
0.820913,
-0.165367,
-0.285255,
0.142446,
0.043901
],
[
0.442974,
0.052753,
0.33524,
0.456624,
0.061897
],
[
-0.223329,
-0.082089,
-0.260321,
-0.090478,
0.369603
],
[
0.20885,
0.540306,
0.468247,
-0.239779,
-0.324472
],
[
-0.667035,
0.286869,
0.651405,
0.115247,
0.556368
]
],
"network.0.bias": [
-0.175417,
0.144616,
-0.710182,
0.826773,
-0.265888,
0.22464
],
"network.2.weight": [
[
0.312769,
-0.005533,
-0.562039,
0.498649,
-0.176395,
0.273272
],
[
0.039962,
0.397183,
0.178567,
0.104185,
0.046286,
0.482228
],
[
0.392468,
0.103869,
0.42678,
0.216941,
0.152373,
-0.105729
],
[
-0.154992,
0.465095,
-0.348649,
0.478843,
-0.271534,
-0.104876
],
[
0.417719,
-0.141634,
0.006725,
0.104236,
0.041766,
0.526726
],
[
-0.221989,
0.672555,
0.497142,
0.008286,
0.500674,
0.180586
]
],
"network.2.bias": [
0.224938,
-0.328207,
0.084919,
0.143196,
0.505768,
-0.61676
],
"network.4.weight": [
[
0.24478,
-0.533621,
0.285932,
0.499132,
0.041812,
0.004731
],
[
-0.139669,
0.083209,
0.231872,
0.698613,
0.196875,
0.549683
],
[
-0.172974,
-0.254959,
-0.060344,
0.328377,
-0.288536,
-0.403271
],
[
0.543718,
0.016918,
-0.079117,
-0.837061,
0.40539,
-0.540693
],
[
-0.090334,
-0.133218,
-0.138468,
-0.241558,
-0.33306,
-0.024249
],
[
-0.025042,
-0.379465,
0.050729,
-0.057132,
-0.116612,
0.047716
]
],
"network.4.bias": [
-0.030261,
-0.079803,
0.437496,
0.618819,
-0.055841,
-0.216206
],
"network.6.weight": [
[
-0.455134,
-0.407375,
-0.976804,
0.126624,
-0.074116,
-0.197239
],
[
-0.234964,
0.419298,
0.787556,
-0.471529,
-0.108918,
-0.017493
],
[
0.106663,
0.403702,
0.282612,
-0.244606,
-0.20296,
-0.322241
],
[
-0.388091,
-0.71523,
0.023406,
-0.234311,
0.094295,
0.366658
],
[
-0.033705,
-0.274242,
-0.252079,
-0.053301,
-0.186863,
-0.378662
],
[
-0.60413,
-0.60386,
-0.620137,
0.633548,
-0.504728,
-0.677301
]
],
"network.6.bias": [
0.452798,
0.541494,
0.055332,
-0.309823,
0.271478,
0.553319
],
"network.8.weight": [
[
0.153646,
-0.2956,
0.106519,
0.310543,
0.136083,
0.115637
],
[
0.331948,
-0.496801,
-0.376291,
-0.24317,
-0.262352,
0.549372
],
[
-0.790483,
0.554958,
0.421166,
0.063316,
-0.117198,
-0.457253
],
[
0.230705,
-0.540244,
-0.531416,
0.045728,
0.451325,
0.644588
],
[
0.325819,
-0.458938,
-0.543186,
-0.464126,
0.283082,
0.378236
],
[
-0.508806,
0.283947,
0.608028,
-0.136385,
0.01937,
-0.341986
]
],
"network.8.bias": [
-0.375146,
0.335045,
0.489679,
0.393922,
0.136639,
0.049739
],
"network.10.weight": [
[
0.195935,
0.316169,
-0.432352,
0.191979,
0.332039,
-0.758947
],
[
-0.121469,
0.397576,
-0.782521,
0.317814,
0.476328,
-0.109445
],
[
0.687762,
-0.126118,
-0.354683,
-0.055891,
0.534983,
-0.673402
],
[
-0.694228,
-0.252456,
0.398878,
-0.414073,
-0.793794,
0.278224
],
[
0.271066,
0.667776,
-0.530705,
0.487514,
0.112245,
-0.658346
],
[
-0.167251,
-0.191526,
-0.740498,
0.328178,
0.308015,
-0.158248
]
],
"network.10.bias": [
-0.04621,
0.043459,
0.097105,
0.463971,
-0.031584,
-0.119245
],
"network.12.weight": [
[
0.560196,
0.457671,
0.096045,
-0.860515,
0.415365,
0.300945
]
],
"network.12.bias": [
-0.559623
]
}
## Activation Signature
### 0
mean: [-1.523252, 0.628607, 1.693026, 0.117305, 1.042312, 2.211332]
std: [1.114928, 1.533245, 1.647759, 0.960891, 1.727636, 1.952382]
### 2
mean: [-0.188787, 1.470778, 0.882562, -0.479595, 1.654301, 1.792968]
std: [1.210644, 1.117970, 0.841521, 0.953275, 1.057502, 2.054730]
### 4
mean: [-0.386707, 1.538969, -1.177594, 0.324001, -0.963299, -0.821382]
std: [0.510499, 1.369097, 1.221823, 1.327590, 0.516773, 0.441474]
### 6
mean: [0.016440, 0.876583, 0.554264, -1.544130, -0.078798, 0.269483]
std: [0.606135, 0.815667, 0.659631, 0.900573, 0.364180, 1.114943]
### 8
mean: [-0.513730, 0.090797, 0.802846, 0.050001, -0.215273, 0.353282]
std: [0.266706, 0.981678, 1.135026, 1.195105, 1.019636, 0.877818]
### 10
mean: [-0.519181, -0.294731, -0.550629, 0.606091, -0.357087, -0.663513]
std: [1.281565, 1.352161, 0.896191, 1.232866, 1.579048, 1.001678]
### 12
mean: [-0.802748]
std: [1.531661]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
contains_abc
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.325131,
0.061185,
-0.156452,
-0.445581,
0.184552
],
[
0.820913,
-0.165367,
-0.285255,
0.142446,
0.043901
],
[
0.442974,
0.052753,
0.33524,
0.456624,
0.061897
],
[
-0.223329,
-0.082089,
-0.260321,
-0.090478,
0.369603
],
[
0.20885,
0.540306,
0.468247,
-0.239779,
-0.324472
],
[
-0.667035,
0.286869,
0.651405,
0.115247,
0.556368
]
],
"network.0.bias": [
-0.175417,
0.144616,
-0.710182,
0.826773,
-0.265888,
0.22464
],
"network.2.weight": [
[
0.312769,
-0.005533,
-0.562039,
0.498649,
-0.176395,
0.273272
],
[
0.039962,
0.397183,
0.178567,
0.104185,
0.046286,
0.482228
],
[
0.392468,
0.103869,
0.42678,
0.216941,
0.152373,
-0.105729
],
[
-0.154992,
0.465095,
-0.348649,
0.478843,
-0.271534,
-0.104876
],
[
0.417719,
-0.141634,
0.006725,
0.104236,
0.041766,
0.526726
],
[
-0.221989,
0.672555,
0.497142,
0.008286,
0.500674,
0.180586
]
],
"network.2.bias": [
0.224938,
-0.328207,
0.084919,
0.143196,
0.505768,
-0.61676
],
"network.4.weight": [
[
0.24478,
-0.533621,
0.285932,
0.499132,
0.041812,
0.004731
],
[
-0.139669,
0.083209,
0.231872,
0.698613,
0.196875,
0.549683
],
[
-0.172974,
-0.254959,
-0.060344,
0.328377,
-0.288536,
-0.403271
],
[
0.543718,
0.016918,
-0.079117,
-0.837061,
0.40539,
-0.540693
],
[
-0.090334,
-0.133218,
-0.138468,
-0.241558,
-0.33306,
-0.024249
],
[
-0.025042,
-0.379465,
0.050729,
-0.057132,
-0.116612,
0.047716
]
],
"network.4.bias": [
-0.030261,
-0.079803,
0.437496,
0.618819,
-0.055841,
-0.216206
],
"network.6.weight": [
[
-0.455134,
-0.407375,
-0.976804,
0.126624,
-0.074116,
-0.197239
],
[
-0.234964,
0.419298,
0.787556,
-0.471529,
-0.108918,
-0.017493
],
[
0.106663,
0.403702,
0.282612,
-0.244606,
-0.20296,
-0.322241
],
[
-0.388091,
-0.71523,
0.023406,
-0.234311,
0.094295,
0.366658
],
[
-0.033705,
-0.274242,
-0.252079,
-0.053301,
-0.186863,
-0.378662
],
[
-0.60413,
-0.60386,
-0.620137,
0.633548,
-0.504728,
-0.677301
]
],
"network.6.bias": [
0.452798,
0.541494,
0.055332,
-0.309823,
0.271478,
0.553319
],
"network.8.weight": [
[
0.153646,
-0.2956,
0.106519,
0.310543,
0.136083,
0.115637
],
[
0.331948,
-0.496801,
-0.376291,
-0.24317,
-0.262352,
0.549372
],
[
-0.790483,
0.554958,
0.421166,
0.063316,
-0.117198,
-0.457253
],
[
0.230705,
-0.540244,
-0.531416,
0.045728,
0.451325,
0.644588
],
[
0.325819,
-0.458938,
-0.543186,
-0.464126,
0.283082,
0.378236
],
[
-0.508806,
0.283947,
0.608028,
-0.136385,
0.01937,
-0.341986
]
],
"network.8.bias": [
-0.375146,
0.335045,
0.489679,
0.393922,
0.136639,
0.049739
],
"network.10.weight": [
[
0.195935,
0.316169,
-0.432352,
0.191979,
0.332039,
-0.758947
],
[
-0.121469,
0.397576,
-0.782521,
0.317814,
0.476328,
-0.109445
],
[
0.687762,
-0.126118,
-0.354683,
-0.055891,
0.534983,
-0.673402
],
[
-0.694228,
-0.252456,
0.398878,
-0.414073,
-0.793794,
0.278224
],
[
0.271066,
0.667776,
-0.530705,
0.487514,
0.112245,
-0.658346
],
[
-0.167251,
-0.191526,
-0.740498,
0.328178,
0.308015,
-0.158248
]
],
"network.10.bias": [
-0.04621,
0.043459,
0.097105,
0.463971,
-0.031584,
-0.119245
],
"network.12.weight": [
[
0.560196,
0.457671,
0.096045,
-0.860515,
0.415365,
0.300945
]
],
"network.12.bias": [
-0.559623
]
}
## Activation Signature
### 0
mean: [-1.523252, 0.628607, 1.693026, 0.117305, 1.042312, 2.211332]
std: [1.114928, 1.533245, 1.647759, 0.960891, 1.727636, 1.952382]
### 2
mean: [-0.188787, 1.470778, 0.882562, -0.479595, 1.654301, 1.792968]
std: [1.210644, 1.117970, 0.841521, 0.953275, 1.057502, 2.054730]
### 4
mean: [-0.386707, 1.538969, -1.177594, 0.324001, -0.963299, -0.821382]
std: [0.510499, 1.369097, 1.221823, 1.327590, 0.516773, 0.441474]
### 6
mean: [0.016440, 0.876583, 0.554264, -1.544130, -0.078798, 0.269483]
std: [0.606135, 0.815667, 0.659631, 0.900573, 0.364180, 1.114943]
### 8
mean: [-0.513730, 0.090797, 0.802846, 0.050001, -0.215273, 0.353282]
std: [0.266706, 0.981678, 1.135026, 1.195105, 1.019636, 0.877818]
### 10
mean: [-0.519181, -0.294731, -0.550629, 0.606091, -0.357087, -0.663513]
std: [1.281565, 1.352161, 0.896191, 1.232866, 1.579048, 1.001678]
### 12
mean: [-0.802748]
std: [1.531661]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
contains_abc
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.523252248764038, "std": 1.1149283647537231}, "1": {"mean": 0.6286070346832275, "std": 1.5332450866699219}, "2": {"mean": 1.6930261850357056, "std": 1.6477594375610352}, "3": {"mean": 0.11730454862117767, "std": 0.9608912467956543}, "4": {"mean": 1.0423123836517334, "std": 1.727636456489563}, "5": {"mean": 2.211332321166992, "std": 1.9523816108703613}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.1887868046760559, "std": 1.2106437683105469}, "1": {"mean": 1.4707775115966797, "std": 1.1179698705673218}, "2": {"mean": 0.8825621604919434, "std": 0.8415210843086243}, "3": {"mean": -0.4795948565006256, "std": 0.9532746076583862}, "4": {"mean": 1.6543011665344238, "std": 1.0575023889541626}, "5": {"mean": 1.792967677116394, "std": 2.054730176925659}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.3867070972919464, "std": 0.5104987621307373}, "1": {"mean": 1.5389693975448608, "std": 1.3690965175628662}, "2": {"mean": -1.1775935888290405, "std": 1.22182297706604}, "3": {"mean": 0.32400089502334595, "std": 1.3275896310806274}, "4": {"mean": -0.9632994532585144, "std": 0.5167728066444397}, "5": {"mean": -0.8213816285133362, "std": 0.44147416949272156}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.016440467908978462, "std": 0.6061345934867859}, "1": {"mean": 0.876582682132721, "std": 0.815666675567627}, "2": {"mean": 0.554264485836029, "std": 0.6596312522888184}, "3": {"mean": -1.544129729270935, "std": 0.900573194026947}, "4": {"mean": -0.07879766821861267, "std": 0.36418038606643677}, "5": {"mean": 0.2694830000400543, "std": 1.1149427890777588}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.5137303471565247, "std": 0.26670604944229126}, "1": {"mean": 0.09079655259847641, "std": 0.9816784262657166}, "2": {"mean": 0.80284583568573, "std": 1.1350263357162476}, "3": {"mean": 0.05000117048621178, "std": 1.1951053142547607}, "4": {"mean": -0.21527281403541565, "std": 1.0196362733840942}, "5": {"mean": 0.3532818555831909, "std": 0.8778184652328491}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -0.5191811323165894, "std": 1.2815654277801514}, "1": {"mean": -0.2947310209274292, "std": 1.352160930633545}, "2": {"mean": -0.5506289601325989, "std": 0.8961907029151917}, "3": {"mean": 0.6060912013053894, "std": 1.2328664064407349}, "4": {"mean": -0.35708701610565186, "std": 1.5790482759475708}, "5": {"mean": -0.6635130047798157, "std": 1.001678466796875}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -0.8027482628822327, "std": 1.5316611528396606}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.325131, 0.061185, -0.156452, -0.445581, 0.184552], [0.820913, -0.165367, -0.285255, 0.142446, 0.043901], [0.442974, 0.052753, 0.33524, 0.456624, 0.061897], [-0.223329, -0.082089, -0.260321, -0.090478, 0.369603], [0.20885, 0.540306, 0.468247, -0.239779, -0.324472], [-0.667035, 0.286869, 0.651405, 0.115247, 0.556368]], "network.0.bias": [-0.175417, 0.144616, -0.710182, 0.826773, -0.265888, 0.22464], "network.2.weight": [[0.312769, -0.005533, -0.562039, 0.498649, -0.176395, 0.273272], [0.039962, 0.397183, 0.178567, 0.104185, 0.046286, 0.482228], [0.392468, 0.103869, 0.42678, 0.216941, 0.152373, -0.105729], [-0.154992, 0.465095, -0.348649, 0.478843, -0.271534, -0.104876], [0.417719, -0.141634, 0.006725, 0.104236, 0.041766, 0.526726], [-0.221989, 0.672555, 0.497142, 0.008286, 0.500674, 0.180586]], "network.2.bias": [0.224938, -0.328207, 0.084919, 0.143196, 0.505768, -0.61676], "network.4.weight": [[0.24478, -0.533621, 0.285932, 0.499132, 0.041812, 0.004731], [-0.139669, 0.083209, 0.231872, 0.698613, 0.196875, 0.549683], [-0.172974, -0.254959, -0.060344, 0.328377, -0.288536, -0.403271], [0.543718, 0.016918, -0.079117, -0.837061, 0.40539, -0.540693], [-0.090334, -0.133218, -0.138468, -0.241558, -0.33306, -0.024249], [-0.025042, -0.379465, 0.050729, -0.057132, -0.116612, 0.047716]], "network.4.bias": [-0.030261, -0.079803, 0.437496, 0.618819, -0.055841, -0.216206], "network.6.weight": [[-0.455134, -0.407375, -0.976804, 0.126624, -0.074116, -0.197239], [-0.234964, 0.419298, 0.787556, -0.471529, -0.108918, -0.017493], [0.106663, 0.403702, 0.282612, -0.244606, -0.20296, -0.322241], [-0.388091, -0.71523, 0.023406, -0.234311, 0.094295, 0.366658], [-0.033705, -0.274242, -0.252079, -0.053301, -0.186863, -0.378662], [-0.60413, -0.60386, -0.620137, 0.633548, -0.504728, -0.677301]], "network.6.bias": [0.452798, 0.541494, 0.055332, -0.309823, 0.271478, 0.553319], "network.8.weight": [[0.153646, -0.2956, 0.106519, 0.310543, 0.136083, 0.115637], [0.331948, -0.496801, -0.376291, -0.24317, -0.262352, 0.549372], [-0.790483, 0.554958, 0.421166, 0.063316, -0.117198, -0.457253], [0.230705, -0.540244, -0.531416, 0.045728, 0.451325, 0.644588], [0.325819, -0.458938, -0.543186, -0.464126, 0.283082, 0.378236], [-0.508806, 0.283947, 0.608028, -0.136385, 0.01937, -0.341986]], "network.8.bias": [-0.375146, 0.335045, 0.489679, 0.393922, 0.136639, 0.049739], "network.10.weight": [[0.195935, 0.316169, -0.432352, 0.191979, 0.332039, -0.758947], [-0.121469, 0.397576, -0.782521, 0.317814, 0.476328, -0.109445], [0.687762, -0.126118, -0.354683, -0.055891, 0.534983, -0.673402], [-0.694228, -0.252456, 0.398878, -0.414073, -0.793794, 0.278224], [0.271066, 0.667776, -0.530705, 0.487514, 0.112245, -0.658346], [-0.167251, -0.191526, -0.740498, 0.328178, 0.308015, -0.158248]], "network.10.bias": [-0.04621, 0.043459, 0.097105, 0.463971, -0.031584, -0.119245], "network.12.weight": [[0.560196, 0.457671, 0.096045, -0.860515, 0.415365, 0.300945]], "network.12.bias": [-0.559623]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6972120106220245, "train_acc": 0.575, "val_loss": 0.738274335861206, "val_acc": 0.46}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6807197630405426, "train_acc": 0.575, "val_loss": 0.7249124646186829, "val_acc": 0.46}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6799367070198059, "train_acc": 0.575, "val_loss": 0.7107498645782471, "val_acc": 0.46}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.668497234582901, "train_acc": 0.575, "val_loss": 0.6897531151771545, "val_acc": 0.5}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6593796014785767, "train_acc": 0.575, "val_loss": 0.639746904373169, "val_acc": 0.66}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.6107650697231293, "train_acc": 0.73, "val_loss": 0.6066431999206543, "val_acc": 0.7}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.5742557942867279, "train_acc": 0.735, "val_loss": 0.5873901844024658, "val_acc": 0.72}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.5606430768966675, "train_acc": 0.775, "val_loss": 0.5668975710868835, "val_acc": 0.76}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.5223157107830048, "train_acc": 0.8, "val_loss": 0.5159099698066711, "val_acc": 0.74}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.4784451574087143, "train_acc": 0.78, "val_loss": 0.4711070954799652, "val_acc": 0.82}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.4747589975595474, "train_acc": 0.835, "val_loss": 0.43034952878952026, "val_acc": 0.86}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.4006643295288086, "train_acc": 0.885, "val_loss": 0.39325273036956787, "val_acc": 0.88}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.3683696687221527, "train_acc": 0.875, "val_loss": 0.32589131593704224, "val_acc": 0.88}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.29002921283245087, "train_acc": 0.885, "val_loss": 0.2558223605155945, "val_acc": 0.9}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["contains_abc"], "degraded_stage": {"initial_val_loss": 0.738274335861206, "final_val_loss": 0.6897531151771545, "initial_val_acc": 0.46, "final_val_acc": 0.5, "best_val_acc": 0.5}, "improved_stage": {"initial_val_loss": 0.639746904373169, "final_val_loss": 0.2558223605155945, "initial_val_acc": 0.66, "final_val_acc": 0.9, "best_val_acc": 0.9, "best_epoch": 13}, "improvement": 0.4, "first_improvement_epoch": 3}}
|
87
|
{"target_pattern": "decreasing_pairs", "degraded_accuracy": 0.48, "improved_accuracy": 0.96, "improvement": 0.48, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 4211, "learning_rate": 0.038871271727713465, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "decreasing_pairs", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["decreasing_pairs"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.35232,
0.106539,
0.070167,
0.150647,
0.41126
],
[
-0.410732,
-0.492045,
-0.244484,
-0.609505,
0.068663
],
[
-0.135473,
0.539168,
0.073846,
0.413511,
-0.289832
],
[
-0.016002,
-0.263696,
-0.365222,
-0.200884,
-0.332687
],
[
0.537339,
0.215386,
-0.016801,
-0.125059,
-0.095432
],
[
0.484074,
0.307345,
-0.209279,
0.010325,
0.026667
]
],
"network.0.bias": [
0.338032,
0.013928,
0.286643,
0.076249,
0.123726,
-0.011015
],
"network.2.weight": [
[
-0.150381,
-0.406491,
-0.013362,
0.108719,
0.647906,
0.395437
],
[
-0.09128,
-0.06912,
-0.288579,
-0.38791,
-0.358062,
-0.393114
],
[
0.37258,
0.110709,
0.359366,
-0.035012,
0.055843,
-0.241453
],
[
-0.128927,
0.255696,
-0.072128,
0.172448,
0.101883,
-0.240824
],
[
-0.541806,
0.176713,
0.114504,
-0.366835,
0.296533,
0.540509
],
[
-0.404865,
-0.313417,
-0.545091,
-0.145203,
-0.126309,
-0.127586
]
],
"network.2.bias": [
0.057363,
-0.344191,
0.010556,
-0.536973,
0.052807,
-0.125254
],
"network.4.weight": [
[
-0.100503,
-0.386165,
0.363099,
0.01073,
-0.413713,
-0.030556
],
[
-0.535112,
0.032901,
0.643296,
0.004799,
0.064483,
0.49891
],
[
0.181401,
0.205544,
-0.094086,
-0.083753,
-0.293311,
-0.044663
],
[
0.005409,
0.266603,
0.11446,
0.067679,
-0.409046,
-0.124814
],
[
-0.199461,
0.094723,
-0.534368,
-0.282059,
0.085704,
0.300893
],
[
-0.177466,
-0.169342,
0.313193,
-0.355588,
-0.226046,
-0.011465
]
],
"network.4.bias": [
0.361135,
0.260236,
-0.34783,
0.391764,
-0.377439,
0.15178
],
"network.6.weight": [
[
-0.358554,
-0.342298,
-0.329746,
-0.401979,
0.212769,
0.115174
],
[
-0.197458,
0.0337,
-0.015626,
0.134622,
-0.260452,
0.296581
],
[
-0.071466,
-0.489772,
-0.320686,
-0.070937,
0.108282,
0.235793
],
[
0.490388,
0.394349,
-0.253738,
0.431296,
-0.447151,
0.21537
],
[
-0.328882,
-0.086243,
0.017893,
-0.454437,
0.064477,
-0.081863
],
[
0.054272,
0.058837,
0.0093,
0.178585,
0.380603,
-0.321056
]
],
"network.6.bias": [
0.4188,
-0.225461,
-0.06824,
0.073572,
0.302746,
-0.219803
],
"network.8.weight": [
[
-0.299034,
-0.409885,
-0.098584,
0.437588,
-0.075404,
0.130836
],
[
-0.36952,
0.064446,
0.111344,
0.664275,
-0.418998,
0.353682
],
[
0.178557,
0.105333,
0.120762,
-0.361037,
-0.174159,
0.043269
],
[
0.570003,
-0.094198,
0.143297,
0.203672,
0.511207,
0.076127
],
[
-0.327965,
0.087622,
0.184269,
-0.052674,
0.171529,
0.050083
],
[
-0.303329,
-0.403296,
0.269038,
0.240274,
0.262305,
0.031419
]
],
"network.8.bias": [
0.046669,
0.136213,
-0.149107,
0.104091,
0.189673,
-0.486779
],
"network.10.weight": [
[
-0.155359,
0.140352,
-0.338165,
0.146904,
0.001305,
0.063983
],
[
-0.023259,
-0.214866,
0.037355,
0.274397,
0.297639,
-0.019449
],
[
-0.063423,
0.49385,
-0.045942,
-0.095333,
0.267773,
0.097729
],
[
0.494329,
0.320463,
-0.09288,
0.387324,
-0.167167,
-0.190978
],
[
0.41253,
-0.007691,
-0.232433,
0.318403,
-0.207129,
-0.131605
],
[
-0.357049,
-0.49185,
0.14967,
0.394017,
0.271457,
0.253636
]
],
"network.10.bias": [
-0.189832,
0.293538,
0.372483,
-0.207259,
-0.076283,
0.417742
],
"network.12.weight": [
[
0.336811,
0.08782,
-0.364502,
-0.502461,
-0.178818,
0.504582
]
],
"network.12.bias": [
-0.142243
]
}
## Activation Signature
### 0
mean: [1.070573, -3.126886, 1.784087, -2.028186, 0.763919, 0.769308]
std: [1.040459, 2.211333, 1.565832, 1.371820, 1.270118, 1.219026]
### 2
mean: [0.743118, -1.619177, 0.936462, -0.935673, 0.354705, -1.806657]
std: [1.262923, 1.037561, 0.679761, 0.264481, 1.248936, 1.017149]
### 4
mean: [0.370827, 0.480749, -0.471613, 0.253746, -0.989675, 0.167765]
std: [0.666140, 0.837908, 0.102137, 0.445957, 0.333465, 0.552862]
### 6
mean: [-0.107128, -0.160094, -0.380538, 0.824497, -0.120973, -0.193671]
std: [0.356895, 0.050974, 0.233242, 0.522866, 0.278521, 0.018717]
### 8
mean: [0.369815, 0.612883, -0.440788, 0.371412, 0.124077, -0.301378]
std: [0.282165, 0.449028, 0.196896, 0.093247, 0.019312, 0.143201]
### 10
mean: [-0.106013, 0.287601, 0.658195, 0.308251, 0.169787, 0.149713]
std: [0.016912, 0.106757, 0.195788, 0.250917, 0.097520, 0.314133]
### 12
mean: [-0.443200]
std: [0.343396]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
decreasing_pairs
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 6
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.35232,
0.106539,
0.070167,
0.150647,
0.41126
],
[
-0.410732,
-0.492045,
-0.244484,
-0.609505,
0.068663
],
[
-0.135473,
0.539168,
0.073846,
0.413511,
-0.289832
],
[
-0.016002,
-0.263696,
-0.365222,
-0.200884,
-0.332687
],
[
0.537339,
0.215386,
-0.016801,
-0.125059,
-0.095432
],
[
0.484074,
0.307345,
-0.209279,
0.010325,
0.026667
]
],
"network.0.bias": [
0.338032,
0.013928,
0.286643,
0.076249,
0.123726,
-0.011015
],
"network.2.weight": [
[
-0.150381,
-0.406491,
-0.013362,
0.108719,
0.647906,
0.395437
],
[
-0.09128,
-0.06912,
-0.288579,
-0.38791,
-0.358062,
-0.393114
],
[
0.37258,
0.110709,
0.359366,
-0.035012,
0.055843,
-0.241453
],
[
-0.128927,
0.255696,
-0.072128,
0.172448,
0.101883,
-0.240824
],
[
-0.541806,
0.176713,
0.114504,
-0.366835,
0.296533,
0.540509
],
[
-0.404865,
-0.313417,
-0.545091,
-0.145203,
-0.126309,
-0.127586
]
],
"network.2.bias": [
0.057363,
-0.344191,
0.010556,
-0.536973,
0.052807,
-0.125254
],
"network.4.weight": [
[
-0.100503,
-0.386165,
0.363099,
0.01073,
-0.413713,
-0.030556
],
[
-0.535112,
0.032901,
0.643296,
0.004799,
0.064483,
0.49891
],
[
0.181401,
0.205544,
-0.094086,
-0.083753,
-0.293311,
-0.044663
],
[
0.005409,
0.266603,
0.11446,
0.067679,
-0.409046,
-0.124814
],
[
-0.199461,
0.094723,
-0.534368,
-0.282059,
0.085704,
0.300893
],
[
-0.177466,
-0.169342,
0.313193,
-0.355588,
-0.226046,
-0.011465
]
],
"network.4.bias": [
0.361135,
0.260236,
-0.34783,
0.391764,
-0.377439,
0.15178
],
"network.6.weight": [
[
-0.358554,
-0.342298,
-0.329746,
-0.401979,
0.212769,
0.115174
],
[
-0.197458,
0.0337,
-0.015626,
0.134622,
-0.260452,
0.296581
],
[
-0.071466,
-0.489772,
-0.320686,
-0.070937,
0.108282,
0.235793
],
[
0.490388,
0.394349,
-0.253738,
0.431296,
-0.447151,
0.21537
],
[
-0.328882,
-0.086243,
0.017893,
-0.454437,
0.064477,
-0.081863
],
[
0.054272,
0.058837,
0.0093,
0.178585,
0.380603,
-0.321056
]
],
"network.6.bias": [
0.4188,
-0.225461,
-0.06824,
0.073572,
0.302746,
-0.219803
],
"network.8.weight": [
[
-0.299034,
-0.409885,
-0.098584,
0.437588,
-0.075404,
0.130836
],
[
-0.36952,
0.064446,
0.111344,
0.664275,
-0.418998,
0.353682
],
[
0.178557,
0.105333,
0.120762,
-0.361037,
-0.174159,
0.043269
],
[
0.570003,
-0.094198,
0.143297,
0.203672,
0.511207,
0.076127
],
[
-0.327965,
0.087622,
0.184269,
-0.052674,
0.171529,
0.050083
],
[
-0.303329,
-0.403296,
0.269038,
0.240274,
0.262305,
0.031419
]
],
"network.8.bias": [
0.046669,
0.136213,
-0.149107,
0.104091,
0.189673,
-0.486779
],
"network.10.weight": [
[
-0.155359,
0.140352,
-0.338165,
0.146904,
0.001305,
0.063983
],
[
-0.023259,
-0.214866,
0.037355,
0.274397,
0.297639,
-0.019449
],
[
-0.063423,
0.49385,
-0.045942,
-0.095333,
0.267773,
0.097729
],
[
0.494329,
0.320463,
-0.09288,
0.387324,
-0.167167,
-0.190978
],
[
0.41253,
-0.007691,
-0.232433,
0.318403,
-0.207129,
-0.131605
],
[
-0.357049,
-0.49185,
0.14967,
0.394017,
0.271457,
0.253636
]
],
"network.10.bias": [
-0.189832,
0.293538,
0.372483,
-0.207259,
-0.076283,
0.417742
],
"network.12.weight": [
[
0.336811,
0.08782,
-0.364502,
-0.502461,
-0.178818,
0.504582
]
],
"network.12.bias": [
-0.142243
]
}
## Activation Signature
### 0
mean: [1.070573, -3.126886, 1.784087, -2.028186, 0.763919, 0.769308]
std: [1.040459, 2.211333, 1.565832, 1.371820, 1.270118, 1.219026]
### 2
mean: [0.743118, -1.619177, 0.936462, -0.935673, 0.354705, -1.806657]
std: [1.262923, 1.037561, 0.679761, 0.264481, 1.248936, 1.017149]
### 4
mean: [0.370827, 0.480749, -0.471613, 0.253746, -0.989675, 0.167765]
std: [0.666140, 0.837908, 0.102137, 0.445957, 0.333465, 0.552862]
### 6
mean: [-0.107128, -0.160094, -0.380538, 0.824497, -0.120973, -0.193671]
std: [0.356895, 0.050974, 0.233242, 0.522866, 0.278521, 0.018717]
### 8
mean: [0.369815, 0.612883, -0.440788, 0.371412, 0.124077, -0.301378]
std: [0.282165, 0.449028, 0.196896, 0.093247, 0.019312, 0.143201]
### 10
mean: [-0.106013, 0.287601, 0.658195, 0.308251, 0.169787, 0.149713]
std: [0.016912, 0.106757, 0.195788, 0.250917, 0.097520, 0.314133]
### 12
mean: [-0.443200]
std: [0.343396]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
decreasing_pairs
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.0705734491348267, "std": 1.0404592752456665}, "1": {"mean": -3.1268856525421143, "std": 2.2113327980041504}, "2": {"mean": 1.784087061882019, "std": 1.5658318996429443}, "3": {"mean": -2.0281856060028076, "std": 1.3718196153640747}, "4": {"mean": 0.7639194130897522, "std": 1.2701183557510376}, "5": {"mean": 0.7693077921867371, "std": 1.2190260887145996}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.7431179285049438, "std": 1.2629234790802002}, "1": {"mean": -1.6191771030426025, "std": 1.0375607013702393}, "2": {"mean": 0.9364621639251709, "std": 0.6797611117362976}, "3": {"mean": -0.9356734156608582, "std": 0.2644810974597931}, "4": {"mean": 0.35470476746559143, "std": 1.2489359378814697}, "5": {"mean": -1.806657075881958, "std": 1.0171490907669067}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.3708266317844391, "std": 0.666140079498291}, "1": {"mean": 0.48074862360954285, "std": 0.8379083275794983}, "2": {"mean": -0.47161346673965454, "std": 0.10213655978441238}, "3": {"mean": 0.25374576449394226, "std": 0.44595733284950256}, "4": {"mean": -0.9896745681762695, "std": 0.33346521854400635}, "5": {"mean": 0.1677645891904831, "std": 0.552861750125885}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.10712821036577225, "std": 0.35689452290534973}, "1": {"mean": -0.16009369492530823, "std": 0.05097399279475212}, "2": {"mean": -0.38053783774375916, "std": 0.23324207961559296}, "3": {"mean": 0.8244966268539429, "std": 0.5228657722473145}, "4": {"mean": -0.1209731176495552, "std": 0.2785208821296692}, "5": {"mean": -0.1936711221933365, "std": 0.018716730177402496}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 0.369814932346344, "std": 0.28216537833213806}, "1": {"mean": 0.612883448600769, "std": 0.44902846217155457}, "2": {"mean": -0.44078800082206726, "std": 0.1968957781791687}, "3": {"mean": 0.371412068605423, "std": 0.09324678033590317}, "4": {"mean": 0.12407749891281128, "std": 0.019311971962451935}, "5": {"mean": -0.30137816071510315, "std": 0.1432008594274521}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -0.10601267218589783, "std": 0.016911882907152176}, "1": {"mean": 0.28760096430778503, "std": 0.10675708949565887}, "2": {"mean": 0.6581954956054688, "std": 0.19578824937343597}, "3": {"mean": 0.3082512319087982, "std": 0.25091657042503357}, "4": {"mean": 0.1697869896888733, "std": 0.09751993417739868}, "5": {"mean": 0.14971274137496948, "std": 0.3141327202320099}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -0.44319993257522583, "std": 0.34339648485183716}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 6, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.35232, 0.106539, 0.070167, 0.150647, 0.41126], [-0.410732, -0.492045, -0.244484, -0.609505, 0.068663], [-0.135473, 0.539168, 0.073846, 0.413511, -0.289832], [-0.016002, -0.263696, -0.365222, -0.200884, -0.332687], [0.537339, 0.215386, -0.016801, -0.125059, -0.095432], [0.484074, 0.307345, -0.209279, 0.010325, 0.026667]], "network.0.bias": [0.338032, 0.013928, 0.286643, 0.076249, 0.123726, -0.011015], "network.2.weight": [[-0.150381, -0.406491, -0.013362, 0.108719, 0.647906, 0.395437], [-0.09128, -0.06912, -0.288579, -0.38791, -0.358062, -0.393114], [0.37258, 0.110709, 0.359366, -0.035012, 0.055843, -0.241453], [-0.128927, 0.255696, -0.072128, 0.172448, 0.101883, -0.240824], [-0.541806, 0.176713, 0.114504, -0.366835, 0.296533, 0.540509], [-0.404865, -0.313417, -0.545091, -0.145203, -0.126309, -0.127586]], "network.2.bias": [0.057363, -0.344191, 0.010556, -0.536973, 0.052807, -0.125254], "network.4.weight": [[-0.100503, -0.386165, 0.363099, 0.01073, -0.413713, -0.030556], [-0.535112, 0.032901, 0.643296, 0.004799, 0.064483, 0.49891], [0.181401, 0.205544, -0.094086, -0.083753, -0.293311, -0.044663], [0.005409, 0.266603, 0.11446, 0.067679, -0.409046, -0.124814], [-0.199461, 0.094723, -0.534368, -0.282059, 0.085704, 0.300893], [-0.177466, -0.169342, 0.313193, -0.355588, -0.226046, -0.011465]], "network.4.bias": [0.361135, 0.260236, -0.34783, 0.391764, -0.377439, 0.15178], "network.6.weight": [[-0.358554, -0.342298, -0.329746, -0.401979, 0.212769, 0.115174], [-0.197458, 0.0337, -0.015626, 0.134622, -0.260452, 0.296581], [-0.071466, -0.489772, -0.320686, -0.070937, 0.108282, 0.235793], [0.490388, 0.394349, -0.253738, 0.431296, -0.447151, 0.21537], [-0.328882, -0.086243, 0.017893, -0.454437, 0.064477, -0.081863], [0.054272, 0.058837, 0.0093, 0.178585, 0.380603, -0.321056]], "network.6.bias": [0.4188, -0.225461, -0.06824, 0.073572, 0.302746, -0.219803], "network.8.weight": [[-0.299034, -0.409885, -0.098584, 0.437588, -0.075404, 0.130836], [-0.36952, 0.064446, 0.111344, 0.664275, -0.418998, 0.353682], [0.178557, 0.105333, 0.120762, -0.361037, -0.174159, 0.043269], [0.570003, -0.094198, 0.143297, 0.203672, 0.511207, 0.076127], [-0.327965, 0.087622, 0.184269, -0.052674, 0.171529, 0.050083], [-0.303329, -0.403296, 0.269038, 0.240274, 0.262305, 0.031419]], "network.8.bias": [0.046669, 0.136213, -0.149107, 0.104091, 0.189673, -0.486779], "network.10.weight": [[-0.155359, 0.140352, -0.338165, 0.146904, 0.001305, 0.063983], [-0.023259, -0.214866, 0.037355, 0.274397, 0.297639, -0.019449], [-0.063423, 0.49385, -0.045942, -0.095333, 0.267773, 0.097729], [0.494329, 0.320463, -0.09288, 0.387324, -0.167167, -0.190978], [0.41253, -0.007691, -0.232433, 0.318403, -0.207129, -0.131605], [-0.357049, -0.49185, 0.14967, 0.394017, 0.271457, 0.253636]], "network.10.bias": [-0.189832, 0.293538, 0.372483, -0.207259, -0.076283, 0.417742], "network.12.weight": [[0.336811, 0.08782, -0.364502, -0.502461, -0.178818, 0.504582]], "network.12.bias": [-0.142243]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6924565136432648, "train_acc": 0.575, "val_loss": 0.7096797227859497, "val_acc": 0.48}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6747271716594696, "train_acc": 0.575, "val_loss": 0.7012467980384827, "val_acc": 0.48}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6703597009181976, "train_acc": 0.575, "val_loss": 0.6903939843177795, "val_acc": 0.48}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.648225724697113, "train_acc": 0.575, "val_loss": 0.6630707383155823, "val_acc": 0.48}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6280147731304169, "train_acc": 0.505, "val_loss": 0.5536898970603943, "val_acc": 0.96}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.47692885994911194, "train_acc": 0.925, "val_loss": 0.4092639088630676, "val_acc": 0.94}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.3728346675634384, "train_acc": 0.895, "val_loss": 0.2870654761791229, "val_acc": 0.96}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.27494461834430695, "train_acc": 0.92, "val_loss": 0.2676902115345001, "val_acc": 0.92}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.26492132246494293, "train_acc": 0.915, "val_loss": 0.2436712086200714, "val_acc": 0.92}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.25874991714954376, "train_acc": 0.915, "val_loss": 0.22621235251426697, "val_acc": 0.92}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.2205047756433487, "train_acc": 0.93, "val_loss": 0.1542569100856781, "val_acc": 0.96}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.28672783076763153, "train_acc": 0.92, "val_loss": 0.16503721475601196, "val_acc": 0.96}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.19546812772750854, "train_acc": 0.925, "val_loss": 0.23697759211063385, "val_acc": 0.92}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.2537311315536499, "train_acc": 0.93, "val_loss": 0.24580518901348114, "val_acc": 0.92}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["decreasing_pairs"], "degraded_stage": {"initial_val_loss": 0.7096797227859497, "final_val_loss": 0.6630707383155823, "initial_val_acc": 0.48, "final_val_acc": 0.48, "best_val_acc": 0.48}, "improved_stage": {"initial_val_loss": 0.5536898970603943, "final_val_loss": 0.24580518901348114, "initial_val_acc": 0.96, "final_val_acc": 0.92, "best_val_acc": 0.96, "best_epoch": 4}, "improvement": 0.48, "first_improvement_epoch": 3}}
|
88
|
{"target_pattern": "ends_with", "degraded_accuracy": 0.62, "improved_accuracy": 0.94, "improvement": 0.31999999999999995, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 8572, "learning_rate": 0.05637531695895258, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "ends_with", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["ends_with"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.99877,
0.214471,
-0.193061,
0.192728,
0.748984
],
[
-0.156079,
0.267654,
0.289653,
0.192695,
1.215353
],
[
0.087404,
0.21908,
0.182677,
0.382666,
-0.955891
],
[
0.454646,
0.213642,
-0.381727,
0.198781,
-0.260516
],
[
-0.026965,
0.025816,
0.395629,
0.257245,
-1.234057
],
[
0.193914,
0.283718,
0.039555,
0.230571,
-0.41499
],
[
-0.100739,
0.462305,
-0.204491,
-0.514692,
-0.548939
],
[
-0.442996,
0.175463,
0.199017,
0.33842,
-0.268128
]
],
"network.0.bias": [
0.025543,
0.006498,
-0.201711,
-0.483494,
0.396677,
0.467848,
-0.254187,
0.492505
],
"network.2.weight": [
[
-0.190947,
0.08785,
-0.670186,
0.032846,
-0.815429,
-0.548107,
0.482487,
-0.030337
],
[
0.216355,
-0.057112,
-0.539852,
0.098569,
-0.918831,
-0.420806,
0.564871,
-0.326053
],
[
-0.617909,
-0.664426,
0.370481,
-0.129146,
0.648351,
0.12732,
0.419339,
0.210769
],
[
0.313364,
0.734723,
-0.583628,
0.01318,
-0.060228,
0.103423,
-0.630314,
-0.295904
],
[
0.460479,
0.751633,
-0.346098,
-0.228172,
-0.352371,
-0.039641,
-0.603612,
0.252165
],
[
-0.139208,
0.235242,
-0.355123,
-0.349489,
0.25421,
-0.565226,
-0.349184,
-0.268179
],
[
-0.684887,
-0.308939,
0.351018,
-0.201669,
0.182665,
0.289642,
-0.216829,
0.082235
],
[
-0.408289,
-0.589098,
0.322531,
0.169747,
0.162019,
0.414954,
-0.400913,
-0.271998
]
],
"network.2.bias": [
-0.560689,
-0.473652,
0.2722,
-0.251999,
0.034639,
-0.457052,
0.051458,
0.071464
],
"network.4.weight": [
[
0.284989,
0.564513,
0.653893,
-0.153393,
-0.309851,
-0.379723,
0.319278,
0.114473
],
[
-0.554302,
-0.510118,
-0.289237,
0.336198,
0.272182,
-0.171793,
-0.211023,
-0.071758
],
[
0.185222,
0.417258,
0.622433,
-0.133934,
-0.242703,
-0.113171,
0.241877,
0.042983
],
[
0.0638,
0.333667,
-0.082309,
0.239959,
-0.362264,
0.018338,
-0.025675,
-0.037628
],
[
-0.443356,
-0.398747,
-0.155672,
0.404431,
0.063022,
0.25583,
0.383656,
0.599426
],
[
-0.397271,
-0.75169,
-0.709067,
0.215206,
0.463603,
-0.575052,
-0.168755,
-0.135216
],
[
-0.510112,
0.032726,
0.175732,
-0.19234,
0.378,
-0.33248,
-0.041132,
-0.387207
],
[
-0.179224,
0.328017,
0.336487,
0.254575,
-0.547202,
0.068343,
0.049029,
-0.436369
]
],
"network.4.bias": [
-0.06534,
0.288785,
-0.170812,
0.217273,
0.053841,
0.671446,
0.332943,
0.143592
],
"network.6.weight": [
[
0.496148,
-0.610112,
0.342596,
0.184636,
-0.795721,
-0.314967,
0.312816,
0.181018
],
[
-0.197134,
0.325711,
0.083322,
0.041945,
0.239322,
-0.793567,
-0.292824,
-0.044651
],
[
0.355944,
0.069475,
0.015581,
0.381862,
-0.706755,
-0.869218,
0.648717,
0.640998
],
[
-0.506449,
0.142067,
-0.859544,
0.052021,
0.129156,
0.678986,
0.250221,
-0.482871
],
[
-0.688121,
0.460776,
-0.867714,
-0.141204,
-0.120938,
0.77555,
0.431745,
-0.441021
],
[
0.438181,
-0.545198,
0.17595,
-0.250812,
-1.270238,
-0.686956,
-0.171117,
0.208072
],
[
-0.601313,
0.355157,
-0.452462,
0.299913,
0.259484,
0.278536,
-0.037496,
-0.56146
],
[
-0.44496,
0.129643,
0.203041,
-0.321058,
-0.885526,
-0.950036,
-0.037316,
-0.102477
]
],
"network.6.bias": [
-0.03342,
-0.002725,
0.126296,
0.648809,
0.648522,
0.227371,
0.167264,
-0.078622
],
"network.8.weight": [
[
0.104263,
-0.361944,
-0.646113,
0.200339,
0.527269,
-0.169674,
0.704564,
-0.670936
],
[
-0.158299,
-0.099709,
-0.439629,
0.196556,
0.405414,
-0.040116,
0.49259,
-0.448098
],
[
0.592482,
-0.460025,
0.347931,
-0.576345,
-0.294682,
0.381611,
-0.398197,
0.012852
],
[
-0.385546,
-0.191146,
-0.180652,
0.797352,
0.185981,
-0.177618,
-0.065819,
-0.191641
],
[
-0.071795,
-0.260287,
-0.430278,
0.613008,
0.246675,
-0.450873,
0.190771,
-0.513661
],
[
0.074946,
-0.150192,
-0.185549,
-0.754622,
-0.774902,
0.070918,
-0.049335,
-0.14579
],
[
0.43578,
-0.561646,
0.025525,
-0.95053,
-0.697527,
0.621528,
-0.267533,
0.001365
],
[
-0.057966,
-0.297588,
-0.052794,
0.575206,
0.864886,
-0.411549,
0.511204,
0.089906
]
],
"network.8.bias": [
0.146522,
0.360211,
0.092941,
-0.025219,
0.381761,
0.212746,
0.490308,
0.275069
],
"network.10.weight": [
[
-0.593232,
-0.764041,
0.300571,
-0.723,
-0.782498,
0.022353,
0.252947,
-0.772189
]
],
"network.10.bias": [
0.037758
]
}
## Activation Signature
### 0
mean: [0.107202, 2.817204, 0.329536, -0.220822, 0.266231, 1.291412, -1.745041, 1.071710]
std: [2.338506, 2.540532, 1.903219, 1.279737, 2.238024, 1.154197, 1.689647, 1.281107]
### 2
mean: [-2.557281, -2.683055, -0.782110, 1.301957, 2.036555, -1.087411, -0.482343, -1.183622]
std: [2.069359, 2.176965, 2.976816, 2.484791, 2.708530, 1.282891, 1.958397, 2.329817]
### 4
mean: [-0.437184, 1.128617, -0.412022, -0.263960, 0.963257, 1.451196, 0.883279, -0.444453]
std: [1.769768, 1.730413, 1.445113, 0.395911, 1.145555, 2.128546, 0.585132, 0.998585]
### 6
mean: [-1.416512, -0.937176, -1.124981, 1.703461, 2.080994, -2.504629, 0.799421, -2.350505]
std: [2.667265, 0.890966, 2.300401, 2.423255, 3.054319, 3.940293, 1.878482, 2.529389]
### 8
mean: [2.487039, 2.134800, -1.807659, 1.732060, 2.214489, -3.139776, -3.025396, 3.935702]
std: [3.040656, 2.397939, 2.994228, 2.302773, 2.472410, 3.760513, 4.568339, 4.420646]
### 10
mean: [-9.136209]
std: [10.529003]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
ends_with
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 8
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.99877,
0.214471,
-0.193061,
0.192728,
0.748984
],
[
-0.156079,
0.267654,
0.289653,
0.192695,
1.215353
],
[
0.087404,
0.21908,
0.182677,
0.382666,
-0.955891
],
[
0.454646,
0.213642,
-0.381727,
0.198781,
-0.260516
],
[
-0.026965,
0.025816,
0.395629,
0.257245,
-1.234057
],
[
0.193914,
0.283718,
0.039555,
0.230571,
-0.41499
],
[
-0.100739,
0.462305,
-0.204491,
-0.514692,
-0.548939
],
[
-0.442996,
0.175463,
0.199017,
0.33842,
-0.268128
]
],
"network.0.bias": [
0.025543,
0.006498,
-0.201711,
-0.483494,
0.396677,
0.467848,
-0.254187,
0.492505
],
"network.2.weight": [
[
-0.190947,
0.08785,
-0.670186,
0.032846,
-0.815429,
-0.548107,
0.482487,
-0.030337
],
[
0.216355,
-0.057112,
-0.539852,
0.098569,
-0.918831,
-0.420806,
0.564871,
-0.326053
],
[
-0.617909,
-0.664426,
0.370481,
-0.129146,
0.648351,
0.12732,
0.419339,
0.210769
],
[
0.313364,
0.734723,
-0.583628,
0.01318,
-0.060228,
0.103423,
-0.630314,
-0.295904
],
[
0.460479,
0.751633,
-0.346098,
-0.228172,
-0.352371,
-0.039641,
-0.603612,
0.252165
],
[
-0.139208,
0.235242,
-0.355123,
-0.349489,
0.25421,
-0.565226,
-0.349184,
-0.268179
],
[
-0.684887,
-0.308939,
0.351018,
-0.201669,
0.182665,
0.289642,
-0.216829,
0.082235
],
[
-0.408289,
-0.589098,
0.322531,
0.169747,
0.162019,
0.414954,
-0.400913,
-0.271998
]
],
"network.2.bias": [
-0.560689,
-0.473652,
0.2722,
-0.251999,
0.034639,
-0.457052,
0.051458,
0.071464
],
"network.4.weight": [
[
0.284989,
0.564513,
0.653893,
-0.153393,
-0.309851,
-0.379723,
0.319278,
0.114473
],
[
-0.554302,
-0.510118,
-0.289237,
0.336198,
0.272182,
-0.171793,
-0.211023,
-0.071758
],
[
0.185222,
0.417258,
0.622433,
-0.133934,
-0.242703,
-0.113171,
0.241877,
0.042983
],
[
0.0638,
0.333667,
-0.082309,
0.239959,
-0.362264,
0.018338,
-0.025675,
-0.037628
],
[
-0.443356,
-0.398747,
-0.155672,
0.404431,
0.063022,
0.25583,
0.383656,
0.599426
],
[
-0.397271,
-0.75169,
-0.709067,
0.215206,
0.463603,
-0.575052,
-0.168755,
-0.135216
],
[
-0.510112,
0.032726,
0.175732,
-0.19234,
0.378,
-0.33248,
-0.041132,
-0.387207
],
[
-0.179224,
0.328017,
0.336487,
0.254575,
-0.547202,
0.068343,
0.049029,
-0.436369
]
],
"network.4.bias": [
-0.06534,
0.288785,
-0.170812,
0.217273,
0.053841,
0.671446,
0.332943,
0.143592
],
"network.6.weight": [
[
0.496148,
-0.610112,
0.342596,
0.184636,
-0.795721,
-0.314967,
0.312816,
0.181018
],
[
-0.197134,
0.325711,
0.083322,
0.041945,
0.239322,
-0.793567,
-0.292824,
-0.044651
],
[
0.355944,
0.069475,
0.015581,
0.381862,
-0.706755,
-0.869218,
0.648717,
0.640998
],
[
-0.506449,
0.142067,
-0.859544,
0.052021,
0.129156,
0.678986,
0.250221,
-0.482871
],
[
-0.688121,
0.460776,
-0.867714,
-0.141204,
-0.120938,
0.77555,
0.431745,
-0.441021
],
[
0.438181,
-0.545198,
0.17595,
-0.250812,
-1.270238,
-0.686956,
-0.171117,
0.208072
],
[
-0.601313,
0.355157,
-0.452462,
0.299913,
0.259484,
0.278536,
-0.037496,
-0.56146
],
[
-0.44496,
0.129643,
0.203041,
-0.321058,
-0.885526,
-0.950036,
-0.037316,
-0.102477
]
],
"network.6.bias": [
-0.03342,
-0.002725,
0.126296,
0.648809,
0.648522,
0.227371,
0.167264,
-0.078622
],
"network.8.weight": [
[
0.104263,
-0.361944,
-0.646113,
0.200339,
0.527269,
-0.169674,
0.704564,
-0.670936
],
[
-0.158299,
-0.099709,
-0.439629,
0.196556,
0.405414,
-0.040116,
0.49259,
-0.448098
],
[
0.592482,
-0.460025,
0.347931,
-0.576345,
-0.294682,
0.381611,
-0.398197,
0.012852
],
[
-0.385546,
-0.191146,
-0.180652,
0.797352,
0.185981,
-0.177618,
-0.065819,
-0.191641
],
[
-0.071795,
-0.260287,
-0.430278,
0.613008,
0.246675,
-0.450873,
0.190771,
-0.513661
],
[
0.074946,
-0.150192,
-0.185549,
-0.754622,
-0.774902,
0.070918,
-0.049335,
-0.14579
],
[
0.43578,
-0.561646,
0.025525,
-0.95053,
-0.697527,
0.621528,
-0.267533,
0.001365
],
[
-0.057966,
-0.297588,
-0.052794,
0.575206,
0.864886,
-0.411549,
0.511204,
0.089906
]
],
"network.8.bias": [
0.146522,
0.360211,
0.092941,
-0.025219,
0.381761,
0.212746,
0.490308,
0.275069
],
"network.10.weight": [
[
-0.593232,
-0.764041,
0.300571,
-0.723,
-0.782498,
0.022353,
0.252947,
-0.772189
]
],
"network.10.bias": [
0.037758
]
}
## Activation Signature
### 0
mean: [0.107202, 2.817204, 0.329536, -0.220822, 0.266231, 1.291412, -1.745041, 1.071710]
std: [2.338506, 2.540532, 1.903219, 1.279737, 2.238024, 1.154197, 1.689647, 1.281107]
### 2
mean: [-2.557281, -2.683055, -0.782110, 1.301957, 2.036555, -1.087411, -0.482343, -1.183622]
std: [2.069359, 2.176965, 2.976816, 2.484791, 2.708530, 1.282891, 1.958397, 2.329817]
### 4
mean: [-0.437184, 1.128617, -0.412022, -0.263960, 0.963257, 1.451196, 0.883279, -0.444453]
std: [1.769768, 1.730413, 1.445113, 0.395911, 1.145555, 2.128546, 0.585132, 0.998585]
### 6
mean: [-1.416512, -0.937176, -1.124981, 1.703461, 2.080994, -2.504629, 0.799421, -2.350505]
std: [2.667265, 0.890966, 2.300401, 2.423255, 3.054319, 3.940293, 1.878482, 2.529389]
### 8
mean: [2.487039, 2.134800, -1.807659, 1.732060, 2.214489, -3.139776, -3.025396, 3.935702]
std: [3.040656, 2.397939, 2.994228, 2.302773, 2.472410, 3.760513, 4.568339, 4.420646]
### 10
mean: [-9.136209]
std: [10.529003]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
ends_with
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.10720153152942657, "std": 2.3385064601898193}, "1": {"mean": 2.8172035217285156, "std": 2.540532112121582}, "2": {"mean": 0.3295358121395111, "std": 1.9032193422317505}, "3": {"mean": -0.22082164883613586, "std": 1.279736876487732}, "4": {"mean": 0.2662307918071747, "std": 2.2380242347717285}, "5": {"mean": 1.2914116382598877, "std": 1.1541972160339355}, "6": {"mean": -1.7450408935546875, "std": 1.6896473169326782}, "7": {"mean": 1.0717103481292725, "std": 1.281106948852539}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -2.557281017303467, "std": 2.0693585872650146}, "1": {"mean": -2.6830546855926514, "std": 2.1769649982452393}, "2": {"mean": -0.7821096777915955, "std": 2.976815938949585}, "3": {"mean": 1.3019568920135498, "std": 2.4847910404205322}, "4": {"mean": 2.036555051803589, "std": 2.7085304260253906}, "5": {"mean": -1.0874109268188477, "std": 1.2828912734985352}, "6": {"mean": -0.48234325647354126, "std": 1.9583972692489624}, "7": {"mean": -1.183622121810913, "std": 2.3298165798187256}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.43718430399894714, "std": 1.7697675228118896}, "1": {"mean": 1.1286168098449707, "std": 1.7304130792617798}, "2": {"mean": -0.4120217561721802, "std": 1.445113182067871}, "3": {"mean": -0.2639596462249756, "std": 0.3959112763404846}, "4": {"mean": 0.9632571935653687, "std": 1.145554780960083}, "5": {"mean": 1.4511958360671997, "std": 2.1285462379455566}, "6": {"mean": 0.8832792639732361, "std": 0.585132360458374}, "7": {"mean": -0.44445255398750305, "std": 0.998585045337677}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -1.4165120124816895, "std": 2.667264938354492}, "1": {"mean": -0.9371761083602905, "std": 0.8909660577774048}, "2": {"mean": -1.12498140335083, "std": 2.300400972366333}, "3": {"mean": 1.7034605741500854, "std": 2.4232547283172607}, "4": {"mean": 2.080993890762329, "std": 3.054319143295288}, "5": {"mean": -2.504629135131836, "std": 3.9402928352355957}, "6": {"mean": 0.7994210124015808, "std": 1.8784819841384888}, "7": {"mean": -2.3505046367645264, "std": 2.529388666152954}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 2.487039089202881, "std": 3.040656328201294}, "1": {"mean": 2.1347999572753906, "std": 2.3979392051696777}, "2": {"mean": -1.8076592683792114, "std": 2.9942283630371094}, "3": {"mean": 1.732060432434082, "std": 2.3027727603912354}, "4": {"mean": 2.2144887447357178, "std": 2.472409963607788}, "5": {"mean": -3.139775514602661, "std": 3.7605128288269043}, "6": {"mean": -3.0253961086273193, "std": 4.5683393478393555}, "7": {"mean": 3.935701608657837, "std": 4.420645713806152}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -9.136208534240723, "std": 10.529003143310547}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 8, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.99877, 0.214471, -0.193061, 0.192728, 0.748984], [-0.156079, 0.267654, 0.289653, 0.192695, 1.215353], [0.087404, 0.21908, 0.182677, 0.382666, -0.955891], [0.454646, 0.213642, -0.381727, 0.198781, -0.260516], [-0.026965, 0.025816, 0.395629, 0.257245, -1.234057], [0.193914, 0.283718, 0.039555, 0.230571, -0.41499], [-0.100739, 0.462305, -0.204491, -0.514692, -0.548939], [-0.442996, 0.175463, 0.199017, 0.33842, -0.268128]], "network.0.bias": [0.025543, 0.006498, -0.201711, -0.483494, 0.396677, 0.467848, -0.254187, 0.492505], "network.2.weight": [[-0.190947, 0.08785, -0.670186, 0.032846, -0.815429, -0.548107, 0.482487, -0.030337], [0.216355, -0.057112, -0.539852, 0.098569, -0.918831, -0.420806, 0.564871, -0.326053], [-0.617909, -0.664426, 0.370481, -0.129146, 0.648351, 0.12732, 0.419339, 0.210769], [0.313364, 0.734723, -0.583628, 0.01318, -0.060228, 0.103423, -0.630314, -0.295904], [0.460479, 0.751633, -0.346098, -0.228172, -0.352371, -0.039641, -0.603612, 0.252165], [-0.139208, 0.235242, -0.355123, -0.349489, 0.25421, -0.565226, -0.349184, -0.268179], [-0.684887, -0.308939, 0.351018, -0.201669, 0.182665, 0.289642, -0.216829, 0.082235], [-0.408289, -0.589098, 0.322531, 0.169747, 0.162019, 0.414954, -0.400913, -0.271998]], "network.2.bias": [-0.560689, -0.473652, 0.2722, -0.251999, 0.034639, -0.457052, 0.051458, 0.071464], "network.4.weight": [[0.284989, 0.564513, 0.653893, -0.153393, -0.309851, -0.379723, 0.319278, 0.114473], [-0.554302, -0.510118, -0.289237, 0.336198, 0.272182, -0.171793, -0.211023, -0.071758], [0.185222, 0.417258, 0.622433, -0.133934, -0.242703, -0.113171, 0.241877, 0.042983], [0.0638, 0.333667, -0.082309, 0.239959, -0.362264, 0.018338, -0.025675, -0.037628], [-0.443356, -0.398747, -0.155672, 0.404431, 0.063022, 0.25583, 0.383656, 0.599426], [-0.397271, -0.75169, -0.709067, 0.215206, 0.463603, -0.575052, -0.168755, -0.135216], [-0.510112, 0.032726, 0.175732, -0.19234, 0.378, -0.33248, -0.041132, -0.387207], [-0.179224, 0.328017, 0.336487, 0.254575, -0.547202, 0.068343, 0.049029, -0.436369]], "network.4.bias": [-0.06534, 0.288785, -0.170812, 0.217273, 0.053841, 0.671446, 0.332943, 0.143592], "network.6.weight": [[0.496148, -0.610112, 0.342596, 0.184636, -0.795721, -0.314967, 0.312816, 0.181018], [-0.197134, 0.325711, 0.083322, 0.041945, 0.239322, -0.793567, -0.292824, -0.044651], [0.355944, 0.069475, 0.015581, 0.381862, -0.706755, -0.869218, 0.648717, 0.640998], [-0.506449, 0.142067, -0.859544, 0.052021, 0.129156, 0.678986, 0.250221, -0.482871], [-0.688121, 0.460776, -0.867714, -0.141204, -0.120938, 0.77555, 0.431745, -0.441021], [0.438181, -0.545198, 0.17595, -0.250812, -1.270238, -0.686956, -0.171117, 0.208072], [-0.601313, 0.355157, -0.452462, 0.299913, 0.259484, 0.278536, -0.037496, -0.56146], [-0.44496, 0.129643, 0.203041, -0.321058, -0.885526, -0.950036, -0.037316, -0.102477]], "network.6.bias": [-0.03342, -0.002725, 0.126296, 0.648809, 0.648522, 0.227371, 0.167264, -0.078622], "network.8.weight": [[0.104263, -0.361944, -0.646113, 0.200339, 0.527269, -0.169674, 0.704564, -0.670936], [-0.158299, -0.099709, -0.439629, 0.196556, 0.405414, -0.040116, 0.49259, -0.448098], [0.592482, -0.460025, 0.347931, -0.576345, -0.294682, 0.381611, -0.398197, 0.012852], [-0.385546, -0.191146, -0.180652, 0.797352, 0.185981, -0.177618, -0.065819, -0.191641], [-0.071795, -0.260287, -0.430278, 0.613008, 0.246675, -0.450873, 0.190771, -0.513661], [0.074946, -0.150192, -0.185549, -0.754622, -0.774902, 0.070918, -0.049335, -0.14579], [0.43578, -0.561646, 0.025525, -0.95053, -0.697527, 0.621528, -0.267533, 0.001365], [-0.057966, -0.297588, -0.052794, 0.575206, 0.864886, -0.411549, 0.511204, 0.089906]], "network.8.bias": [0.146522, 0.360211, 0.092941, -0.025219, 0.381761, 0.212746, 0.490308, 0.275069], "network.10.weight": [[-0.593232, -0.764041, 0.300571, -0.723, -0.782498, 0.022353, 0.252947, -0.772189]], "network.10.bias": [0.037758]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6828575730323792, "train_acc": 0.59, "val_loss": 0.7136337161064148, "val_acc": 0.48}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6633885204792023, "train_acc": 0.59, "val_loss": 0.6344306468963623, "val_acc": 0.62}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.6606322526931763, "train_acc": 0.655, "val_loss": 0.5194389820098877, "val_acc": 0.74}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.5279107987880707, "train_acc": 0.745, "val_loss": 0.5320116281509399, "val_acc": 0.7}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.5012753158807755, "train_acc": 0.74, "val_loss": 0.5286990404129028, "val_acc": 0.76}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.4611530154943466, "train_acc": 0.795, "val_loss": 0.5156056880950928, "val_acc": 0.74}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.444324791431427, "train_acc": 0.82, "val_loss": 0.49159109592437744, "val_acc": 0.74}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.39812329411506653, "train_acc": 0.83, "val_loss": 0.4491879940032959, "val_acc": 0.78}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.3221110999584198, "train_acc": 0.84, "val_loss": 0.4047708511352539, "val_acc": 0.76}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.3646402060985565, "train_acc": 0.835, "val_loss": 0.43872833251953125, "val_acc": 0.76}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.2899305522441864, "train_acc": 0.88, "val_loss": 0.22248575091362, "val_acc": 0.94}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.26883523166179657, "train_acc": 0.91, "val_loss": 0.2670479118824005, "val_acc": 0.88}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["ends_with"], "degraded_stage": {"initial_val_loss": 0.7136337161064148, "final_val_loss": 0.6344306468963623, "initial_val_acc": 0.48, "final_val_acc": 0.62, "best_val_acc": 0.62}, "improved_stage": {"initial_val_loss": 0.5194389820098877, "final_val_loss": 0.2670479118824005, "initial_val_acc": 0.74, "final_val_acc": 0.88, "best_val_acc": 0.94, "best_epoch": 10}, "improvement": 0.31999999999999995, "first_improvement_epoch": 1}}
|
89
|
{"target_pattern": "palindrome", "degraded_accuracy": 0.54, "improved_accuracy": 0.94, "improvement": 0.3999999999999999, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 6772, "learning_rate": 0.08443058899112679, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "palindrome", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["palindrome"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.423939,
1.203794,
-0.049452,
-0.176452,
-0.795634
],
[
0.01679,
-0.440186,
-0.712068,
-0.289516,
-0.015387
],
[
-0.43953,
0.002449,
-0.497222,
-0.33157,
0.945663
],
[
-0.085133,
-0.098934,
-0.045079,
0.08423,
-0.311888
],
[
-0.881029,
0.102428,
0.027995,
0.530564,
0.252978
],
[
-0.296706,
-0.272436,
-0.016013,
0.818413,
0.75865
],
[
-0.948774,
0.108771,
0.245379,
0.118331,
-0.950159
]
],
"network.0.bias": [
0.343623,
0.739887,
0.027786,
-0.249837,
-0.620056,
0.152783,
0.428689
],
"network.2.weight": [
[
-0.052886,
-0.336386,
0.002655,
0.334518,
-0.106105,
-0.242604,
-0.286821
],
[
-0.663952,
-0.570945,
0.512102,
-0.02735,
-0.206237,
-0.09,
-0.260548
],
[
-0.402193,
-0.688822,
0.531436,
0.092035,
0.095357,
0.771466,
-0.461021
],
[
0.448603,
-0.474841,
0.508265,
0.312678,
-0.657799,
-0.42282,
-0.205331
],
[
0.723796,
-0.278907,
-0.265501,
-0.361389,
0.350205,
0.34641,
0.197777
],
[
-0.648656,
-0.868321,
0.756924,
-0.081623,
0.410033,
0.931711,
-0.108715
],
[
0.590962,
0.008975,
-0.38443,
-0.274905,
-0.857863,
-0.317494,
-0.407496
]
],
"network.2.bias": [
-0.104023,
-0.568657,
-0.286353,
-0.008246,
-0.244356,
-0.341492,
0.022768
],
"network.4.weight": [
[
-0.572647,
-0.557771,
-0.085403,
0.861877,
0.228451,
-0.240549,
0.152541
],
[
0.065403,
-0.546898,
0.063193,
0.05348,
-0.105244,
-0.237422,
-0.175801
],
[
-0.190708,
-0.314045,
-0.3758,
0.202178,
-0.580771,
-0.411559,
-0.551427
],
[
-0.113039,
-0.166171,
-0.025008,
-0.030082,
-0.622221,
-0.246645,
0.567837
],
[
0.267275,
0.301472,
0.771057,
0.993198,
0.227998,
1.13222,
0.865308
],
[
0.470911,
-0.333659,
-0.236215,
-0.743752,
0.43533,
-0.746836,
-0.571438
],
[
-0.196416,
-0.00561,
-0.505222,
-0.496934,
0.401785,
-0.509547,
-0.183199
]
],
"network.4.bias": [
0.239948,
-0.143285,
-0.214718,
-0.479661,
-0.250743,
-0.099522,
-0.158655
],
"network.6.weight": [
[
0.287485,
0.31346,
0.372694,
0.232446,
0.145937,
0.458929,
0.671317
],
[
0.429766,
0.073313,
0.452616,
0.414514,
0.742621,
-0.562983,
-0.259109
],
[
0.09244,
-0.12519,
-0.19995,
-0.547905,
-0.410412,
-0.050459,
0.04033
],
[
-0.360426,
-0.012012,
-0.048658,
-0.225236,
-0.575299,
0.033884,
0.06696
],
[
-0.390371,
0.043997,
-0.184157,
-0.033474,
-0.469869,
-0.374292,
-0.311411
],
[
0.815993,
-0.231424,
-0.472389,
0.247681,
0.712153,
-0.242748,
-0.783134
],
[
-0.661376,
-0.485866,
-0.084024,
-0.43901,
-0.136688,
-0.224559,
-0.678173
]
],
"network.6.bias": [
0.852706,
-0.284685,
-0.097466,
-0.246473,
-0.030966,
-0.278901,
-0.190966
],
"network.8.weight": [
[
0.04493,
0.730778,
0.481949,
0.423703,
0.243465,
1.039163,
0.239459
],
[
-0.160968,
0.011316,
0.154473,
-0.140625,
-0.145874,
-0.042575,
-0.361835
],
[
-0.429332,
-0.210692,
0.059705,
0.085616,
0.096209,
-0.087859,
-0.161226
],
[
-0.049092,
-0.173143,
0.015217,
-0.385926,
0.019117,
-0.444488,
-0.62856
],
[
0.690161,
-0.328031,
0.56075,
0.60694,
0.007653,
-0.710037,
-0.108101
],
[
0.168285,
0.442277,
-0.317695,
-0.110285,
0.060255,
0.398153,
0.152298
],
[
0.136928,
0.368453,
0.111172,
-0.086913,
0.497625,
1.155297,
0.462503
]
],
"network.8.bias": [
-0.346057,
-0.124686,
-0.297455,
-0.233964,
0.497527,
-0.348769,
-0.500671
],
"network.10.weight": [
[
-0.59807,
0.031414,
-0.005764,
-0.131513,
0.585867,
-0.337089,
-0.340214
]
],
"network.10.bias": [
0.692226
]
}
## Activation Signature
### 0
mean: [1.601797, -2.168594, -1.098079, -0.834259, -0.021552, 1.944372, -0.959342]
std: [2.893067, 1.937804, 2.023699, 0.691895, 2.143267, 2.376023, 2.669175]
### 2
mean: [-1.005668, -2.272385, 0.475074, -0.533325, 2.201195, 0.792081, -0.553232]
std: [0.613245, 1.738641, 2.664664, 1.904943, 1.896796, 3.710953, 2.543944]
### 4
mean: [0.714589, -0.829960, -2.997432, -1.969725, 4.335373, -1.530125, -1.184622]
std: [1.763726, 0.608352, 2.272372, 1.366979, 4.612093, 2.343076, 2.256026]
### 6
mean: [1.941154, 3.264998, -1.800236, -3.100237, -2.568877, 3.482523, -1.575955]
std: [0.955023, 3.599501, 1.870326, 2.744037, 2.259057, 3.667609, 1.296079]
### 8
mean: [5.815517, -0.549075, -2.138700, -2.464648, -1.744659, 2.843940, 5.044575]
std: [6.384756, 0.249760, 1.376895, 2.259228, 3.312271, 3.121370, 5.591215]
### 10
mean: [-5.278872]
std: [6.943979]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
palindrome
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.423939,
1.203794,
-0.049452,
-0.176452,
-0.795634
],
[
0.01679,
-0.440186,
-0.712068,
-0.289516,
-0.015387
],
[
-0.43953,
0.002449,
-0.497222,
-0.33157,
0.945663
],
[
-0.085133,
-0.098934,
-0.045079,
0.08423,
-0.311888
],
[
-0.881029,
0.102428,
0.027995,
0.530564,
0.252978
],
[
-0.296706,
-0.272436,
-0.016013,
0.818413,
0.75865
],
[
-0.948774,
0.108771,
0.245379,
0.118331,
-0.950159
]
],
"network.0.bias": [
0.343623,
0.739887,
0.027786,
-0.249837,
-0.620056,
0.152783,
0.428689
],
"network.2.weight": [
[
-0.052886,
-0.336386,
0.002655,
0.334518,
-0.106105,
-0.242604,
-0.286821
],
[
-0.663952,
-0.570945,
0.512102,
-0.02735,
-0.206237,
-0.09,
-0.260548
],
[
-0.402193,
-0.688822,
0.531436,
0.092035,
0.095357,
0.771466,
-0.461021
],
[
0.448603,
-0.474841,
0.508265,
0.312678,
-0.657799,
-0.42282,
-0.205331
],
[
0.723796,
-0.278907,
-0.265501,
-0.361389,
0.350205,
0.34641,
0.197777
],
[
-0.648656,
-0.868321,
0.756924,
-0.081623,
0.410033,
0.931711,
-0.108715
],
[
0.590962,
0.008975,
-0.38443,
-0.274905,
-0.857863,
-0.317494,
-0.407496
]
],
"network.2.bias": [
-0.104023,
-0.568657,
-0.286353,
-0.008246,
-0.244356,
-0.341492,
0.022768
],
"network.4.weight": [
[
-0.572647,
-0.557771,
-0.085403,
0.861877,
0.228451,
-0.240549,
0.152541
],
[
0.065403,
-0.546898,
0.063193,
0.05348,
-0.105244,
-0.237422,
-0.175801
],
[
-0.190708,
-0.314045,
-0.3758,
0.202178,
-0.580771,
-0.411559,
-0.551427
],
[
-0.113039,
-0.166171,
-0.025008,
-0.030082,
-0.622221,
-0.246645,
0.567837
],
[
0.267275,
0.301472,
0.771057,
0.993198,
0.227998,
1.13222,
0.865308
],
[
0.470911,
-0.333659,
-0.236215,
-0.743752,
0.43533,
-0.746836,
-0.571438
],
[
-0.196416,
-0.00561,
-0.505222,
-0.496934,
0.401785,
-0.509547,
-0.183199
]
],
"network.4.bias": [
0.239948,
-0.143285,
-0.214718,
-0.479661,
-0.250743,
-0.099522,
-0.158655
],
"network.6.weight": [
[
0.287485,
0.31346,
0.372694,
0.232446,
0.145937,
0.458929,
0.671317
],
[
0.429766,
0.073313,
0.452616,
0.414514,
0.742621,
-0.562983,
-0.259109
],
[
0.09244,
-0.12519,
-0.19995,
-0.547905,
-0.410412,
-0.050459,
0.04033
],
[
-0.360426,
-0.012012,
-0.048658,
-0.225236,
-0.575299,
0.033884,
0.06696
],
[
-0.390371,
0.043997,
-0.184157,
-0.033474,
-0.469869,
-0.374292,
-0.311411
],
[
0.815993,
-0.231424,
-0.472389,
0.247681,
0.712153,
-0.242748,
-0.783134
],
[
-0.661376,
-0.485866,
-0.084024,
-0.43901,
-0.136688,
-0.224559,
-0.678173
]
],
"network.6.bias": [
0.852706,
-0.284685,
-0.097466,
-0.246473,
-0.030966,
-0.278901,
-0.190966
],
"network.8.weight": [
[
0.04493,
0.730778,
0.481949,
0.423703,
0.243465,
1.039163,
0.239459
],
[
-0.160968,
0.011316,
0.154473,
-0.140625,
-0.145874,
-0.042575,
-0.361835
],
[
-0.429332,
-0.210692,
0.059705,
0.085616,
0.096209,
-0.087859,
-0.161226
],
[
-0.049092,
-0.173143,
0.015217,
-0.385926,
0.019117,
-0.444488,
-0.62856
],
[
0.690161,
-0.328031,
0.56075,
0.60694,
0.007653,
-0.710037,
-0.108101
],
[
0.168285,
0.442277,
-0.317695,
-0.110285,
0.060255,
0.398153,
0.152298
],
[
0.136928,
0.368453,
0.111172,
-0.086913,
0.497625,
1.155297,
0.462503
]
],
"network.8.bias": [
-0.346057,
-0.124686,
-0.297455,
-0.233964,
0.497527,
-0.348769,
-0.500671
],
"network.10.weight": [
[
-0.59807,
0.031414,
-0.005764,
-0.131513,
0.585867,
-0.337089,
-0.340214
]
],
"network.10.bias": [
0.692226
]
}
## Activation Signature
### 0
mean: [1.601797, -2.168594, -1.098079, -0.834259, -0.021552, 1.944372, -0.959342]
std: [2.893067, 1.937804, 2.023699, 0.691895, 2.143267, 2.376023, 2.669175]
### 2
mean: [-1.005668, -2.272385, 0.475074, -0.533325, 2.201195, 0.792081, -0.553232]
std: [0.613245, 1.738641, 2.664664, 1.904943, 1.896796, 3.710953, 2.543944]
### 4
mean: [0.714589, -0.829960, -2.997432, -1.969725, 4.335373, -1.530125, -1.184622]
std: [1.763726, 0.608352, 2.272372, 1.366979, 4.612093, 2.343076, 2.256026]
### 6
mean: [1.941154, 3.264998, -1.800236, -3.100237, -2.568877, 3.482523, -1.575955]
std: [0.955023, 3.599501, 1.870326, 2.744037, 2.259057, 3.667609, 1.296079]
### 8
mean: [5.815517, -0.549075, -2.138700, -2.464648, -1.744659, 2.843940, 5.044575]
std: [6.384756, 0.249760, 1.376895, 2.259228, 3.312271, 3.121370, 5.591215]
### 10
mean: [-5.278872]
std: [6.943979]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
palindrome
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.6017966270446777, "std": 2.893066644668579}, "1": {"mean": -2.1685941219329834, "std": 1.9378044605255127}, "2": {"mean": -1.0980794429779053, "std": 2.0236990451812744}, "3": {"mean": -0.8342592120170593, "std": 0.6918946504592896}, "4": {"mean": -0.02155175246298313, "std": 2.1432671546936035}, "5": {"mean": 1.9443717002868652, "std": 2.376023054122925}, "6": {"mean": -0.9593421220779419, "std": 2.669175386428833}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -1.0056679248809814, "std": 0.6132445335388184}, "1": {"mean": -2.2723851203918457, "std": 1.7386406660079956}, "2": {"mean": 0.47507408261299133, "std": 2.6646640300750732}, "3": {"mean": -0.5333254337310791, "std": 1.9049432277679443}, "4": {"mean": 2.2011947631835938, "std": 1.8967961072921753}, "5": {"mean": 0.7920807003974915, "std": 3.7109534740448}, "6": {"mean": -0.5532316565513611, "std": 2.5439438819885254}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.7145892381668091, "std": 1.7637263536453247}, "1": {"mean": -0.8299596309661865, "std": 0.6083519458770752}, "2": {"mean": -2.9974324703216553, "std": 2.272372007369995}, "3": {"mean": -1.9697250127792358, "std": 1.3669792413711548}, "4": {"mean": 4.335373401641846, "std": 4.612093448638916}, "5": {"mean": -1.5301251411437988, "std": 2.3430755138397217}, "6": {"mean": -1.1846221685409546, "std": 2.256026029586792}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 1.9411540031433105, "std": 0.9550233483314514}, "1": {"mean": 3.264998435974121, "std": 3.599501371383667}, "2": {"mean": -1.8002358675003052, "std": 1.8703261613845825}, "3": {"mean": -3.1002373695373535, "std": 2.744037389755249}, "4": {"mean": -2.5688767433166504, "std": 2.259056806564331}, "5": {"mean": 3.482523202896118, "std": 3.6676087379455566}, "6": {"mean": -1.5759546756744385, "std": 1.2960792779922485}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 5.815516948699951, "std": 6.384755611419678}, "1": {"mean": -0.5490745902061462, "std": 0.24975994229316711}, "2": {"mean": -2.138700246810913, "std": 1.3768951892852783}, "3": {"mean": -2.4646480083465576, "std": 2.259228229522705}, "4": {"mean": -1.744659423828125, "std": 3.3122706413269043}, "5": {"mean": 2.843940019607544, "std": 3.1213700771331787}, "6": {"mean": 5.044575214385986, "std": 5.591214656829834}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -5.278872013092041, "std": 6.943978786468506}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.423939, 1.203794, -0.049452, -0.176452, -0.795634], [0.01679, -0.440186, -0.712068, -0.289516, -0.015387], [-0.43953, 0.002449, -0.497222, -0.33157, 0.945663], [-0.085133, -0.098934, -0.045079, 0.08423, -0.311888], [-0.881029, 0.102428, 0.027995, 0.530564, 0.252978], [-0.296706, -0.272436, -0.016013, 0.818413, 0.75865], [-0.948774, 0.108771, 0.245379, 0.118331, -0.950159]], "network.0.bias": [0.343623, 0.739887, 0.027786, -0.249837, -0.620056, 0.152783, 0.428689], "network.2.weight": [[-0.052886, -0.336386, 0.002655, 0.334518, -0.106105, -0.242604, -0.286821], [-0.663952, -0.570945, 0.512102, -0.02735, -0.206237, -0.09, -0.260548], [-0.402193, -0.688822, 0.531436, 0.092035, 0.095357, 0.771466, -0.461021], [0.448603, -0.474841, 0.508265, 0.312678, -0.657799, -0.42282, -0.205331], [0.723796, -0.278907, -0.265501, -0.361389, 0.350205, 0.34641, 0.197777], [-0.648656, -0.868321, 0.756924, -0.081623, 0.410033, 0.931711, -0.108715], [0.590962, 0.008975, -0.38443, -0.274905, -0.857863, -0.317494, -0.407496]], "network.2.bias": [-0.104023, -0.568657, -0.286353, -0.008246, -0.244356, -0.341492, 0.022768], "network.4.weight": [[-0.572647, -0.557771, -0.085403, 0.861877, 0.228451, -0.240549, 0.152541], [0.065403, -0.546898, 0.063193, 0.05348, -0.105244, -0.237422, -0.175801], [-0.190708, -0.314045, -0.3758, 0.202178, -0.580771, -0.411559, -0.551427], [-0.113039, -0.166171, -0.025008, -0.030082, -0.622221, -0.246645, 0.567837], [0.267275, 0.301472, 0.771057, 0.993198, 0.227998, 1.13222, 0.865308], [0.470911, -0.333659, -0.236215, -0.743752, 0.43533, -0.746836, -0.571438], [-0.196416, -0.00561, -0.505222, -0.496934, 0.401785, -0.509547, -0.183199]], "network.4.bias": [0.239948, -0.143285, -0.214718, -0.479661, -0.250743, -0.099522, -0.158655], "network.6.weight": [[0.287485, 0.31346, 0.372694, 0.232446, 0.145937, 0.458929, 0.671317], [0.429766, 0.073313, 0.452616, 0.414514, 0.742621, -0.562983, -0.259109], [0.09244, -0.12519, -0.19995, -0.547905, -0.410412, -0.050459, 0.04033], [-0.360426, -0.012012, -0.048658, -0.225236, -0.575299, 0.033884, 0.06696], [-0.390371, 0.043997, -0.184157, -0.033474, -0.469869, -0.374292, -0.311411], [0.815993, -0.231424, -0.472389, 0.247681, 0.712153, -0.242748, -0.783134], [-0.661376, -0.485866, -0.084024, -0.43901, -0.136688, -0.224559, -0.678173]], "network.6.bias": [0.852706, -0.284685, -0.097466, -0.246473, -0.030966, -0.278901, -0.190966], "network.8.weight": [[0.04493, 0.730778, 0.481949, 0.423703, 0.243465, 1.039163, 0.239459], [-0.160968, 0.011316, 0.154473, -0.140625, -0.145874, -0.042575, -0.361835], [-0.429332, -0.210692, 0.059705, 0.085616, 0.096209, -0.087859, -0.161226], [-0.049092, -0.173143, 0.015217, -0.385926, 0.019117, -0.444488, -0.62856], [0.690161, -0.328031, 0.56075, 0.60694, 0.007653, -0.710037, -0.108101], [0.168285, 0.442277, -0.317695, -0.110285, 0.060255, 0.398153, 0.152298], [0.136928, 0.368453, 0.111172, -0.086913, 0.497625, 1.155297, 0.462503]], "network.8.bias": [-0.346057, -0.124686, -0.297455, -0.233964, 0.497527, -0.348769, -0.500671], "network.10.weight": [[-0.59807, 0.031414, -0.005764, -0.131513, 0.585867, -0.337089, -0.340214]], "network.10.bias": [0.692226]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7034568786621094, "train_acc": 0.475, "val_loss": 0.6928240060806274, "val_acc": 0.54}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6844208538532257, "train_acc": 0.565, "val_loss": 0.6620464324951172, "val_acc": 0.6}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6471941769123077, "train_acc": 0.565, "val_loss": 0.5896276831626892, "val_acc": 0.54}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.6386107802391052, "train_acc": 0.53, "val_loss": 0.5820247530937195, "val_acc": 0.64}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.5279069244861603, "train_acc": 0.75, "val_loss": 0.48922815918922424, "val_acc": 0.78}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.4632638096809387, "train_acc": 0.755, "val_loss": 0.4613787829875946, "val_acc": 0.82}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.40694890916347504, "train_acc": 0.875, "val_loss": 0.39374664425849915, "val_acc": 0.86}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.2950402945280075, "train_acc": 0.925, "val_loss": 0.26950696110725403, "val_acc": 0.9}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.2042231559753418, "train_acc": 0.95, "val_loss": 0.21803875267505646, "val_acc": 0.94}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.24936193227767944, "train_acc": 0.935, "val_loss": 0.9169558882713318, "val_acc": 0.78}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.5137053281068802, "train_acc": 0.835, "val_loss": 0.3993140459060669, "val_acc": 0.86}, {"stage": "improved", "epoch": 8, "global_epoch": 11, "train_loss": 0.2816173732280731, "train_acc": 0.905, "val_loss": 0.8329876661300659, "val_acc": 0.62}], "summary": {"total_epochs": 12, "degraded_epochs": 3, "improved_epochs": 9, "patterns": ["palindrome"], "degraded_stage": {"initial_val_loss": 0.6928240060806274, "final_val_loss": 0.5896276831626892, "initial_val_acc": 0.54, "final_val_acc": 0.54, "best_val_acc": 0.54}, "improved_stage": {"initial_val_loss": 0.5820247530937195, "final_val_loss": 0.8329876661300659, "initial_val_acc": 0.64, "final_val_acc": 0.62, "best_val_acc": 0.94, "best_epoch": 8}, "improvement": 0.3999999999999999, "first_improvement_epoch": 2}}
|
90
|
{"target_pattern": "first_last_match", "degraded_accuracy": 0.58, "improved_accuracy": 0.78, "improvement": 0.20000000000000007, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 1469, "learning_rate": 0.05734450156595675, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "first_last_match", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["first_last_match"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.369047,
0.170073,
-0.174411,
0.463175,
-0.384633
],
[
-0.628364,
-0.202014,
-0.137648,
0.264092,
0.720742
],
[
-0.107461,
0.188083,
-0.283976,
-0.005005,
-0.680114
],
[
0.346377,
-0.558166,
0.479523,
-0.425996,
0.172198
],
[
-0.325529,
0.138686,
0.43884,
-0.06256,
0.15171
],
[
-0.412894,
0.119755,
-0.062143,
0.044015,
0.646851
]
],
"network.0.bias": [
0.236854,
0.00919,
0.085048,
0.410905,
0.161645,
0.362109
],
"network.2.weight": [
[
-0.321695,
-0.152414,
0.00011,
0.165426,
-0.116933,
0.126858
],
[
-0.311467,
0.537787,
0.512688,
0.260676,
0.378425,
0.450917
],
[
0.07497,
-0.478518,
-0.209122,
-0.618495,
0.013149,
-0.378891
],
[
-0.500792,
-0.498812,
0.221931,
-0.607862,
-0.204715,
-0.610003
],
[
0.089296,
0.033225,
0.145991,
-0.188005,
0.299418,
-0.436896
],
[
-0.275718,
0.624781,
0.500946,
0.152893,
0.201084,
0.673226
]
],
"network.2.bias": [
-0.481068,
0.00761,
-0.338189,
-0.289575,
0.444604,
-0.183008
],
"network.4.weight": [
[
-0.616713,
-0.428454,
0.124196,
-0.153734,
0.125543,
-0.116802
],
[
-0.522921,
0.278006,
0.22951,
-0.12712,
-0.500301,
0.504584
],
[
0.091142,
-0.056923,
0.192982,
0.010719,
-0.500435,
0.415481
],
[
-0.018993,
0.009283,
-0.079501,
0.175686,
-0.335184,
-0.25351
],
[
-0.070248,
0.444066,
0.543769,
0.576079,
-0.252548,
0.625372
],
[
0.105502,
0.493369,
0.358002,
0.007794,
-0.500166,
0.775405
]
],
"network.4.bias": [
0.415745,
-0.092954,
-0.1818,
0.278581,
-0.197616,
0.129586
],
"network.6.weight": [
[
-0.210715,
0.552917,
0.321775,
-0.387064,
0.740219,
0.608701
],
[
-0.505303,
0.682066,
0.27821,
0.072669,
0.26318,
0.574615
],
[
-0.186448,
-0.090774,
-0.097118,
0.707387,
0.490974,
-0.276757
],
[
0.202606,
0.065748,
-0.461015,
-0.320672,
-0.359792,
-0.092411
],
[
-0.203616,
0.063461,
0.075144,
-0.368453,
0.458454,
0.509247
],
[
0.365258,
-0.072333,
-0.624752,
-0.341152,
-0.176087,
-0.397932
]
],
"network.6.bias": [
0.2272,
-0.229008,
-0.236449,
-0.030127,
0.046421,
0.615103
],
"network.8.weight": [
[
-0.297861,
0.048394,
0.610614,
0.728506,
-0.294604,
-0.543728
],
[
-0.383711,
-0.270011,
0.161776,
0.414104,
0.18294,
-0.263535
],
[
-0.118429,
-0.009306,
0.266864,
0.217971,
-0.339764,
-0.165482
],
[
0.730559,
0.446197,
0.210105,
0.124198,
0.70353,
-0.254592
],
[
0.528319,
0.358492,
-0.02237,
-0.215453,
0.108242,
-0.296804
],
[
0.193697,
-0.319612,
-0.046657,
0.280241,
-0.275888,
-0.150059
]
],
"network.8.bias": [
-0.84612,
0.139614,
-0.30143,
-0.131505,
0.093005,
0.261466
],
"network.10.weight": [
[
0.195984,
0.135343,
-0.268892,
-0.744443,
-0.333335,
0.091205
]
],
"network.10.bias": [
0.259742
]
}
## Activation Signature
### 0
mean: [1.160571, 0.026829, -1.148685, 0.125375, 0.978900, 0.828864]
std: [1.411778, 1.966559, 1.544125, 1.896105, 0.971504, 1.325184]
### 2
mean: [-0.857630, 0.859149, -1.246504, -2.319102, 0.348366, 0.724167]
std: [0.505936, 1.496730, 1.195092, 1.316156, 0.493574, 1.708101]
### 4
mean: [0.041936, 0.465572, -0.075875, -0.024277, 0.555201, 1.008249]
std: [0.818649, 1.294824, 0.700928, 0.331119, 1.701679, 2.035151]
### 6
mean: [1.659023, 0.847358, -0.261477, -0.359800, 0.889521, 0.036783]
std: [3.311719, 2.628312, 0.053290, 0.941297, 1.926463, 1.540045]
### 8
mean: [-1.856788, -0.756835, -0.920126, 2.094670, 1.337415, -0.071864]
std: [1.306021, 1.538500, 1.021312, 4.911207, 2.918426, 0.657299]
### 10
mean: [-1.884325]
std: [4.552058]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
first_last_match
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 5
Neurons per Layer: 6
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
0.369047,
0.170073,
-0.174411,
0.463175,
-0.384633
],
[
-0.628364,
-0.202014,
-0.137648,
0.264092,
0.720742
],
[
-0.107461,
0.188083,
-0.283976,
-0.005005,
-0.680114
],
[
0.346377,
-0.558166,
0.479523,
-0.425996,
0.172198
],
[
-0.325529,
0.138686,
0.43884,
-0.06256,
0.15171
],
[
-0.412894,
0.119755,
-0.062143,
0.044015,
0.646851
]
],
"network.0.bias": [
0.236854,
0.00919,
0.085048,
0.410905,
0.161645,
0.362109
],
"network.2.weight": [
[
-0.321695,
-0.152414,
0.00011,
0.165426,
-0.116933,
0.126858
],
[
-0.311467,
0.537787,
0.512688,
0.260676,
0.378425,
0.450917
],
[
0.07497,
-0.478518,
-0.209122,
-0.618495,
0.013149,
-0.378891
],
[
-0.500792,
-0.498812,
0.221931,
-0.607862,
-0.204715,
-0.610003
],
[
0.089296,
0.033225,
0.145991,
-0.188005,
0.299418,
-0.436896
],
[
-0.275718,
0.624781,
0.500946,
0.152893,
0.201084,
0.673226
]
],
"network.2.bias": [
-0.481068,
0.00761,
-0.338189,
-0.289575,
0.444604,
-0.183008
],
"network.4.weight": [
[
-0.616713,
-0.428454,
0.124196,
-0.153734,
0.125543,
-0.116802
],
[
-0.522921,
0.278006,
0.22951,
-0.12712,
-0.500301,
0.504584
],
[
0.091142,
-0.056923,
0.192982,
0.010719,
-0.500435,
0.415481
],
[
-0.018993,
0.009283,
-0.079501,
0.175686,
-0.335184,
-0.25351
],
[
-0.070248,
0.444066,
0.543769,
0.576079,
-0.252548,
0.625372
],
[
0.105502,
0.493369,
0.358002,
0.007794,
-0.500166,
0.775405
]
],
"network.4.bias": [
0.415745,
-0.092954,
-0.1818,
0.278581,
-0.197616,
0.129586
],
"network.6.weight": [
[
-0.210715,
0.552917,
0.321775,
-0.387064,
0.740219,
0.608701
],
[
-0.505303,
0.682066,
0.27821,
0.072669,
0.26318,
0.574615
],
[
-0.186448,
-0.090774,
-0.097118,
0.707387,
0.490974,
-0.276757
],
[
0.202606,
0.065748,
-0.461015,
-0.320672,
-0.359792,
-0.092411
],
[
-0.203616,
0.063461,
0.075144,
-0.368453,
0.458454,
0.509247
],
[
0.365258,
-0.072333,
-0.624752,
-0.341152,
-0.176087,
-0.397932
]
],
"network.6.bias": [
0.2272,
-0.229008,
-0.236449,
-0.030127,
0.046421,
0.615103
],
"network.8.weight": [
[
-0.297861,
0.048394,
0.610614,
0.728506,
-0.294604,
-0.543728
],
[
-0.383711,
-0.270011,
0.161776,
0.414104,
0.18294,
-0.263535
],
[
-0.118429,
-0.009306,
0.266864,
0.217971,
-0.339764,
-0.165482
],
[
0.730559,
0.446197,
0.210105,
0.124198,
0.70353,
-0.254592
],
[
0.528319,
0.358492,
-0.02237,
-0.215453,
0.108242,
-0.296804
],
[
0.193697,
-0.319612,
-0.046657,
0.280241,
-0.275888,
-0.150059
]
],
"network.8.bias": [
-0.84612,
0.139614,
-0.30143,
-0.131505,
0.093005,
0.261466
],
"network.10.weight": [
[
0.195984,
0.135343,
-0.268892,
-0.744443,
-0.333335,
0.091205
]
],
"network.10.bias": [
0.259742
]
}
## Activation Signature
### 0
mean: [1.160571, 0.026829, -1.148685, 0.125375, 0.978900, 0.828864]
std: [1.411778, 1.966559, 1.544125, 1.896105, 0.971504, 1.325184]
### 2
mean: [-0.857630, 0.859149, -1.246504, -2.319102, 0.348366, 0.724167]
std: [0.505936, 1.496730, 1.195092, 1.316156, 0.493574, 1.708101]
### 4
mean: [0.041936, 0.465572, -0.075875, -0.024277, 0.555201, 1.008249]
std: [0.818649, 1.294824, 0.700928, 0.331119, 1.701679, 2.035151]
### 6
mean: [1.659023, 0.847358, -0.261477, -0.359800, 0.889521, 0.036783]
std: [3.311719, 2.628312, 0.053290, 0.941297, 1.926463, 1.540045]
### 8
mean: [-1.856788, -0.756835, -0.920126, 2.094670, 1.337415, -0.071864]
std: [1.306021, 1.538500, 1.021312, 4.911207, 2.918426, 0.657299]
### 10
mean: [-1.884325]
std: [4.552058]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
first_last_match
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 1.1605710983276367, "std": 1.4117778539657593}, "1": {"mean": 0.026828672736883163, "std": 1.966558575630188}, "2": {"mean": -1.1486846208572388, "std": 1.5441248416900635}, "3": {"mean": 0.12537455558776855, "std": 1.8961046934127808}, "4": {"mean": 0.9789000749588013, "std": 0.971504271030426}, "5": {"mean": 0.8288642168045044, "std": 1.3251837491989136}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.8576298952102661, "std": 0.505935549736023}, "1": {"mean": 0.8591493368148804, "std": 1.4967303276062012}, "2": {"mean": -1.2465038299560547, "std": 1.1950918436050415}, "3": {"mean": -2.319101572036743, "std": 1.3161559104919434}, "4": {"mean": 0.3483659327030182, "std": 0.49357420206069946}, "5": {"mean": 0.7241670489311218, "std": 1.7081010341644287}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.04193614423274994, "std": 0.8186489343643188}, "1": {"mean": 0.46557173132896423, "std": 1.2948241233825684}, "2": {"mean": -0.07587526738643646, "std": 0.7009284496307373}, "3": {"mean": -0.02427714318037033, "std": 0.331118643283844}, "4": {"mean": 0.5552005171775818, "std": 1.7016792297363281}, "5": {"mean": 1.0082486867904663, "std": 2.0351505279541016}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 1.659023404121399, "std": 3.3117194175720215}, "1": {"mean": 0.8473579287528992, "std": 2.6283118724823}, "2": {"mean": -0.261476993560791, "std": 0.05328989028930664}, "3": {"mean": -0.3598000705242157, "std": 0.9412973523139954}, "4": {"mean": 0.8895210027694702, "std": 1.9264633655548096}, "5": {"mean": 0.03678318113088608, "std": 1.5400453805923462}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.8567878007888794, "std": 1.3060210943222046}, "1": {"mean": -0.7568351030349731, "std": 1.5385003089904785}, "2": {"mean": -0.9201260209083557, "std": 1.0213117599487305}, "3": {"mean": 2.094669818878174, "std": 4.91120719909668}, "4": {"mean": 1.3374149799346924, "std": 2.9184255599975586}, "5": {"mean": -0.07186447829008102, "std": 0.6572988629341125}}, "layer_info": {"num_neurons": 6, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -1.8843251466751099, "std": 4.55205774307251}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 5, "neurons_per_layer": 6, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[0.369047, 0.170073, -0.174411, 0.463175, -0.384633], [-0.628364, -0.202014, -0.137648, 0.264092, 0.720742], [-0.107461, 0.188083, -0.283976, -0.005005, -0.680114], [0.346377, -0.558166, 0.479523, -0.425996, 0.172198], [-0.325529, 0.138686, 0.43884, -0.06256, 0.15171], [-0.412894, 0.119755, -0.062143, 0.044015, 0.646851]], "network.0.bias": [0.236854, 0.00919, 0.085048, 0.410905, 0.161645, 0.362109], "network.2.weight": [[-0.321695, -0.152414, 0.00011, 0.165426, -0.116933, 0.126858], [-0.311467, 0.537787, 0.512688, 0.260676, 0.378425, 0.450917], [0.07497, -0.478518, -0.209122, -0.618495, 0.013149, -0.378891], [-0.500792, -0.498812, 0.221931, -0.607862, -0.204715, -0.610003], [0.089296, 0.033225, 0.145991, -0.188005, 0.299418, -0.436896], [-0.275718, 0.624781, 0.500946, 0.152893, 0.201084, 0.673226]], "network.2.bias": [-0.481068, 0.00761, -0.338189, -0.289575, 0.444604, -0.183008], "network.4.weight": [[-0.616713, -0.428454, 0.124196, -0.153734, 0.125543, -0.116802], [-0.522921, 0.278006, 0.22951, -0.12712, -0.500301, 0.504584], [0.091142, -0.056923, 0.192982, 0.010719, -0.500435, 0.415481], [-0.018993, 0.009283, -0.079501, 0.175686, -0.335184, -0.25351], [-0.070248, 0.444066, 0.543769, 0.576079, -0.252548, 0.625372], [0.105502, 0.493369, 0.358002, 0.007794, -0.500166, 0.775405]], "network.4.bias": [0.415745, -0.092954, -0.1818, 0.278581, -0.197616, 0.129586], "network.6.weight": [[-0.210715, 0.552917, 0.321775, -0.387064, 0.740219, 0.608701], [-0.505303, 0.682066, 0.27821, 0.072669, 0.26318, 0.574615], [-0.186448, -0.090774, -0.097118, 0.707387, 0.490974, -0.276757], [0.202606, 0.065748, -0.461015, -0.320672, -0.359792, -0.092411], [-0.203616, 0.063461, 0.075144, -0.368453, 0.458454, 0.509247], [0.365258, -0.072333, -0.624752, -0.341152, -0.176087, -0.397932]], "network.6.bias": [0.2272, -0.229008, -0.236449, -0.030127, 0.046421, 0.615103], "network.8.weight": [[-0.297861, 0.048394, 0.610614, 0.728506, -0.294604, -0.543728], [-0.383711, -0.270011, 0.161776, 0.414104, 0.18294, -0.263535], [-0.118429, -0.009306, 0.266864, 0.217971, -0.339764, -0.165482], [0.730559, 0.446197, 0.210105, 0.124198, 0.70353, -0.254592], [0.528319, 0.358492, -0.02237, -0.215453, 0.108242, -0.296804], [0.193697, -0.319612, -0.046657, 0.280241, -0.275888, -0.150059]], "network.8.bias": [-0.84612, 0.139614, -0.30143, -0.131505, 0.093005, 0.261466], "network.10.weight": [[0.195984, 0.135343, -0.268892, -0.744443, -0.333335, 0.091205]], "network.10.bias": [0.259742]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7016567289829254, "train_acc": 0.44, "val_loss": 0.6465933322906494, "val_acc": 0.58}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6765626966953278, "train_acc": 0.555, "val_loss": 0.5933147668838501, "val_acc": 0.58}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.6261861026287079, "train_acc": 0.48, "val_loss": 0.497810959815979, "val_acc": 0.72}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.5258068740367889, "train_acc": 0.76, "val_loss": 0.4406237304210663, "val_acc": 0.78}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.5296743810176849, "train_acc": 0.715, "val_loss": 0.45905518531799316, "val_acc": 0.74}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.4946480095386505, "train_acc": 0.74, "val_loss": 0.48228561878204346, "val_acc": 0.72}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.5081653594970703, "train_acc": 0.74, "val_loss": 0.4651734530925751, "val_acc": 0.74}], "summary": {"total_epochs": 7, "degraded_epochs": 2, "improved_epochs": 5, "patterns": ["first_last_match"], "degraded_stage": {"initial_val_loss": 0.6465933322906494, "final_val_loss": 0.5933147668838501, "initial_val_acc": 0.58, "final_val_acc": 0.58, "best_val_acc": 0.58}, "improved_stage": {"initial_val_loss": 0.497810959815979, "final_val_loss": 0.4651734530925751, "initial_val_acc": 0.72, "final_val_acc": 0.74, "best_val_acc": 0.78, "best_epoch": 3}, "improvement": 0.20000000000000007, "first_improvement_epoch": 1}}
|
91
|
{"target_pattern": "mountain_pattern", "degraded_accuracy": 0.64, "improved_accuracy": 0.84, "improvement": 0.19999999999999996, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 9217, "learning_rate": 0.03413339280459434, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "mountain_pattern", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["mountain_pattern"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.147766,
-0.408518,
0.430708,
0.240523,
-0.026108
],
[
-0.17618,
0.348834,
0.089448,
0.052008,
-0.083194
],
[
-0.055695,
0.496102,
0.352443,
-0.213119,
0.220963
],
[
0.219576,
-0.06344,
-0.091946,
0.036346,
-0.112992
],
[
-0.227764,
-0.323152,
-0.16907,
-0.012625,
-0.0948
],
[
-0.112461,
-0.260045,
0.29566,
0.172087,
0.012556
],
[
0.14737,
0.055561,
-0.488982,
-0.147292,
-0.534412
],
[
0.065767,
-0.21531,
0.216812,
-0.567253,
-0.547954
]
],
"network.0.bias": [
0.285955,
-0.355869,
0.57814,
0.380587,
-0.297232,
-0.187628,
-0.579158,
-0.622876
],
"network.2.weight": [
[
-0.26663,
0.113506,
-0.136481,
0.048528,
-0.127674,
0.097647,
-0.079164,
-0.004227
],
[
-0.217548,
0.292038,
0.15957,
0.281812,
0.019285,
-0.232626,
0.008668,
-0.222703
],
[
-0.157493,
0.390344,
0.361167,
-0.089549,
0.275287,
-0.371854,
0.246084,
0.026631
],
[
0.455727,
-0.138506,
-0.077637,
-0.419754,
0.345635,
-0.145655,
-0.342737,
-0.141594
],
[
0.39569,
0.0296,
-0.027499,
-0.078814,
-0.178116,
0.44698,
-0.128694,
-0.270358
],
[
-0.212828,
0.050709,
0.16424,
0.272741,
-0.370267,
-0.358056,
0.30994,
0.107765
],
[
0.392458,
0.06132,
-0.066911,
-0.304994,
0.132039,
0.122711,
-0.047899,
-0.13201
],
[
-0.399933,
0.062248,
0.409615,
0.433837,
-0.244798,
-0.062649,
0.259381,
0.030563
]
],
"network.2.bias": [
0.533085,
0.427972,
0.260957,
-0.055488,
-0.111158,
0.301865,
0.129429,
0.492858
],
"network.4.weight": [
[
0.20838,
-0.172428,
-0.13866,
0.117641,
0.482251,
-0.120584,
0.274337,
0.124626
],
[
0.407263,
0.115612,
0.203285,
-0.12273,
0.100545,
-0.068845,
-0.343914,
0.423072
],
[
0.225436,
-0.340723,
-0.091198,
-0.093027,
-0.440076,
-0.166402,
-0.026798,
0.216566
],
[
-0.174171,
0.116644,
-0.214921,
-0.242981,
-0.026129,
0.090122,
0.054292,
0.52614
],
[
-0.330818,
-0.041005,
0.016066,
0.121781,
0.420393,
0.191326,
0.293054,
-0.323414
],
[
-0.211829,
-0.18066,
-0.070787,
0.208531,
-0.267152,
-0.013556,
-0.000668,
-0.011648
],
[
0.450772,
0.50971,
0.216192,
-0.216523,
-0.364062,
0.30529,
-0.067298,
0.443892
],
[
-0.220088,
-0.183147,
-0.353751,
-0.288388,
0.009332,
0.241686,
0.130107,
-0.058932
]
],
"network.4.bias": [
-0.056945,
0.140441,
-0.410874,
-0.039373,
-0.260125,
-0.363827,
0.399518,
-0.226475
],
"network.6.weight": [
[
-0.51946,
0.340244,
-0.192006,
-0.265788,
0.101166,
0.015231,
0.297718,
-0.164233
],
[
-0.319778,
0.151185,
-0.313265,
0.272443,
-0.40037,
-0.210375,
-0.054215,
0.249531
],
[
-0.053716,
0.479769,
0.097767,
0.338578,
-0.290313,
-0.413945,
0.454651,
-0.261994
],
[
0.180133,
0.261067,
-0.048435,
0.083908,
0.116077,
-0.010004,
0.411995,
0.056488
],
[
-0.466231,
0.118632,
0.260523,
0.32157,
-0.252041,
-0.196266,
0.470357,
0.247811
],
[
0.249825,
-0.434805,
-0.124064,
0.359546,
0.216655,
-0.156502,
-0.16453,
-0.022874
],
[
0.001366,
-0.049941,
-0.333728,
0.088554,
-0.158481,
0.237791,
-0.050529,
-0.25357
],
[
-0.183361,
0.483675,
-0.242503,
0.139785,
-0.280328,
0.247165,
0.577447,
-0.042964
]
],
"network.6.bias": [
0.236184,
-0.132641,
0.182386,
-0.182842,
0.173187,
0.338475,
0.087492,
0.084167
],
"network.8.weight": [
[
-0.31156,
-0.235717,
-0.333329,
-0.147435,
-0.346223,
0.313994,
0.074213,
-0.501354
]
],
"network.8.bias": [
0.175577
]
}
## Activation Signature
### 0
mean: [0.740947, 0.256307, 1.964553, 0.284757, -1.666687, 0.200676, -2.290915, -2.372732]
std: [1.039454, 0.720991, 1.270262, 0.431211, 1.067284, 0.717643, 1.498477, 1.774249]
### 2
mean: [0.134591, 0.673131, 0.816264, -0.074432, 0.337266, 0.414289, 0.305509, 1.102319]
std: [0.248033, 0.479074, 0.689971, 0.429484, 0.570888, 0.453156, 0.452993, 0.684176]
### 4
mean: [0.137418, 0.805808, -0.716261, 0.428635, -0.316411, -0.685873, 1.452531, -0.632820]
std: [0.450622, 0.529181, 0.213028, 0.340504, 0.495502, 0.102850, 0.974905, 0.267309]
### 6
mean: [0.718340, -0.088224, 1.349233, 0.729418, 0.965435, -0.016203, -0.004393, 1.317002]
std: [0.518995, 0.293007, 0.839198, 0.480956, 0.786915, 0.385631, 0.035617, 0.932812]
### 8
mean: [-1.616155]
std: [1.247188]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
mountain_pattern
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.147766,
-0.408518,
0.430708,
0.240523,
-0.026108
],
[
-0.17618,
0.348834,
0.089448,
0.052008,
-0.083194
],
[
-0.055695,
0.496102,
0.352443,
-0.213119,
0.220963
],
[
0.219576,
-0.06344,
-0.091946,
0.036346,
-0.112992
],
[
-0.227764,
-0.323152,
-0.16907,
-0.012625,
-0.0948
],
[
-0.112461,
-0.260045,
0.29566,
0.172087,
0.012556
],
[
0.14737,
0.055561,
-0.488982,
-0.147292,
-0.534412
],
[
0.065767,
-0.21531,
0.216812,
-0.567253,
-0.547954
]
],
"network.0.bias": [
0.285955,
-0.355869,
0.57814,
0.380587,
-0.297232,
-0.187628,
-0.579158,
-0.622876
],
"network.2.weight": [
[
-0.26663,
0.113506,
-0.136481,
0.048528,
-0.127674,
0.097647,
-0.079164,
-0.004227
],
[
-0.217548,
0.292038,
0.15957,
0.281812,
0.019285,
-0.232626,
0.008668,
-0.222703
],
[
-0.157493,
0.390344,
0.361167,
-0.089549,
0.275287,
-0.371854,
0.246084,
0.026631
],
[
0.455727,
-0.138506,
-0.077637,
-0.419754,
0.345635,
-0.145655,
-0.342737,
-0.141594
],
[
0.39569,
0.0296,
-0.027499,
-0.078814,
-0.178116,
0.44698,
-0.128694,
-0.270358
],
[
-0.212828,
0.050709,
0.16424,
0.272741,
-0.370267,
-0.358056,
0.30994,
0.107765
],
[
0.392458,
0.06132,
-0.066911,
-0.304994,
0.132039,
0.122711,
-0.047899,
-0.13201
],
[
-0.399933,
0.062248,
0.409615,
0.433837,
-0.244798,
-0.062649,
0.259381,
0.030563
]
],
"network.2.bias": [
0.533085,
0.427972,
0.260957,
-0.055488,
-0.111158,
0.301865,
0.129429,
0.492858
],
"network.4.weight": [
[
0.20838,
-0.172428,
-0.13866,
0.117641,
0.482251,
-0.120584,
0.274337,
0.124626
],
[
0.407263,
0.115612,
0.203285,
-0.12273,
0.100545,
-0.068845,
-0.343914,
0.423072
],
[
0.225436,
-0.340723,
-0.091198,
-0.093027,
-0.440076,
-0.166402,
-0.026798,
0.216566
],
[
-0.174171,
0.116644,
-0.214921,
-0.242981,
-0.026129,
0.090122,
0.054292,
0.52614
],
[
-0.330818,
-0.041005,
0.016066,
0.121781,
0.420393,
0.191326,
0.293054,
-0.323414
],
[
-0.211829,
-0.18066,
-0.070787,
0.208531,
-0.267152,
-0.013556,
-0.000668,
-0.011648
],
[
0.450772,
0.50971,
0.216192,
-0.216523,
-0.364062,
0.30529,
-0.067298,
0.443892
],
[
-0.220088,
-0.183147,
-0.353751,
-0.288388,
0.009332,
0.241686,
0.130107,
-0.058932
]
],
"network.4.bias": [
-0.056945,
0.140441,
-0.410874,
-0.039373,
-0.260125,
-0.363827,
0.399518,
-0.226475
],
"network.6.weight": [
[
-0.51946,
0.340244,
-0.192006,
-0.265788,
0.101166,
0.015231,
0.297718,
-0.164233
],
[
-0.319778,
0.151185,
-0.313265,
0.272443,
-0.40037,
-0.210375,
-0.054215,
0.249531
],
[
-0.053716,
0.479769,
0.097767,
0.338578,
-0.290313,
-0.413945,
0.454651,
-0.261994
],
[
0.180133,
0.261067,
-0.048435,
0.083908,
0.116077,
-0.010004,
0.411995,
0.056488
],
[
-0.466231,
0.118632,
0.260523,
0.32157,
-0.252041,
-0.196266,
0.470357,
0.247811
],
[
0.249825,
-0.434805,
-0.124064,
0.359546,
0.216655,
-0.156502,
-0.16453,
-0.022874
],
[
0.001366,
-0.049941,
-0.333728,
0.088554,
-0.158481,
0.237791,
-0.050529,
-0.25357
],
[
-0.183361,
0.483675,
-0.242503,
0.139785,
-0.280328,
0.247165,
0.577447,
-0.042964
]
],
"network.6.bias": [
0.236184,
-0.132641,
0.182386,
-0.182842,
0.173187,
0.338475,
0.087492,
0.084167
],
"network.8.weight": [
[
-0.31156,
-0.235717,
-0.333329,
-0.147435,
-0.346223,
0.313994,
0.074213,
-0.501354
]
],
"network.8.bias": [
0.175577
]
}
## Activation Signature
### 0
mean: [0.740947, 0.256307, 1.964553, 0.284757, -1.666687, 0.200676, -2.290915, -2.372732]
std: [1.039454, 0.720991, 1.270262, 0.431211, 1.067284, 0.717643, 1.498477, 1.774249]
### 2
mean: [0.134591, 0.673131, 0.816264, -0.074432, 0.337266, 0.414289, 0.305509, 1.102319]
std: [0.248033, 0.479074, 0.689971, 0.429484, 0.570888, 0.453156, 0.452993, 0.684176]
### 4
mean: [0.137418, 0.805808, -0.716261, 0.428635, -0.316411, -0.685873, 1.452531, -0.632820]
std: [0.450622, 0.529181, 0.213028, 0.340504, 0.495502, 0.102850, 0.974905, 0.267309]
### 6
mean: [0.718340, -0.088224, 1.349233, 0.729418, 0.965435, -0.016203, -0.004393, 1.317002]
std: [0.518995, 0.293007, 0.839198, 0.480956, 0.786915, 0.385631, 0.035617, 0.932812]
### 8
mean: [-1.616155]
std: [1.247188]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
mountain_pattern
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": 0.7409474849700928, "std": 1.0394535064697266}, "1": {"mean": 0.2563072741031647, "std": 0.7209912538528442}, "2": {"mean": 1.9645527601242065, "std": 1.2702616453170776}, "3": {"mean": 0.28475722670555115, "std": 0.43121057748794556}, "4": {"mean": -1.6666873693466187, "std": 1.0672837495803833}, "5": {"mean": 0.20067615807056427, "std": 0.7176433205604553}, "6": {"mean": -2.2909154891967773, "std": 1.4984767436981201}, "7": {"mean": -2.372732162475586, "std": 1.7742493152618408}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 0.1345909684896469, "std": 0.2480325549840927}, "1": {"mean": 0.6731314063072205, "std": 0.47907400131225586}, "2": {"mean": 0.8162643313407898, "std": 0.6899711489677429}, "3": {"mean": -0.07443246990442276, "std": 0.4294840693473816}, "4": {"mean": 0.3372657299041748, "std": 0.5708878040313721}, "5": {"mean": 0.41428884863853455, "std": 0.45315560698509216}, "6": {"mean": 0.30550915002822876, "std": 0.452993243932724}, "7": {"mean": 1.1023191213607788, "std": 0.6841758489608765}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.13741780817508698, "std": 0.4506216049194336}, "1": {"mean": 0.8058084845542908, "std": 0.5291812419891357}, "2": {"mean": -0.7162607908248901, "std": 0.2130281776189804}, "3": {"mean": 0.4286353588104248, "std": 0.3405035734176636}, "4": {"mean": -0.31641098856925964, "std": 0.4955017566680908}, "5": {"mean": -0.6858726143836975, "std": 0.1028502881526947}, "6": {"mean": 1.452531099319458, "std": 0.9749045968055725}, "7": {"mean": -0.6328195929527283, "std": 0.26730936765670776}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.7183403968811035, "std": 0.5189948081970215}, "1": {"mean": -0.08822400867938995, "std": 0.2930069863796234}, "2": {"mean": 1.3492330312728882, "std": 0.8391982913017273}, "3": {"mean": 0.7294183373451233, "std": 0.48095613718032837}, "4": {"mean": 0.9654354453086853, "std": 0.7869153022766113}, "5": {"mean": -0.016202768310904503, "std": 0.3856305778026581}, "6": {"mean": -0.004393343813717365, "std": 0.0356169268488884}, "7": {"mean": 1.3170017004013062, "std": 0.9328118562698364}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.6161550283432007, "std": 1.2471884489059448}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.147766, -0.408518, 0.430708, 0.240523, -0.026108], [-0.17618, 0.348834, 0.089448, 0.052008, -0.083194], [-0.055695, 0.496102, 0.352443, -0.213119, 0.220963], [0.219576, -0.06344, -0.091946, 0.036346, -0.112992], [-0.227764, -0.323152, -0.16907, -0.012625, -0.0948], [-0.112461, -0.260045, 0.29566, 0.172087, 0.012556], [0.14737, 0.055561, -0.488982, -0.147292, -0.534412], [0.065767, -0.21531, 0.216812, -0.567253, -0.547954]], "network.0.bias": [0.285955, -0.355869, 0.57814, 0.380587, -0.297232, -0.187628, -0.579158, -0.622876], "network.2.weight": [[-0.26663, 0.113506, -0.136481, 0.048528, -0.127674, 0.097647, -0.079164, -0.004227], [-0.217548, 0.292038, 0.15957, 0.281812, 0.019285, -0.232626, 0.008668, -0.222703], [-0.157493, 0.390344, 0.361167, -0.089549, 0.275287, -0.371854, 0.246084, 0.026631], [0.455727, -0.138506, -0.077637, -0.419754, 0.345635, -0.145655, -0.342737, -0.141594], [0.39569, 0.0296, -0.027499, -0.078814, -0.178116, 0.44698, -0.128694, -0.270358], [-0.212828, 0.050709, 0.16424, 0.272741, -0.370267, -0.358056, 0.30994, 0.107765], [0.392458, 0.06132, -0.066911, -0.304994, 0.132039, 0.122711, -0.047899, -0.13201], [-0.399933, 0.062248, 0.409615, 0.433837, -0.244798, -0.062649, 0.259381, 0.030563]], "network.2.bias": [0.533085, 0.427972, 0.260957, -0.055488, -0.111158, 0.301865, 0.129429, 0.492858], "network.4.weight": [[0.20838, -0.172428, -0.13866, 0.117641, 0.482251, -0.120584, 0.274337, 0.124626], [0.407263, 0.115612, 0.203285, -0.12273, 0.100545, -0.068845, -0.343914, 0.423072], [0.225436, -0.340723, -0.091198, -0.093027, -0.440076, -0.166402, -0.026798, 0.216566], [-0.174171, 0.116644, -0.214921, -0.242981, -0.026129, 0.090122, 0.054292, 0.52614], [-0.330818, -0.041005, 0.016066, 0.121781, 0.420393, 0.191326, 0.293054, -0.323414], [-0.211829, -0.18066, -0.070787, 0.208531, -0.267152, -0.013556, -0.000668, -0.011648], [0.450772, 0.50971, 0.216192, -0.216523, -0.364062, 0.30529, -0.067298, 0.443892], [-0.220088, -0.183147, -0.353751, -0.288388, 0.009332, 0.241686, 0.130107, -0.058932]], "network.4.bias": [-0.056945, 0.140441, -0.410874, -0.039373, -0.260125, -0.363827, 0.399518, -0.226475], "network.6.weight": [[-0.51946, 0.340244, -0.192006, -0.265788, 0.101166, 0.015231, 0.297718, -0.164233], [-0.319778, 0.151185, -0.313265, 0.272443, -0.40037, -0.210375, -0.054215, 0.249531], [-0.053716, 0.479769, 0.097767, 0.338578, -0.290313, -0.413945, 0.454651, -0.261994], [0.180133, 0.261067, -0.048435, 0.083908, 0.116077, -0.010004, 0.411995, 0.056488], [-0.466231, 0.118632, 0.260523, 0.32157, -0.252041, -0.196266, 0.470357, 0.247811], [0.249825, -0.434805, -0.124064, 0.359546, 0.216655, -0.156502, -0.16453, -0.022874], [0.001366, -0.049941, -0.333728, 0.088554, -0.158481, 0.237791, -0.050529, -0.25357], [-0.183361, 0.483675, -0.242503, 0.139785, -0.280328, 0.247165, 0.577447, -0.042964]], "network.6.bias": [0.236184, -0.132641, 0.182386, -0.182842, 0.173187, 0.338475, 0.087492, 0.084167], "network.8.weight": [[-0.31156, -0.235717, -0.333329, -0.147435, -0.346223, 0.313994, 0.074213, -0.501354]], "network.8.bias": [0.175577]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6909227073192596, "train_acc": 0.54, "val_loss": 0.6645048260688782, "val_acc": 0.6}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6699088513851166, "train_acc": 0.54, "val_loss": 0.629912793636322, "val_acc": 0.64}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.6121538281440735, "train_acc": 0.62, "val_loss": 0.5510894060134888, "val_acc": 0.68}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.5015057325363159, "train_acc": 0.815, "val_loss": 0.4459746181964874, "val_acc": 0.84}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.3971627950668335, "train_acc": 0.86, "val_loss": 0.37387970089912415, "val_acc": 0.82}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.3151974081993103, "train_acc": 0.88, "val_loss": 0.43805503845214844, "val_acc": 0.8}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.3344157487154007, "train_acc": 0.88, "val_loss": 0.4803382158279419, "val_acc": 0.82}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.3315945714712143, "train_acc": 0.875, "val_loss": 0.4730793237686157, "val_acc": 0.8}], "summary": {"total_epochs": 8, "degraded_epochs": 2, "improved_epochs": 6, "patterns": ["mountain_pattern"], "degraded_stage": {"initial_val_loss": 0.6645048260688782, "final_val_loss": 0.629912793636322, "initial_val_acc": 0.6, "final_val_acc": 0.64, "best_val_acc": 0.64}, "improved_stage": {"initial_val_loss": 0.5510894060134888, "final_val_loss": 0.4730793237686157, "initial_val_acc": 0.68, "final_val_acc": 0.8, "best_val_acc": 0.84, "best_epoch": 3}, "improvement": 0.19999999999999996, "first_improvement_epoch": 1}}
|
92
|
{"target_pattern": "sorted_descending", "degraded_accuracy": 0.62, "improved_accuracy": 0.94, "improvement": 0.31999999999999995, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 1428, "learning_rate": 0.04296735297046316, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_descending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_descending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.189294,
-0.446234,
0.129863,
-0.221898,
0.639235
],
[
-0.328196,
-0.272392,
-0.256819,
0.469097,
0.396866
],
[
-0.793456,
0.365878,
0.252106,
0.24807,
-0.164742
],
[
-0.566874,
-0.199461,
-0.027138,
0.004696,
-0.041705
],
[
-0.347603,
0.285349,
0.160032,
-0.387877,
0.464475
]
],
"network.0.bias": [
0.141572,
-0.112561,
0.447393,
-0.583741,
0.494013
],
"network.2.weight": [
[
-0.512114,
-0.650587,
-0.444313,
0.287473,
-0.190084
],
[
0.811213,
0.430007,
0.827471,
-0.996673,
0.540846
],
[
0.938448,
0.915623,
0.560398,
-0.963317,
0.38616
],
[
0.345613,
0.537044,
0.73473,
-0.089465,
0.572707
],
[
-0.278365,
-0.55723,
-0.39453,
-0.347167,
0.106247
]
],
"network.2.bias": [
0.93765,
0.312951,
-0.2319,
-0.177489,
0.523666
],
"network.4.weight": [
[
0.444441,
0.080082,
-0.411165,
-0.442576,
0.306441
],
[
-0.559489,
0.777774,
0.543595,
0.716301,
-0.203174
],
[
-0.146661,
1.054516,
0.323895,
0.644362,
-0.222075
],
[
0.94977,
-0.107589,
-0.87409,
-0.580674,
0.333783
],
[
-0.468875,
0.238521,
0.522877,
0.453344,
-0.96019
]
],
"network.4.bias": [
-0.067854,
0.434996,
0.102401,
0.336095,
0.07468
],
"network.6.weight": [
[
-0.009868,
0.090458,
0.388063,
-0.403083,
0.844776
],
[
-0.421175,
-0.287297,
-0.412053,
-0.38323,
0.184674
],
[
-0.379908,
0.631096,
0.847868,
-0.879883,
0.107672
],
[
-0.078115,
0.546534,
0.007846,
-0.59976,
0.655141
],
[
-0.615884,
0.833658,
0.632266,
-0.765685,
0.656857
]
],
"network.6.bias": [
-0.053264,
-0.412,
0.133862,
-0.034923,
0.28583
],
"network.8.weight": [
[
0.669644,
0.971753,
-0.289524,
0.696743,
-0.60195
],
[
-0.493011,
-0.400753,
-0.665464,
-0.997686,
-0.295998
],
[
0.288053,
0.524344,
0.744967,
0.986902,
0.58422
],
[
0.718724,
0.221941,
0.260986,
0.313871,
0.769717
],
[
-0.743966,
-0.342793,
-0.604722,
-0.775511,
-0.177799
]
],
"network.8.bias": [
-0.405954,
0.801419,
-0.161702,
-0.378933,
0.703845
],
"network.10.weight": [
[
-0.60403,
0.86891,
-0.35459,
-0.005979,
0.220638
],
[
-0.199014,
0.514747,
-0.213702,
-0.774197,
0.390063
],
[
-0.41844,
-0.381112,
0.396512,
0.595269,
-0.678438
],
[
-0.333205,
0.646385,
-0.359564,
-0.308731,
-0.081327
],
[
-0.523673,
0.658788,
-0.627551,
-0.519252,
0.638276
]
],
"network.10.bias": [
0.350959,
0.438635,
0.225525,
0.361786,
0.75448
],
"network.12.weight": [
[
0.443819,
0.462996,
-0.414173,
0.631273,
0.798923
]
],
"network.12.bias": [
0.015052
]
}
## Activation Signature
### 0
mean: [-0.323321, -0.058389, 0.982104, -1.750699, 0.656769]
std: [1.608197, 1.611121, 1.719918, 1.352068, 1.138539]
### 2
mean: [-0.304778, 2.375539, 1.671516, 1.565931, -0.229348]
std: [1.161960, 1.646469, 1.796094, 1.396848, 0.844959]
### 4
mean: [-1.137501, 4.127585, 4.033920, -2.057399, 2.021542]
std: [1.365784, 3.417463, 3.291591, 2.820100, 2.234002]
### 6
mean: [3.560242, -2.974609, 6.104825, 3.469265, 7.416931]
std: [3.495203, 1.800260, 5.536493, 3.479999, 6.643283]
### 8
mean: [-1.866888, -10.962607, 13.482962, 10.801326, -9.905848]
std: [0.829636, 10.468536, 12.034622, 9.864090, 9.490090]
### 10
mean: [-4.245805, -10.678758, 11.890451, -7.717168, -13.090313]
std: [4.605264, 10.346911, 10.781985, 7.481828, 12.918340]
### 12
mean: [-4.168627]
std: [5.421268]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_descending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.189294,
-0.446234,
0.129863,
-0.221898,
0.639235
],
[
-0.328196,
-0.272392,
-0.256819,
0.469097,
0.396866
],
[
-0.793456,
0.365878,
0.252106,
0.24807,
-0.164742
],
[
-0.566874,
-0.199461,
-0.027138,
0.004696,
-0.041705
],
[
-0.347603,
0.285349,
0.160032,
-0.387877,
0.464475
]
],
"network.0.bias": [
0.141572,
-0.112561,
0.447393,
-0.583741,
0.494013
],
"network.2.weight": [
[
-0.512114,
-0.650587,
-0.444313,
0.287473,
-0.190084
],
[
0.811213,
0.430007,
0.827471,
-0.996673,
0.540846
],
[
0.938448,
0.915623,
0.560398,
-0.963317,
0.38616
],
[
0.345613,
0.537044,
0.73473,
-0.089465,
0.572707
],
[
-0.278365,
-0.55723,
-0.39453,
-0.347167,
0.106247
]
],
"network.2.bias": [
0.93765,
0.312951,
-0.2319,
-0.177489,
0.523666
],
"network.4.weight": [
[
0.444441,
0.080082,
-0.411165,
-0.442576,
0.306441
],
[
-0.559489,
0.777774,
0.543595,
0.716301,
-0.203174
],
[
-0.146661,
1.054516,
0.323895,
0.644362,
-0.222075
],
[
0.94977,
-0.107589,
-0.87409,
-0.580674,
0.333783
],
[
-0.468875,
0.238521,
0.522877,
0.453344,
-0.96019
]
],
"network.4.bias": [
-0.067854,
0.434996,
0.102401,
0.336095,
0.07468
],
"network.6.weight": [
[
-0.009868,
0.090458,
0.388063,
-0.403083,
0.844776
],
[
-0.421175,
-0.287297,
-0.412053,
-0.38323,
0.184674
],
[
-0.379908,
0.631096,
0.847868,
-0.879883,
0.107672
],
[
-0.078115,
0.546534,
0.007846,
-0.59976,
0.655141
],
[
-0.615884,
0.833658,
0.632266,
-0.765685,
0.656857
]
],
"network.6.bias": [
-0.053264,
-0.412,
0.133862,
-0.034923,
0.28583
],
"network.8.weight": [
[
0.669644,
0.971753,
-0.289524,
0.696743,
-0.60195
],
[
-0.493011,
-0.400753,
-0.665464,
-0.997686,
-0.295998
],
[
0.288053,
0.524344,
0.744967,
0.986902,
0.58422
],
[
0.718724,
0.221941,
0.260986,
0.313871,
0.769717
],
[
-0.743966,
-0.342793,
-0.604722,
-0.775511,
-0.177799
]
],
"network.8.bias": [
-0.405954,
0.801419,
-0.161702,
-0.378933,
0.703845
],
"network.10.weight": [
[
-0.60403,
0.86891,
-0.35459,
-0.005979,
0.220638
],
[
-0.199014,
0.514747,
-0.213702,
-0.774197,
0.390063
],
[
-0.41844,
-0.381112,
0.396512,
0.595269,
-0.678438
],
[
-0.333205,
0.646385,
-0.359564,
-0.308731,
-0.081327
],
[
-0.523673,
0.658788,
-0.627551,
-0.519252,
0.638276
]
],
"network.10.bias": [
0.350959,
0.438635,
0.225525,
0.361786,
0.75448
],
"network.12.weight": [
[
0.443819,
0.462996,
-0.414173,
0.631273,
0.798923
]
],
"network.12.bias": [
0.015052
]
}
## Activation Signature
### 0
mean: [-0.323321, -0.058389, 0.982104, -1.750699, 0.656769]
std: [1.608197, 1.611121, 1.719918, 1.352068, 1.138539]
### 2
mean: [-0.304778, 2.375539, 1.671516, 1.565931, -0.229348]
std: [1.161960, 1.646469, 1.796094, 1.396848, 0.844959]
### 4
mean: [-1.137501, 4.127585, 4.033920, -2.057399, 2.021542]
std: [1.365784, 3.417463, 3.291591, 2.820100, 2.234002]
### 6
mean: [3.560242, -2.974609, 6.104825, 3.469265, 7.416931]
std: [3.495203, 1.800260, 5.536493, 3.479999, 6.643283]
### 8
mean: [-1.866888, -10.962607, 13.482962, 10.801326, -9.905848]
std: [0.829636, 10.468536, 12.034622, 9.864090, 9.490090]
### 10
mean: [-4.245805, -10.678758, 11.890451, -7.717168, -13.090313]
std: [4.605264, 10.346911, 10.781985, 7.481828, 12.918340]
### 12
mean: [-4.168627]
std: [5.421268]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_descending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -0.32332053780555725, "std": 1.6081973314285278}, "1": {"mean": -0.05838869884610176, "std": 1.611121416091919}, "2": {"mean": 0.9821040630340576, "std": 1.7199177742004395}, "3": {"mean": -1.7506994009017944, "std": 1.3520680665969849}, "4": {"mean": 0.6567692756652832, "std": 1.13853919506073}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.30477771162986755, "std": 1.1619600057601929}, "1": {"mean": 2.3755385875701904, "std": 1.646469235420227}, "2": {"mean": 1.6715155839920044, "std": 1.7960938215255737}, "3": {"mean": 1.5659312009811401, "std": 1.396848440170288}, "4": {"mean": -0.22934812307357788, "std": 0.844959020614624}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -1.1375008821487427, "std": 1.3657842874526978}, "1": {"mean": 4.127585411071777, "std": 3.4174630641937256}, "2": {"mean": 4.0339202880859375, "std": 3.291591167449951}, "3": {"mean": -2.057398796081543, "std": 2.8200995922088623}, "4": {"mean": 2.0215420722961426, "std": 2.2340023517608643}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 3.560241937637329, "std": 3.4952025413513184}, "1": {"mean": -2.974609136581421, "std": 1.8002598285675049}, "2": {"mean": 6.104824542999268, "std": 5.536493301391602}, "3": {"mean": 3.469264507293701, "std": 3.479998826980591}, "4": {"mean": 7.416930675506592, "std": 6.643282890319824}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.8668880462646484, "std": 0.8296359777450562}, "1": {"mean": -10.962607383728027, "std": 10.468536376953125}, "2": {"mean": 13.482961654663086, "std": 12.034622192382812}, "3": {"mean": 10.801325798034668, "std": 9.864089965820312}, "4": {"mean": -9.905847549438477, "std": 9.490090370178223}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -4.245804786682129, "std": 4.605263710021973}, "1": {"mean": -10.678757667541504, "std": 10.346911430358887}, "2": {"mean": 11.890451431274414, "std": 10.78198528289795}, "3": {"mean": -7.717167854309082, "std": 7.481828212738037}, "4": {"mean": -13.090312957763672, "std": 12.918339729309082}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -4.16862678527832, "std": 5.421267509460449}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.189294, -0.446234, 0.129863, -0.221898, 0.639235], [-0.328196, -0.272392, -0.256819, 0.469097, 0.396866], [-0.793456, 0.365878, 0.252106, 0.24807, -0.164742], [-0.566874, -0.199461, -0.027138, 0.004696, -0.041705], [-0.347603, 0.285349, 0.160032, -0.387877, 0.464475]], "network.0.bias": [0.141572, -0.112561, 0.447393, -0.583741, 0.494013], "network.2.weight": [[-0.512114, -0.650587, -0.444313, 0.287473, -0.190084], [0.811213, 0.430007, 0.827471, -0.996673, 0.540846], [0.938448, 0.915623, 0.560398, -0.963317, 0.38616], [0.345613, 0.537044, 0.73473, -0.089465, 0.572707], [-0.278365, -0.55723, -0.39453, -0.347167, 0.106247]], "network.2.bias": [0.93765, 0.312951, -0.2319, -0.177489, 0.523666], "network.4.weight": [[0.444441, 0.080082, -0.411165, -0.442576, 0.306441], [-0.559489, 0.777774, 0.543595, 0.716301, -0.203174], [-0.146661, 1.054516, 0.323895, 0.644362, -0.222075], [0.94977, -0.107589, -0.87409, -0.580674, 0.333783], [-0.468875, 0.238521, 0.522877, 0.453344, -0.96019]], "network.4.bias": [-0.067854, 0.434996, 0.102401, 0.336095, 0.07468], "network.6.weight": [[-0.009868, 0.090458, 0.388063, -0.403083, 0.844776], [-0.421175, -0.287297, -0.412053, -0.38323, 0.184674], [-0.379908, 0.631096, 0.847868, -0.879883, 0.107672], [-0.078115, 0.546534, 0.007846, -0.59976, 0.655141], [-0.615884, 0.833658, 0.632266, -0.765685, 0.656857]], "network.6.bias": [-0.053264, -0.412, 0.133862, -0.034923, 0.28583], "network.8.weight": [[0.669644, 0.971753, -0.289524, 0.696743, -0.60195], [-0.493011, -0.400753, -0.665464, -0.997686, -0.295998], [0.288053, 0.524344, 0.744967, 0.986902, 0.58422], [0.718724, 0.221941, 0.260986, 0.313871, 0.769717], [-0.743966, -0.342793, -0.604722, -0.775511, -0.177799]], "network.8.bias": [-0.405954, 0.801419, -0.161702, -0.378933, 0.703845], "network.10.weight": [[-0.60403, 0.86891, -0.35459, -0.005979, 0.220638], [-0.199014, 0.514747, -0.213702, -0.774197, 0.390063], [-0.41844, -0.381112, 0.396512, 0.595269, -0.678438], [-0.333205, 0.646385, -0.359564, -0.308731, -0.081327], [-0.523673, 0.658788, -0.627551, -0.519252, 0.638276]], "network.10.bias": [0.350959, 0.438635, 0.225525, 0.361786, 0.75448], "network.12.weight": [[0.443819, 0.462996, -0.414173, 0.631273, 0.798923]], "network.12.bias": [0.015052]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6888768672943115, "train_acc": 0.55, "val_loss": 0.6813093423843384, "val_acc": 0.62}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6898618042469025, "train_acc": 0.55, "val_loss": 0.679975152015686, "val_acc": 0.62}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6850316524505615, "train_acc": 0.55, "val_loss": 0.6642927527427673, "val_acc": 0.62}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.678237795829773, "train_acc": 0.55, "val_loss": 0.6347686648368835, "val_acc": 0.62}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6566454768180847, "train_acc": 0.47, "val_loss": 0.5583257675170898, "val_acc": 0.92}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.5424829423427582, "train_acc": 0.925, "val_loss": 0.4085559844970703, "val_acc": 0.9}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.3705996870994568, "train_acc": 0.925, "val_loss": 0.23264531791210175, "val_acc": 0.9}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.20392496138811111, "train_acc": 0.93, "val_loss": 0.35453999042510986, "val_acc": 0.82}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.2078409343957901, "train_acc": 0.91, "val_loss": 0.24585632979869843, "val_acc": 0.94}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.25147728621959686, "train_acc": 0.94, "val_loss": 0.2531092166900635, "val_acc": 0.94}], "summary": {"total_epochs": 10, "degraded_epochs": 4, "improved_epochs": 6, "patterns": ["sorted_descending"], "degraded_stage": {"initial_val_loss": 0.6813093423843384, "final_val_loss": 0.6347686648368835, "initial_val_acc": 0.62, "final_val_acc": 0.62, "best_val_acc": 0.62}, "improved_stage": {"initial_val_loss": 0.5583257675170898, "final_val_loss": 0.2531092166900635, "initial_val_acc": 0.92, "final_val_acc": 0.94, "best_val_acc": 0.94, "best_epoch": 8}, "improvement": 0.31999999999999995, "first_improvement_epoch": 3}}
|
93
|
{"target_pattern": "palindrome", "degraded_accuracy": 0.76, "improved_accuracy": 0.96, "improvement": 0.19999999999999996, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 3126, "learning_rate": 0.08469175679902498, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "palindrome", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["palindrome"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.221422,
-0.67326,
-1.051407,
-0.356931,
-0.542766
],
[
-0.579778,
0.391329,
-0.003073,
-0.296823,
-0.888908
],
[
-0.453531,
0.109615,
0.404605,
-0.262648,
-0.843453
],
[
-1.201813,
0.300609,
0.360209,
0.149191,
-0.603419
],
[
-0.095195,
-1.061817,
0.698652,
-0.741373,
-0.709674
],
[
-1.664916,
0.425329,
0.138538,
0.053644,
-1.396626
],
[
0.028574,
-0.251265,
-0.013109,
-0.037124,
-0.371232
]
],
"network.0.bias": [
-1.044552,
-0.573145,
0.461836,
-0.089988,
0.728709,
-0.031594,
-0.447473
],
"network.2.weight": [
[
0.194219,
-0.010725,
0.009383,
0.040091,
-0.722704,
-0.846214,
1.055093
],
[
-0.198266,
0.417742,
0.677983,
0.086525,
1.024962,
0.65644,
-0.950511
],
[
-0.13479,
-0.099948,
0.542764,
-0.06071,
1.1243,
-0.123002,
-1.49083
],
[
0.031135,
0.610308,
0.285683,
-0.56336,
-0.478714,
-0.245472,
0.793974
],
[
-0.212774,
-0.116457,
0.660917,
0.169369,
1.179182,
0.344666,
-0.791985
],
[
-0.3835,
0.198593,
0.47813,
0.21813,
1.023052,
0.402692,
-0.784325
],
[
0.734399,
-0.475261,
0.182287,
-0.951079,
-0.337298,
-0.05914,
1.188735
]
],
"network.2.bias": [
0.150203,
0.017674,
-0.075625,
-0.434879,
0.345886,
0.112231,
-0.371997
],
"network.4.weight": [
[
-0.121316,
-0.08514,
-0.216318,
-0.209785,
-0.259314,
-0.305866,
-0.269684
],
[
0.22984,
-0.702824,
-0.629322,
0.782301,
-0.6735,
-0.244032,
1.377117
],
[
-0.13999,
-0.882711,
-0.107782,
0.021497,
-1.333618,
-0.721045,
0.337813
],
[
-0.54617,
0.309314,
0.326128,
-0.455772,
0.258599,
0.364627,
-1.009438
],
[
-0.813158,
-0.900657,
-0.935935,
0.732759,
-0.719281,
-0.565638,
1.242941
],
[
0.285945,
-1.405045,
-0.363381,
-0.074535,
-1.061358,
-1.249785,
-0.059737
],
[
-0.272846,
0.58309,
0.456998,
-0.636052,
0.348076,
0.108838,
-1.135583
]
],
"network.4.bias": [
0.519001,
-0.570377,
-0.364399,
0.000886,
-0.55401,
-0.48457,
-0.161401
],
"network.6.weight": [
[
0.656429,
-0.649767,
-1.075547,
-0.254704,
-0.845249,
-0.351179,
-0.27226
],
[
-0.316572,
0.648012,
-0.037822,
-0.059941,
0.236134,
0.686881,
0.710163
],
[
0.861074,
-0.70442,
-0.68249,
-0.411499,
-0.552574,
-0.261632,
-0.58918
],
[
-0.079963,
0.457154,
0.710123,
-1.047284,
0.645448,
0.267935,
-0.558358
],
[
-0.269345,
0.053551,
0.372948,
0.368612,
0.082484,
-0.225469,
0.324728
],
[
0.041461,
0.627862,
0.323765,
0.491917,
0.360927,
1.38679,
0.232641
],
[
0.494348,
-0.420137,
-0.348972,
-0.111032,
-0.739599,
-0.745217,
-0.286738
]
],
"network.6.bias": [
0.56179,
0.283011,
0.764556,
-0.736984,
0.396928,
0.171586,
0.164923
],
"network.8.weight": [
[
0.847329,
-0.165841,
0.677764,
-0.78068,
-0.282508,
-0.601578,
0.706608
],
[
1.032283,
-0.371188,
1.172119,
-0.959113,
-0.126604,
-0.593829,
0.461667
],
[
-0.60714,
0.159538,
-0.497883,
0.11611,
-0.278534,
0.454409,
-0.026236
],
[
-0.229881,
-0.407365,
-0.073205,
0.053938,
-0.501399,
-0.312649,
-0.758964
],
[
0.760466,
-0.221054,
0.288729,
0.158251,
-0.015944,
-0.342373,
0.359867
],
[
-0.326029,
0.718565,
-0.45904,
0.093084,
0.524952,
0.488431,
-0.690325
],
[
0.767913,
0.039353,
0.437896,
0.493568,
-0.448924,
-0.171572,
0.628516
]
],
"network.8.bias": [
0.818028,
0.848126,
0.072548,
0.052276,
-0.07606,
0.140771,
-0.281602
],
"network.10.weight": [
[
0.021931,
0.266779,
0.36099,
-0.358916,
-0.230742,
-0.587446,
-0.283575
],
[
0.862065,
1.066176,
-0.969611,
0.138067,
0.645466,
-0.17587,
-0.130908
],
[
-0.46675,
-0.949026,
0.044193,
-0.635463,
1.067391,
0.062264,
0.682605
],
[
-0.023297,
0.188642,
0.761664,
0.396583,
0.148712,
0.014555,
0.075637
],
[
0.831148,
0.986911,
-0.244731,
0.210655,
0.246962,
0.077012,
-0.116005
],
[
0.566163,
-0.063433,
0.316829,
-0.714945,
-0.725768,
-1.380225,
-0.062319
],
[
-0.442854,
-0.882596,
0.215222,
-0.867325,
0.147959,
0.763075,
0.056322
]
],
"network.10.bias": [
-1.061658,
-0.144889,
-0.042807,
-0.57096,
-0.217571,
-1.225204,
0.193468
],
"network.12.weight": [
[
0.288868,
-0.210652,
-0.004938,
-0.169185,
-0.162567,
0.572792,
0.500927
]
],
"network.12.bias": [
-0.661149
]
}
## Activation Signature
### 0
mean: [-6.178030, -2.320817, -0.661134, -0.709650, -2.330313, -2.646524, -1.434620]
std: [3.379354, 2.249664, 1.958353, 2.647944, 3.256851, 4.401028, 0.844577]
### 2
mean: [-0.442024, 0.900415, 0.504237, -0.995441, 1.205141, 0.909920, -1.107598]
std: [0.912355, 1.575818, 1.363429, 0.746880, 1.638767, 1.432099, 0.963789]
### 4
mean: [-0.133313, -2.677081, -3.259522, 1.194727, -3.248499, -4.034444, 1.245930]
std: [1.295538, 3.308539, 4.769358, 1.817802, 4.533704, 6.242818, 2.177866]
### 6
mean: [0.324847, 0.861830, -0.040277, -2.722862, 1.107782, 0.758907, -0.018708]
std: [1.289453, 1.550941, 2.251639, 3.070503, 1.436023, 1.533943, 1.001843]
### 8
mean: [1.025061, 1.404363, -0.377163, -1.335726, 0.123756, 1.005863, -0.051318]
std: [2.369889, 2.724412, 1.021309, 1.664457, 1.370190, 3.023469, 1.474295]
### 10
mean: [-1.471021, 3.099739, -1.554615, -0.040255, 3.089183, -2.651369, -0.880620]
std: [1.600401, 4.086599, 1.436001, 0.420968, 2.926208, 3.849633, 3.825473]
### 12
mean: [-1.409259]
std: [2.112575]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
palindrome
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 7
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.221422,
-0.67326,
-1.051407,
-0.356931,
-0.542766
],
[
-0.579778,
0.391329,
-0.003073,
-0.296823,
-0.888908
],
[
-0.453531,
0.109615,
0.404605,
-0.262648,
-0.843453
],
[
-1.201813,
0.300609,
0.360209,
0.149191,
-0.603419
],
[
-0.095195,
-1.061817,
0.698652,
-0.741373,
-0.709674
],
[
-1.664916,
0.425329,
0.138538,
0.053644,
-1.396626
],
[
0.028574,
-0.251265,
-0.013109,
-0.037124,
-0.371232
]
],
"network.0.bias": [
-1.044552,
-0.573145,
0.461836,
-0.089988,
0.728709,
-0.031594,
-0.447473
],
"network.2.weight": [
[
0.194219,
-0.010725,
0.009383,
0.040091,
-0.722704,
-0.846214,
1.055093
],
[
-0.198266,
0.417742,
0.677983,
0.086525,
1.024962,
0.65644,
-0.950511
],
[
-0.13479,
-0.099948,
0.542764,
-0.06071,
1.1243,
-0.123002,
-1.49083
],
[
0.031135,
0.610308,
0.285683,
-0.56336,
-0.478714,
-0.245472,
0.793974
],
[
-0.212774,
-0.116457,
0.660917,
0.169369,
1.179182,
0.344666,
-0.791985
],
[
-0.3835,
0.198593,
0.47813,
0.21813,
1.023052,
0.402692,
-0.784325
],
[
0.734399,
-0.475261,
0.182287,
-0.951079,
-0.337298,
-0.05914,
1.188735
]
],
"network.2.bias": [
0.150203,
0.017674,
-0.075625,
-0.434879,
0.345886,
0.112231,
-0.371997
],
"network.4.weight": [
[
-0.121316,
-0.08514,
-0.216318,
-0.209785,
-0.259314,
-0.305866,
-0.269684
],
[
0.22984,
-0.702824,
-0.629322,
0.782301,
-0.6735,
-0.244032,
1.377117
],
[
-0.13999,
-0.882711,
-0.107782,
0.021497,
-1.333618,
-0.721045,
0.337813
],
[
-0.54617,
0.309314,
0.326128,
-0.455772,
0.258599,
0.364627,
-1.009438
],
[
-0.813158,
-0.900657,
-0.935935,
0.732759,
-0.719281,
-0.565638,
1.242941
],
[
0.285945,
-1.405045,
-0.363381,
-0.074535,
-1.061358,
-1.249785,
-0.059737
],
[
-0.272846,
0.58309,
0.456998,
-0.636052,
0.348076,
0.108838,
-1.135583
]
],
"network.4.bias": [
0.519001,
-0.570377,
-0.364399,
0.000886,
-0.55401,
-0.48457,
-0.161401
],
"network.6.weight": [
[
0.656429,
-0.649767,
-1.075547,
-0.254704,
-0.845249,
-0.351179,
-0.27226
],
[
-0.316572,
0.648012,
-0.037822,
-0.059941,
0.236134,
0.686881,
0.710163
],
[
0.861074,
-0.70442,
-0.68249,
-0.411499,
-0.552574,
-0.261632,
-0.58918
],
[
-0.079963,
0.457154,
0.710123,
-1.047284,
0.645448,
0.267935,
-0.558358
],
[
-0.269345,
0.053551,
0.372948,
0.368612,
0.082484,
-0.225469,
0.324728
],
[
0.041461,
0.627862,
0.323765,
0.491917,
0.360927,
1.38679,
0.232641
],
[
0.494348,
-0.420137,
-0.348972,
-0.111032,
-0.739599,
-0.745217,
-0.286738
]
],
"network.6.bias": [
0.56179,
0.283011,
0.764556,
-0.736984,
0.396928,
0.171586,
0.164923
],
"network.8.weight": [
[
0.847329,
-0.165841,
0.677764,
-0.78068,
-0.282508,
-0.601578,
0.706608
],
[
1.032283,
-0.371188,
1.172119,
-0.959113,
-0.126604,
-0.593829,
0.461667
],
[
-0.60714,
0.159538,
-0.497883,
0.11611,
-0.278534,
0.454409,
-0.026236
],
[
-0.229881,
-0.407365,
-0.073205,
0.053938,
-0.501399,
-0.312649,
-0.758964
],
[
0.760466,
-0.221054,
0.288729,
0.158251,
-0.015944,
-0.342373,
0.359867
],
[
-0.326029,
0.718565,
-0.45904,
0.093084,
0.524952,
0.488431,
-0.690325
],
[
0.767913,
0.039353,
0.437896,
0.493568,
-0.448924,
-0.171572,
0.628516
]
],
"network.8.bias": [
0.818028,
0.848126,
0.072548,
0.052276,
-0.07606,
0.140771,
-0.281602
],
"network.10.weight": [
[
0.021931,
0.266779,
0.36099,
-0.358916,
-0.230742,
-0.587446,
-0.283575
],
[
0.862065,
1.066176,
-0.969611,
0.138067,
0.645466,
-0.17587,
-0.130908
],
[
-0.46675,
-0.949026,
0.044193,
-0.635463,
1.067391,
0.062264,
0.682605
],
[
-0.023297,
0.188642,
0.761664,
0.396583,
0.148712,
0.014555,
0.075637
],
[
0.831148,
0.986911,
-0.244731,
0.210655,
0.246962,
0.077012,
-0.116005
],
[
0.566163,
-0.063433,
0.316829,
-0.714945,
-0.725768,
-1.380225,
-0.062319
],
[
-0.442854,
-0.882596,
0.215222,
-0.867325,
0.147959,
0.763075,
0.056322
]
],
"network.10.bias": [
-1.061658,
-0.144889,
-0.042807,
-0.57096,
-0.217571,
-1.225204,
0.193468
],
"network.12.weight": [
[
0.288868,
-0.210652,
-0.004938,
-0.169185,
-0.162567,
0.572792,
0.500927
]
],
"network.12.bias": [
-0.661149
]
}
## Activation Signature
### 0
mean: [-6.178030, -2.320817, -0.661134, -0.709650, -2.330313, -2.646524, -1.434620]
std: [3.379354, 2.249664, 1.958353, 2.647944, 3.256851, 4.401028, 0.844577]
### 2
mean: [-0.442024, 0.900415, 0.504237, -0.995441, 1.205141, 0.909920, -1.107598]
std: [0.912355, 1.575818, 1.363429, 0.746880, 1.638767, 1.432099, 0.963789]
### 4
mean: [-0.133313, -2.677081, -3.259522, 1.194727, -3.248499, -4.034444, 1.245930]
std: [1.295538, 3.308539, 4.769358, 1.817802, 4.533704, 6.242818, 2.177866]
### 6
mean: [0.324847, 0.861830, -0.040277, -2.722862, 1.107782, 0.758907, -0.018708]
std: [1.289453, 1.550941, 2.251639, 3.070503, 1.436023, 1.533943, 1.001843]
### 8
mean: [1.025061, 1.404363, -0.377163, -1.335726, 0.123756, 1.005863, -0.051318]
std: [2.369889, 2.724412, 1.021309, 1.664457, 1.370190, 3.023469, 1.474295]
### 10
mean: [-1.471021, 3.099739, -1.554615, -0.040255, 3.089183, -2.651369, -0.880620]
std: [1.600401, 4.086599, 1.436001, 0.420968, 2.926208, 3.849633, 3.825473]
### 12
mean: [-1.409259]
std: [2.112575]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
palindrome
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -6.178030014038086, "std": 3.3793537616729736}, "1": {"mean": -2.320816993713379, "std": 2.249664306640625}, "2": {"mean": -0.6611344218254089, "std": 1.95835280418396}, "3": {"mean": -0.7096496820449829, "std": 2.647944211959839}, "4": {"mean": -2.3303134441375732, "std": 3.2568511962890625}, "5": {"mean": -2.64652419090271, "std": 4.401027679443359}, "6": {"mean": -1.4346203804016113, "std": 0.844576895236969}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.44202443957328796, "std": 0.9123554229736328}, "1": {"mean": 0.9004151225090027, "std": 1.5758183002471924}, "2": {"mean": 0.5042365789413452, "std": 1.3634285926818848}, "3": {"mean": -0.9954410195350647, "std": 0.746880054473877}, "4": {"mean": 1.2051405906677246, "std": 1.6387666463851929}, "5": {"mean": 0.9099203944206238, "std": 1.4320985078811646}, "6": {"mean": -1.1075984239578247, "std": 0.9637894630432129}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -0.13331305980682373, "std": 1.2955375909805298}, "1": {"mean": -2.6770808696746826, "std": 3.3085386753082275}, "2": {"mean": -3.2595224380493164, "std": 4.769358158111572}, "3": {"mean": 1.1947271823883057, "std": 1.817801594734192}, "4": {"mean": -3.2484991550445557, "std": 4.5337042808532715}, "5": {"mean": -4.0344438552856445, "std": 6.242818355560303}, "6": {"mean": 1.2459304332733154, "std": 2.177866220474243}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 0.3248465955257416, "std": 1.2894532680511475}, "1": {"mean": 0.8618304133415222, "std": 1.5509408712387085}, "2": {"mean": -0.040277279913425446, "std": 2.251638889312744}, "3": {"mean": -2.7228617668151855, "std": 3.0705032348632812}, "4": {"mean": 1.1077823638916016, "std": 1.4360227584838867}, "5": {"mean": 0.7589067816734314, "std": 1.5339428186416626}, "6": {"mean": -0.018708273768424988, "std": 1.0018434524536133}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 1.0250608921051025, "std": 2.3698890209198}, "1": {"mean": 1.4043633937835693, "std": 2.724412202835083}, "2": {"mean": -0.3771628439426422, "std": 1.021309494972229}, "3": {"mean": -1.3357256650924683, "std": 1.664457082748413}, "4": {"mean": 0.1237555742263794, "std": 1.3701900243759155}, "5": {"mean": 1.0058634281158447, "std": 3.0234694480895996}, "6": {"mean": -0.05131768062710762, "std": 1.4742952585220337}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -1.4710206985473633, "std": 1.6004012823104858}, "1": {"mean": 3.099738836288452, "std": 4.086599349975586}, "2": {"mean": -1.5546151399612427, "std": 1.4360007047653198}, "3": {"mean": -0.04025476053357124, "std": 0.42096757888793945}, "4": {"mean": 3.0891826152801514, "std": 2.926208257675171}, "5": {"mean": -2.6513686180114746, "std": 3.849632501602173}, "6": {"mean": -0.8806196451187134, "std": 3.8254730701446533}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -1.4092594385147095, "std": 2.112574815750122}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.221422, -0.67326, -1.051407, -0.356931, -0.542766], [-0.579778, 0.391329, -0.003073, -0.296823, -0.888908], [-0.453531, 0.109615, 0.404605, -0.262648, -0.843453], [-1.201813, 0.300609, 0.360209, 0.149191, -0.603419], [-0.095195, -1.061817, 0.698652, -0.741373, -0.709674], [-1.664916, 0.425329, 0.138538, 0.053644, -1.396626], [0.028574, -0.251265, -0.013109, -0.037124, -0.371232]], "network.0.bias": [-1.044552, -0.573145, 0.461836, -0.089988, 0.728709, -0.031594, -0.447473], "network.2.weight": [[0.194219, -0.010725, 0.009383, 0.040091, -0.722704, -0.846214, 1.055093], [-0.198266, 0.417742, 0.677983, 0.086525, 1.024962, 0.65644, -0.950511], [-0.13479, -0.099948, 0.542764, -0.06071, 1.1243, -0.123002, -1.49083], [0.031135, 0.610308, 0.285683, -0.56336, -0.478714, -0.245472, 0.793974], [-0.212774, -0.116457, 0.660917, 0.169369, 1.179182, 0.344666, -0.791985], [-0.3835, 0.198593, 0.47813, 0.21813, 1.023052, 0.402692, -0.784325], [0.734399, -0.475261, 0.182287, -0.951079, -0.337298, -0.05914, 1.188735]], "network.2.bias": [0.150203, 0.017674, -0.075625, -0.434879, 0.345886, 0.112231, -0.371997], "network.4.weight": [[-0.121316, -0.08514, -0.216318, -0.209785, -0.259314, -0.305866, -0.269684], [0.22984, -0.702824, -0.629322, 0.782301, -0.6735, -0.244032, 1.377117], [-0.13999, -0.882711, -0.107782, 0.021497, -1.333618, -0.721045, 0.337813], [-0.54617, 0.309314, 0.326128, -0.455772, 0.258599, 0.364627, -1.009438], [-0.813158, -0.900657, -0.935935, 0.732759, -0.719281, -0.565638, 1.242941], [0.285945, -1.405045, -0.363381, -0.074535, -1.061358, -1.249785, -0.059737], [-0.272846, 0.58309, 0.456998, -0.636052, 0.348076, 0.108838, -1.135583]], "network.4.bias": [0.519001, -0.570377, -0.364399, 0.000886, -0.55401, -0.48457, -0.161401], "network.6.weight": [[0.656429, -0.649767, -1.075547, -0.254704, -0.845249, -0.351179, -0.27226], [-0.316572, 0.648012, -0.037822, -0.059941, 0.236134, 0.686881, 0.710163], [0.861074, -0.70442, -0.68249, -0.411499, -0.552574, -0.261632, -0.58918], [-0.079963, 0.457154, 0.710123, -1.047284, 0.645448, 0.267935, -0.558358], [-0.269345, 0.053551, 0.372948, 0.368612, 0.082484, -0.225469, 0.324728], [0.041461, 0.627862, 0.323765, 0.491917, 0.360927, 1.38679, 0.232641], [0.494348, -0.420137, -0.348972, -0.111032, -0.739599, -0.745217, -0.286738]], "network.6.bias": [0.56179, 0.283011, 0.764556, -0.736984, 0.396928, 0.171586, 0.164923], "network.8.weight": [[0.847329, -0.165841, 0.677764, -0.78068, -0.282508, -0.601578, 0.706608], [1.032283, -0.371188, 1.172119, -0.959113, -0.126604, -0.593829, 0.461667], [-0.60714, 0.159538, -0.497883, 0.11611, -0.278534, 0.454409, -0.026236], [-0.229881, -0.407365, -0.073205, 0.053938, -0.501399, -0.312649, -0.758964], [0.760466, -0.221054, 0.288729, 0.158251, -0.015944, -0.342373, 0.359867], [-0.326029, 0.718565, -0.45904, 0.093084, 0.524952, 0.488431, -0.690325], [0.767913, 0.039353, 0.437896, 0.493568, -0.448924, -0.171572, 0.628516]], "network.8.bias": [0.818028, 0.848126, 0.072548, 0.052276, -0.07606, 0.140771, -0.281602], "network.10.weight": [[0.021931, 0.266779, 0.36099, -0.358916, -0.230742, -0.587446, -0.283575], [0.862065, 1.066176, -0.969611, 0.138067, 0.645466, -0.17587, -0.130908], [-0.46675, -0.949026, 0.044193, -0.635463, 1.067391, 0.062264, 0.682605], [-0.023297, 0.188642, 0.761664, 0.396583, 0.148712, 0.014555, 0.075637], [0.831148, 0.986911, -0.244731, 0.210655, 0.246962, 0.077012, -0.116005], [0.566163, -0.063433, 0.316829, -0.714945, -0.725768, -1.380225, -0.062319], [-0.442854, -0.882596, 0.215222, -0.867325, 0.147959, 0.763075, 0.056322]], "network.10.bias": [-1.061658, -0.144889, -0.042807, -0.57096, -0.217571, -1.225204, 0.193468], "network.12.weight": [[0.288868, -0.210652, -0.004938, -0.169185, -0.162567, 0.572792, 0.500927]], "network.12.bias": [-0.661149]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.697492390871048, "train_acc": 0.585, "val_loss": 0.7135833501815796, "val_acc": 0.48}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6806480884552002, "train_acc": 0.585, "val_loss": 0.6081570386886597, "val_acc": 0.76}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.6392010152339935, "train_acc": 0.67, "val_loss": 0.5557979345321655, "val_acc": 0.8}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.6491268277168274, "train_acc": 0.72, "val_loss": 0.7357564568519592, "val_acc": 0.48}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.6945227384567261, "train_acc": 0.51, "val_loss": 0.6851651072502136, "val_acc": 0.5}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.6163231432437897, "train_acc": 0.615, "val_loss": 0.4786622226238251, "val_acc": 0.86}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.5355322659015656, "train_acc": 0.79, "val_loss": 0.4097675681114197, "val_acc": 0.9}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.540764182806015, "train_acc": 0.8, "val_loss": 0.44005709886550903, "val_acc": 0.9}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.49229849874973297, "train_acc": 0.815, "val_loss": 0.4717344641685486, "val_acc": 0.82}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.4630115181207657, "train_acc": 0.825, "val_loss": 0.34323006868362427, "val_acc": 0.88}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.4064553380012512, "train_acc": 0.85, "val_loss": 0.2161167860031128, "val_acc": 0.96}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.4456498622894287, "train_acc": 0.855, "val_loss": 0.2102137804031372, "val_acc": 0.96}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["palindrome"], "degraded_stage": {"initial_val_loss": 0.7135833501815796, "final_val_loss": 0.6081570386886597, "initial_val_acc": 0.48, "final_val_acc": 0.76, "best_val_acc": 0.76}, "improved_stage": {"initial_val_loss": 0.5557979345321655, "final_val_loss": 0.2102137804031372, "initial_val_acc": 0.8, "final_val_acc": 0.96, "best_val_acc": 0.96, "best_epoch": 10}, "improvement": 0.19999999999999996, "first_improvement_epoch": 1}}
|
94
|
{"target_pattern": "sorted_ascending", "degraded_accuracy": 0.68, "improved_accuracy": 0.94, "improvement": 0.2599999999999999, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 4908, "learning_rate": 0.07820043132053234, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_ascending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_ascending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.151775,
-0.059857,
-0.068186,
-0.095639,
-0.394961
],
[
-1.205371,
-0.786052,
-0.608968,
0.296726,
0.524108
],
[
0.860498,
0.530286,
0.095924,
-0.29192,
-0.168628
],
[
-0.781661,
-1.346989,
-0.422029,
0.255054,
0.355188
],
[
-0.810135,
-0.949166,
-0.068033,
-0.296937,
0.645574
]
],
"network.0.bias": [
-0.2225,
0.01936,
0.181163,
0.307308,
0.146852
],
"network.2.weight": [
[
-0.019745,
0.447754,
-0.205076,
0.071683,
-0.456709
],
[
0.366043,
0.210735,
-0.203847,
0.203887,
-0.542468
],
[
0.004636,
0.876047,
-0.629004,
0.334053,
0.258218
],
[
0.272861,
0.38164,
-0.654997,
0.277237,
0.249452
],
[
0.20717,
0.456218,
-0.667458,
0.948762,
0.375345
]
],
"network.2.bias": [
0.218775,
-0.235393,
0.082811,
0.246127,
0.453891
],
"network.4.weight": [
[
-1.021499,
-0.369447,
0.509622,
0.262597,
0.796182
],
[
-0.358978,
0.299295,
-0.087209,
0.09016,
-0.60906
],
[
-0.216,
0.133885,
-0.331028,
-0.308479,
-0.304639
],
[
-0.157251,
0.298345,
-0.358622,
-0.031719,
-0.391069
],
[
-0.220715,
0.170381,
0.163363,
-0.205153,
-0.394304
]
],
"network.4.bias": [
-0.227825,
-0.011755,
-0.257563,
-0.516093,
-0.236089
],
"network.6.weight": [
[
-0.042127,
-0.465679,
-0.470258,
-0.196322,
0.128895
],
[
1.007784,
0.374313,
-0.147033,
-0.015715,
-0.122353
],
[
0.432743,
-0.318056,
-0.160762,
-0.445011,
-0.308509
],
[
0.340809,
-0.016885,
-0.068663,
0.39103,
0.141808
],
[
-0.374824,
-0.129411,
-0.330502,
0.082211,
0.052028
]
],
"network.6.bias": [
-0.048549,
0.186436,
-0.328514,
0.390323,
-0.28405
],
"network.8.weight": [
[
0.050321,
-0.388125,
0.037718,
-0.124911,
-0.107027
],
[
0.347513,
0.08728,
0.335453,
0.426182,
-0.17585
],
[
0.260371,
0.571911,
0.489411,
0.184274,
-0.158506
],
[
0.087971,
-0.242801,
0.124935,
-0.613978,
-0.217582
],
[
0.218746,
-0.241364,
-0.021965,
0.177095,
0.129602
]
],
"network.8.bias": [
-0.405845,
-0.29877,
-0.226859,
0.046574,
-0.439126
],
"network.10.weight": [
[
0.050986,
-0.358228,
0.098848,
0.318,
-0.196852
],
[
-0.156583,
0.117455,
0.474129,
-0.036744,
0.029508
],
[
0.240726,
-0.290944,
-0.530335,
0.227197,
0.268918
],
[
-0.176356,
0.30123,
0.279669,
0.005378,
-0.411836
],
[
0.183437,
0.35377,
0.004211,
-0.190498,
0.32837
]
],
"network.10.bias": [
-0.251292,
-0.087476,
1.231301,
0.02906,
-0.231036
],
"network.12.weight": [
[
0.15346,
0.616833,
-1.518517,
0.482397,
0.226028
]
],
"network.12.bias": [
-0.417011
]
}
## Activation Signature
### 0
mean: [-1.355093, -2.902372, 1.584697, -3.016497, -2.573577]
std: [0.961840, 3.758484, 2.340024, 3.624263, 3.062252]
### 2
mean: [-0.093647, -0.586532, -0.713893, -0.713305, -0.374280]
std: [0.524939, 0.473957, 1.817434, 1.643425, 1.944847]
### 4
mean: [0.227060, -0.359191, -0.607535, -0.839451, -0.458011]
std: [1.212784, 0.741819, 0.835068, 0.777017, 0.432385]
### 6
mean: [-0.064382, 0.565200, -0.165873, 0.518412, -0.424923]
std: [0.048880, 1.169315, 0.502105, 0.395435, 0.434903]
### 8
mean: [-0.685968, 0.007077, 0.243823, -0.395700, -0.486066]
std: [0.487372, 0.412469, 0.948224, 0.474219, 0.221449]
### 10
mean: [-0.259632, 0.056422, 1.054359, 0.137175, -0.194422]
std: [0.046973, 0.489887, 0.609546, 0.378522, 0.140847]
### 12
mean: [-1.966468]
std: [0.925215]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_ascending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.151775,
-0.059857,
-0.068186,
-0.095639,
-0.394961
],
[
-1.205371,
-0.786052,
-0.608968,
0.296726,
0.524108
],
[
0.860498,
0.530286,
0.095924,
-0.29192,
-0.168628
],
[
-0.781661,
-1.346989,
-0.422029,
0.255054,
0.355188
],
[
-0.810135,
-0.949166,
-0.068033,
-0.296937,
0.645574
]
],
"network.0.bias": [
-0.2225,
0.01936,
0.181163,
0.307308,
0.146852
],
"network.2.weight": [
[
-0.019745,
0.447754,
-0.205076,
0.071683,
-0.456709
],
[
0.366043,
0.210735,
-0.203847,
0.203887,
-0.542468
],
[
0.004636,
0.876047,
-0.629004,
0.334053,
0.258218
],
[
0.272861,
0.38164,
-0.654997,
0.277237,
0.249452
],
[
0.20717,
0.456218,
-0.667458,
0.948762,
0.375345
]
],
"network.2.bias": [
0.218775,
-0.235393,
0.082811,
0.246127,
0.453891
],
"network.4.weight": [
[
-1.021499,
-0.369447,
0.509622,
0.262597,
0.796182
],
[
-0.358978,
0.299295,
-0.087209,
0.09016,
-0.60906
],
[
-0.216,
0.133885,
-0.331028,
-0.308479,
-0.304639
],
[
-0.157251,
0.298345,
-0.358622,
-0.031719,
-0.391069
],
[
-0.220715,
0.170381,
0.163363,
-0.205153,
-0.394304
]
],
"network.4.bias": [
-0.227825,
-0.011755,
-0.257563,
-0.516093,
-0.236089
],
"network.6.weight": [
[
-0.042127,
-0.465679,
-0.470258,
-0.196322,
0.128895
],
[
1.007784,
0.374313,
-0.147033,
-0.015715,
-0.122353
],
[
0.432743,
-0.318056,
-0.160762,
-0.445011,
-0.308509
],
[
0.340809,
-0.016885,
-0.068663,
0.39103,
0.141808
],
[
-0.374824,
-0.129411,
-0.330502,
0.082211,
0.052028
]
],
"network.6.bias": [
-0.048549,
0.186436,
-0.328514,
0.390323,
-0.28405
],
"network.8.weight": [
[
0.050321,
-0.388125,
0.037718,
-0.124911,
-0.107027
],
[
0.347513,
0.08728,
0.335453,
0.426182,
-0.17585
],
[
0.260371,
0.571911,
0.489411,
0.184274,
-0.158506
],
[
0.087971,
-0.242801,
0.124935,
-0.613978,
-0.217582
],
[
0.218746,
-0.241364,
-0.021965,
0.177095,
0.129602
]
],
"network.8.bias": [
-0.405845,
-0.29877,
-0.226859,
0.046574,
-0.439126
],
"network.10.weight": [
[
0.050986,
-0.358228,
0.098848,
0.318,
-0.196852
],
[
-0.156583,
0.117455,
0.474129,
-0.036744,
0.029508
],
[
0.240726,
-0.290944,
-0.530335,
0.227197,
0.268918
],
[
-0.176356,
0.30123,
0.279669,
0.005378,
-0.411836
],
[
0.183437,
0.35377,
0.004211,
-0.190498,
0.32837
]
],
"network.10.bias": [
-0.251292,
-0.087476,
1.231301,
0.02906,
-0.231036
],
"network.12.weight": [
[
0.15346,
0.616833,
-1.518517,
0.482397,
0.226028
]
],
"network.12.bias": [
-0.417011
]
}
## Activation Signature
### 0
mean: [-1.355093, -2.902372, 1.584697, -3.016497, -2.573577]
std: [0.961840, 3.758484, 2.340024, 3.624263, 3.062252]
### 2
mean: [-0.093647, -0.586532, -0.713893, -0.713305, -0.374280]
std: [0.524939, 0.473957, 1.817434, 1.643425, 1.944847]
### 4
mean: [0.227060, -0.359191, -0.607535, -0.839451, -0.458011]
std: [1.212784, 0.741819, 0.835068, 0.777017, 0.432385]
### 6
mean: [-0.064382, 0.565200, -0.165873, 0.518412, -0.424923]
std: [0.048880, 1.169315, 0.502105, 0.395435, 0.434903]
### 8
mean: [-0.685968, 0.007077, 0.243823, -0.395700, -0.486066]
std: [0.487372, 0.412469, 0.948224, 0.474219, 0.221449]
### 10
mean: [-0.259632, 0.056422, 1.054359, 0.137175, -0.194422]
std: [0.046973, 0.489887, 0.609546, 0.378522, 0.140847]
### 12
mean: [-1.966468]
std: [0.925215]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_ascending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.3550925254821777, "std": 0.9618397355079651}, "1": {"mean": -2.902371644973755, "std": 3.758484363555908}, "2": {"mean": 1.5846972465515137, "std": 2.340023994445801}, "3": {"mean": -3.0164973735809326, "std": 3.624262571334839}, "4": {"mean": -2.5735769271850586, "std": 3.062251567840576}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.09364670515060425, "std": 0.5249388217926025}, "1": {"mean": -0.586531937122345, "std": 0.4739571511745453}, "2": {"mean": -0.7138932943344116, "std": 1.8174339532852173}, "3": {"mean": -0.7133053541183472, "std": 1.6434248685836792}, "4": {"mean": -0.3742801249027252, "std": 1.9448472261428833}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 0.22706018388271332, "std": 1.2127840518951416}, "1": {"mean": -0.35919100046157837, "std": 0.7418187260627747}, "2": {"mean": -0.6075353622436523, "std": 0.8350675702095032}, "3": {"mean": -0.8394514918327332, "std": 0.77701735496521}, "4": {"mean": -0.4580114483833313, "std": 0.4323849678039551}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -0.06438247114419937, "std": 0.04887952283024788}, "1": {"mean": 0.5651998519897461, "std": 1.1693153381347656}, "2": {"mean": -0.16587284207344055, "std": 0.502104640007019}, "3": {"mean": 0.5184118747711182, "std": 0.39543482661247253}, "4": {"mean": -0.4249228239059448, "std": 0.4349026679992676}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.6859682202339172, "std": 0.4873718023300171}, "1": {"mean": 0.0070773945190012455, "std": 0.412468820810318}, "2": {"mean": 0.2438233494758606, "std": 0.9482235908508301}, "3": {"mean": -0.3956996202468872, "std": 0.47421902418136597}, "4": {"mean": -0.4860661029815674, "std": 0.22144901752471924}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -0.25963249802589417, "std": 0.04697268456220627}, "1": {"mean": 0.05642189458012581, "std": 0.4898874759674072}, "2": {"mean": 1.0543594360351562, "std": 0.6095460653305054}, "3": {"mean": 0.1371753215789795, "std": 0.37852200865745544}, "4": {"mean": -0.1944224238395691, "std": 0.14084745943546295}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -1.966468334197998, "std": 0.9252153635025024}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.151775, -0.059857, -0.068186, -0.095639, -0.394961], [-1.205371, -0.786052, -0.608968, 0.296726, 0.524108], [0.860498, 0.530286, 0.095924, -0.29192, -0.168628], [-0.781661, -1.346989, -0.422029, 0.255054, 0.355188], [-0.810135, -0.949166, -0.068033, -0.296937, 0.645574]], "network.0.bias": [-0.2225, 0.01936, 0.181163, 0.307308, 0.146852], "network.2.weight": [[-0.019745, 0.447754, -0.205076, 0.071683, -0.456709], [0.366043, 0.210735, -0.203847, 0.203887, -0.542468], [0.004636, 0.876047, -0.629004, 0.334053, 0.258218], [0.272861, 0.38164, -0.654997, 0.277237, 0.249452], [0.20717, 0.456218, -0.667458, 0.948762, 0.375345]], "network.2.bias": [0.218775, -0.235393, 0.082811, 0.246127, 0.453891], "network.4.weight": [[-1.021499, -0.369447, 0.509622, 0.262597, 0.796182], [-0.358978, 0.299295, -0.087209, 0.09016, -0.60906], [-0.216, 0.133885, -0.331028, -0.308479, -0.304639], [-0.157251, 0.298345, -0.358622, -0.031719, -0.391069], [-0.220715, 0.170381, 0.163363, -0.205153, -0.394304]], "network.4.bias": [-0.227825, -0.011755, -0.257563, -0.516093, -0.236089], "network.6.weight": [[-0.042127, -0.465679, -0.470258, -0.196322, 0.128895], [1.007784, 0.374313, -0.147033, -0.015715, -0.122353], [0.432743, -0.318056, -0.160762, -0.445011, -0.308509], [0.340809, -0.016885, -0.068663, 0.39103, 0.141808], [-0.374824, -0.129411, -0.330502, 0.082211, 0.052028]], "network.6.bias": [-0.048549, 0.186436, -0.328514, 0.390323, -0.28405], "network.8.weight": [[0.050321, -0.388125, 0.037718, -0.124911, -0.107027], [0.347513, 0.08728, 0.335453, 0.426182, -0.17585], [0.260371, 0.571911, 0.489411, 0.184274, -0.158506], [0.087971, -0.242801, 0.124935, -0.613978, -0.217582], [0.218746, -0.241364, -0.021965, 0.177095, 0.129602]], "network.8.bias": [-0.405845, -0.29877, -0.226859, 0.046574, -0.439126], "network.10.weight": [[0.050986, -0.358228, 0.098848, 0.318, -0.196852], [-0.156583, 0.117455, 0.474129, -0.036744, 0.029508], [0.240726, -0.290944, -0.530335, 0.227197, 0.268918], [-0.176356, 0.30123, 0.279669, 0.005378, -0.411836], [0.183437, 0.35377, 0.004211, -0.190498, 0.32837]], "network.10.bias": [-0.251292, -0.087476, 1.231301, 0.02906, -0.231036], "network.12.weight": [[0.15346, 0.616833, -1.518517, 0.482397, 0.226028]], "network.12.bias": [-0.417011]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.7002740502357483, "train_acc": 0.465, "val_loss": 0.6883068680763245, "val_acc": 0.56}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6886576414108276, "train_acc": 0.535, "val_loss": 0.687092125415802, "val_acc": 0.56}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6898380517959595, "train_acc": 0.535, "val_loss": 0.676719069480896, "val_acc": 0.56}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6745440363883972, "train_acc": 0.55, "val_loss": 0.644247829914093, "val_acc": 0.68}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.6010787487030029, "train_acc": 0.73, "val_loss": 0.617027997970581, "val_acc": 0.7}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.561182290315628, "train_acc": 0.74, "val_loss": 0.4836534559726715, "val_acc": 0.78}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.4909709692001343, "train_acc": 0.775, "val_loss": 0.3813009560108185, "val_acc": 0.88}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.43797364830970764, "train_acc": 0.835, "val_loss": 0.33192405104637146, "val_acc": 0.92}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.41049644351005554, "train_acc": 0.845, "val_loss": 0.31850650906562805, "val_acc": 0.92}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.34310562908649445, "train_acc": 0.86, "val_loss": 0.2763127088546753, "val_acc": 0.94}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.34799911081790924, "train_acc": 0.87, "val_loss": 0.2633906900882721, "val_acc": 0.92}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.3126722127199173, "train_acc": 0.875, "val_loss": 0.2582755386829376, "val_acc": 0.92}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.2870781272649765, "train_acc": 0.89, "val_loss": 0.2516591548919678, "val_acc": 0.92}, {"stage": "improved", "epoch": 9, "global_epoch": 13, "train_loss": 0.27057476341724396, "train_acc": 0.9, "val_loss": 0.22699283063411713, "val_acc": 0.94}], "summary": {"total_epochs": 14, "degraded_epochs": 4, "improved_epochs": 10, "patterns": ["sorted_ascending"], "degraded_stage": {"initial_val_loss": 0.6883068680763245, "final_val_loss": 0.644247829914093, "initial_val_acc": 0.56, "final_val_acc": 0.68, "best_val_acc": 0.68}, "improved_stage": {"initial_val_loss": 0.617027997970581, "final_val_loss": 0.22699283063411713, "initial_val_acc": 0.7, "final_val_acc": 0.94, "best_val_acc": 0.94, "best_epoch": 9}, "improvement": 0.2599999999999999, "first_improvement_epoch": 3}}
|
95
|
{"target_pattern": "mountain_pattern", "degraded_accuracy": 0.58, "improved_accuracy": 0.92, "improvement": 0.3400000000000001, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 5234, "learning_rate": 0.05054527604583981, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "mountain_pattern", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["mountain_pattern"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.314425,
-0.417277,
-0.026703,
-0.014668,
-0.659435
],
[
-1.416766,
-0.805619,
0.274216,
0.483875,
0.027112
],
[
0.760073,
0.803172,
-0.078247,
-0.263218,
0.333036
],
[
0.125868,
-0.029179,
0.226605,
-0.711747,
0.229402
],
[
-0.414445,
-0.46763,
0.175761,
0.055566,
-0.209686
],
[
-0.755682,
-0.099184,
0.059371,
-0.531853,
-0.206547
],
[
0.862806,
-0.071103,
-0.56626,
-0.002331,
0.396109
]
],
"network.0.bias": [
0.066554,
0.138342,
0.456079,
-0.24033,
-0.164227,
-0.285014,
0.323291
],
"network.2.weight": [
[
0.018306,
-0.446642,
0.706644,
0.735987,
-0.271927,
0.114859,
0.300388
],
[
-0.380401,
-0.382377,
-0.380972,
0.061962,
-0.493432,
0.173018,
0.122134
],
[
0.14727,
-0.040644,
-0.359542,
-0.18267,
-0.103081,
0.427276,
-0.036055
],
[
-0.177045,
-0.207144,
0.245874,
0.23559,
0.242214,
0.023774,
0.163363
],
[
-0.323718,
-0.171233,
0.02076,
-0.254871,
-0.569992,
-0.071728,
-0.397009
],
[
0.148756,
0.375187,
-0.105666,
-0.085403,
0.13249,
-0.136469,
-0.623099
],
[
0.227943,
-0.52875,
0.479731,
0.154326,
0.195866,
-0.106165,
0.258899
]
],
"network.2.bias": [
0.013863,
-0.170617,
-0.193343,
0.147354,
-0.187647,
-0.117722,
0.47627
],
"network.4.weight": [
[
-0.281718,
0.309665,
0.269901,
-0.197056,
-0.045141,
-0.283916,
-0.210933
],
[
-0.161653,
-0.323992,
-0.490237,
-0.541604,
0.207791,
-0.057667,
-0.213319
],
[
-0.277192,
-0.095696,
-0.158831,
-0.19221,
-0.27014,
-0.107159,
-0.552276
],
[
0.587403,
-0.145809,
-0.292626,
0.040892,
-0.120977,
-0.398269,
0.711691
],
[
-0.802478,
-0.052872,
0.029904,
-0.090592,
-0.136366,
0.22729,
-0.320134
],
[
-0.065815,
-0.137403,
0.10031,
-0.216028,
0.00352,
0.117371,
-0.344361
],
[
-0.364916,
0.161233,
0.163758,
-0.412789,
-0.250923,
-0.566865,
-0.221111
]
],
"network.4.bias": [
-0.314816,
-0.175662,
-0.235518,
-0.023393,
0.048511,
-0.430548,
-0.429115
],
"network.6.weight": [
[
-0.176521,
-0.254027,
0.480333,
0.871775,
0.010755,
0.059408,
-0.019601
],
[
0.00971,
0.199531,
-0.340913,
-0.018486,
0.085514,
0.046551,
-0.123556
],
[
-0.017338,
-0.097165,
-0.069209,
0.494743,
-0.345829,
-0.476132,
0.333956
],
[
-0.012462,
0.034535,
-0.029322,
0.727583,
-0.222568,
0.10189,
-0.107995
],
[
0.207192,
0.089232,
-0.204887,
-0.33795,
0.220345,
-0.260333,
-0.089602
],
[
0.077076,
0.369405,
-0.010598,
-0.114706,
0.213149,
0.112216,
0.029057
],
[
-0.068569,
0.330292,
0.324333,
-0.058566,
-0.136655,
0.301137,
-0.280385
]
],
"network.6.bias": [
-0.04579,
-0.357835,
-0.068215,
0.139131,
0.141968,
-0.271122,
-0.338105
],
"network.8.weight": [
[
0.726907,
-0.176365,
0.103124,
0.43727,
-0.137379,
-0.349505,
0.20274
],
[
0.177858,
0.285072,
0.231711,
0.112785,
-0.214589,
-0.123061,
-0.092085
],
[
-0.148527,
-0.327418,
-0.244924,
-0.16735,
-0.116902,
-0.243723,
-0.072062
],
[
0.503501,
0.058622,
0.711018,
0.490414,
-0.085195,
-0.272593,
0.20645
],
[
-0.145363,
-0.283947,
0.105326,
-0.003862,
-0.278628,
-0.192685,
-0.201527
],
[
0.520622,
0.16077,
0.512029,
0.450253,
0.257112,
0.366304,
0.090954
],
[
-0.56889,
-0.011933,
-0.212973,
-0.306643,
0.505373,
-0.065965,
0.270248
]
],
"network.8.bias": [
0.0215,
0.059759,
-0.714695,
0.085714,
0.103721,
-0.144209,
0.539037
],
"network.10.weight": [
[
-0.006876,
0.135178,
0.428271,
-0.452046,
-0.491178,
-0.418462,
-0.57707
],
[
0.090728,
0.165262,
0.088841,
0.239886,
0.014108,
0.431333,
-0.545028
],
[
-0.174005,
-0.144001,
0.017763,
0.201371,
-0.072619,
-0.424914,
-0.368714
],
[
0.55844,
0.192664,
0.026249,
0.563504,
0.149264,
0.521144,
-0.095257
],
[
0.386899,
-0.0978,
0.052936,
0.417483,
-0.075349,
0.247062,
-0.513789
],
[
0.101448,
-0.135925,
-0.289569,
0.000331,
-0.24478,
-0.051023,
0.035995
],
[
-0.159098,
-0.302213,
0.29959,
-0.015111,
0.003603,
-0.198617,
0.077576
]
],
"network.10.bias": [
-0.201809,
0.011813,
-0.210794,
-0.311132,
-0.24093,
-0.292703,
-0.070399
],
"network.12.weight": [
[
0.089586,
-0.196018,
-0.132675,
-0.396877,
-0.311907,
0.173247,
0.371508
]
],
"network.12.bias": [
0.579689
]
}
## Activation Signature
### 0
mean: [-1.985638, -1.448875, 2.548344, -0.906885, -1.304414, -2.677407, 0.568121]
std: [1.783149, 3.547228, 2.509776, 1.646343, 1.376338, 2.032847, 1.907434]
### 2
mean: [2.095607, -1.234522, -1.230045, 0.914088, -0.727830, -0.858662, 1.759281]
std: [2.418010, 0.705366, 0.933397, 0.916661, 0.564942, 1.247538, 1.738678]
### 4
mean: [-1.528613, -1.430495, -2.038531, 2.558280, -2.355527, -1.391919, -2.073568]
std: [1.148637, 1.202621, 1.714685, 2.610888, 2.485059, 0.927173, 1.525774]
### 6
mean: [2.217567, -0.403510, 1.206751, 2.021866, -0.729368, -0.563127, -0.493835]
std: [2.241169, 0.050156, 1.281686, 1.876853, 0.875046, 0.301047, 0.147106]
### 8
mean: [2.645131, 0.961127, -1.687857, 3.067564, -0.107055, 2.561133, -1.592078]
std: [2.578772, 0.907900, 0.950969, 2.944145, 0.190999, 2.645418, 2.130904]
### 10
mean: [-2.628834, 2.192300, -1.327767, 4.411078, 2.543152, -0.285963, -1.329431]
std: [2.258632, 2.291321, 1.066923, 4.656289, 2.850976, 0.006811, 1.261783]
### 12
mean: [-2.446954]
std: [3.133005]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
mountain_pattern
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.314425,
-0.417277,
-0.026703,
-0.014668,
-0.659435
],
[
-1.416766,
-0.805619,
0.274216,
0.483875,
0.027112
],
[
0.760073,
0.803172,
-0.078247,
-0.263218,
0.333036
],
[
0.125868,
-0.029179,
0.226605,
-0.711747,
0.229402
],
[
-0.414445,
-0.46763,
0.175761,
0.055566,
-0.209686
],
[
-0.755682,
-0.099184,
0.059371,
-0.531853,
-0.206547
],
[
0.862806,
-0.071103,
-0.56626,
-0.002331,
0.396109
]
],
"network.0.bias": [
0.066554,
0.138342,
0.456079,
-0.24033,
-0.164227,
-0.285014,
0.323291
],
"network.2.weight": [
[
0.018306,
-0.446642,
0.706644,
0.735987,
-0.271927,
0.114859,
0.300388
],
[
-0.380401,
-0.382377,
-0.380972,
0.061962,
-0.493432,
0.173018,
0.122134
],
[
0.14727,
-0.040644,
-0.359542,
-0.18267,
-0.103081,
0.427276,
-0.036055
],
[
-0.177045,
-0.207144,
0.245874,
0.23559,
0.242214,
0.023774,
0.163363
],
[
-0.323718,
-0.171233,
0.02076,
-0.254871,
-0.569992,
-0.071728,
-0.397009
],
[
0.148756,
0.375187,
-0.105666,
-0.085403,
0.13249,
-0.136469,
-0.623099
],
[
0.227943,
-0.52875,
0.479731,
0.154326,
0.195866,
-0.106165,
0.258899
]
],
"network.2.bias": [
0.013863,
-0.170617,
-0.193343,
0.147354,
-0.187647,
-0.117722,
0.47627
],
"network.4.weight": [
[
-0.281718,
0.309665,
0.269901,
-0.197056,
-0.045141,
-0.283916,
-0.210933
],
[
-0.161653,
-0.323992,
-0.490237,
-0.541604,
0.207791,
-0.057667,
-0.213319
],
[
-0.277192,
-0.095696,
-0.158831,
-0.19221,
-0.27014,
-0.107159,
-0.552276
],
[
0.587403,
-0.145809,
-0.292626,
0.040892,
-0.120977,
-0.398269,
0.711691
],
[
-0.802478,
-0.052872,
0.029904,
-0.090592,
-0.136366,
0.22729,
-0.320134
],
[
-0.065815,
-0.137403,
0.10031,
-0.216028,
0.00352,
0.117371,
-0.344361
],
[
-0.364916,
0.161233,
0.163758,
-0.412789,
-0.250923,
-0.566865,
-0.221111
]
],
"network.4.bias": [
-0.314816,
-0.175662,
-0.235518,
-0.023393,
0.048511,
-0.430548,
-0.429115
],
"network.6.weight": [
[
-0.176521,
-0.254027,
0.480333,
0.871775,
0.010755,
0.059408,
-0.019601
],
[
0.00971,
0.199531,
-0.340913,
-0.018486,
0.085514,
0.046551,
-0.123556
],
[
-0.017338,
-0.097165,
-0.069209,
0.494743,
-0.345829,
-0.476132,
0.333956
],
[
-0.012462,
0.034535,
-0.029322,
0.727583,
-0.222568,
0.10189,
-0.107995
],
[
0.207192,
0.089232,
-0.204887,
-0.33795,
0.220345,
-0.260333,
-0.089602
],
[
0.077076,
0.369405,
-0.010598,
-0.114706,
0.213149,
0.112216,
0.029057
],
[
-0.068569,
0.330292,
0.324333,
-0.058566,
-0.136655,
0.301137,
-0.280385
]
],
"network.6.bias": [
-0.04579,
-0.357835,
-0.068215,
0.139131,
0.141968,
-0.271122,
-0.338105
],
"network.8.weight": [
[
0.726907,
-0.176365,
0.103124,
0.43727,
-0.137379,
-0.349505,
0.20274
],
[
0.177858,
0.285072,
0.231711,
0.112785,
-0.214589,
-0.123061,
-0.092085
],
[
-0.148527,
-0.327418,
-0.244924,
-0.16735,
-0.116902,
-0.243723,
-0.072062
],
[
0.503501,
0.058622,
0.711018,
0.490414,
-0.085195,
-0.272593,
0.20645
],
[
-0.145363,
-0.283947,
0.105326,
-0.003862,
-0.278628,
-0.192685,
-0.201527
],
[
0.520622,
0.16077,
0.512029,
0.450253,
0.257112,
0.366304,
0.090954
],
[
-0.56889,
-0.011933,
-0.212973,
-0.306643,
0.505373,
-0.065965,
0.270248
]
],
"network.8.bias": [
0.0215,
0.059759,
-0.714695,
0.085714,
0.103721,
-0.144209,
0.539037
],
"network.10.weight": [
[
-0.006876,
0.135178,
0.428271,
-0.452046,
-0.491178,
-0.418462,
-0.57707
],
[
0.090728,
0.165262,
0.088841,
0.239886,
0.014108,
0.431333,
-0.545028
],
[
-0.174005,
-0.144001,
0.017763,
0.201371,
-0.072619,
-0.424914,
-0.368714
],
[
0.55844,
0.192664,
0.026249,
0.563504,
0.149264,
0.521144,
-0.095257
],
[
0.386899,
-0.0978,
0.052936,
0.417483,
-0.075349,
0.247062,
-0.513789
],
[
0.101448,
-0.135925,
-0.289569,
0.000331,
-0.24478,
-0.051023,
0.035995
],
[
-0.159098,
-0.302213,
0.29959,
-0.015111,
0.003603,
-0.198617,
0.077576
]
],
"network.10.bias": [
-0.201809,
0.011813,
-0.210794,
-0.311132,
-0.24093,
-0.292703,
-0.070399
],
"network.12.weight": [
[
0.089586,
-0.196018,
-0.132675,
-0.396877,
-0.311907,
0.173247,
0.371508
]
],
"network.12.bias": [
0.579689
]
}
## Activation Signature
### 0
mean: [-1.985638, -1.448875, 2.548344, -0.906885, -1.304414, -2.677407, 0.568121]
std: [1.783149, 3.547228, 2.509776, 1.646343, 1.376338, 2.032847, 1.907434]
### 2
mean: [2.095607, -1.234522, -1.230045, 0.914088, -0.727830, -0.858662, 1.759281]
std: [2.418010, 0.705366, 0.933397, 0.916661, 0.564942, 1.247538, 1.738678]
### 4
mean: [-1.528613, -1.430495, -2.038531, 2.558280, -2.355527, -1.391919, -2.073568]
std: [1.148637, 1.202621, 1.714685, 2.610888, 2.485059, 0.927173, 1.525774]
### 6
mean: [2.217567, -0.403510, 1.206751, 2.021866, -0.729368, -0.563127, -0.493835]
std: [2.241169, 0.050156, 1.281686, 1.876853, 0.875046, 0.301047, 0.147106]
### 8
mean: [2.645131, 0.961127, -1.687857, 3.067564, -0.107055, 2.561133, -1.592078]
std: [2.578772, 0.907900, 0.950969, 2.944145, 0.190999, 2.645418, 2.130904]
### 10
mean: [-2.628834, 2.192300, -1.327767, 4.411078, 2.543152, -0.285963, -1.329431]
std: [2.258632, 2.291321, 1.066923, 4.656289, 2.850976, 0.006811, 1.261783]
### 12
mean: [-2.446954]
std: [3.133005]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
mountain_pattern
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -1.9856383800506592, "std": 1.7831486463546753}, "1": {"mean": -1.4488753080368042, "std": 3.547227621078491}, "2": {"mean": 2.5483436584472656, "std": 2.5097763538360596}, "3": {"mean": -0.9068846702575684, "std": 1.6463425159454346}, "4": {"mean": -1.3044136762619019, "std": 1.3763376474380493}, "5": {"mean": -2.6774070262908936, "std": 2.0328469276428223}, "6": {"mean": 0.568121075630188, "std": 1.9074338674545288}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 2.095606565475464, "std": 2.4180104732513428}, "1": {"mean": -1.2345223426818848, "std": 0.7053660750389099}, "2": {"mean": -1.2300454378128052, "std": 0.9333967566490173}, "3": {"mean": 0.9140881896018982, "std": 0.9166613221168518}, "4": {"mean": -0.7278300523757935, "std": 0.5649415254592896}, "5": {"mean": -0.8586618304252625, "std": 1.2475380897521973}, "6": {"mean": 1.7592813968658447, "std": 1.738677978515625}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -1.5286133289337158, "std": 1.1486365795135498}, "1": {"mean": -1.430495023727417, "std": 1.2026206254959106}, "2": {"mean": -2.0385313034057617, "std": 1.7146847248077393}, "3": {"mean": 2.558279514312744, "std": 2.6108884811401367}, "4": {"mean": -2.355527400970459, "std": 2.4850592613220215}, "5": {"mean": -1.391918659210205, "std": 0.9271734952926636}, "6": {"mean": -2.073568105697632, "std": 1.5257740020751953}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 2.2175674438476562, "std": 2.2411694526672363}, "1": {"mean": -0.40350964665412903, "std": 0.050155915319919586}, "2": {"mean": 1.2067514657974243, "std": 1.2816860675811768}, "3": {"mean": 2.0218658447265625, "std": 1.8768529891967773}, "4": {"mean": -0.7293678522109985, "std": 0.8750459551811218}, "5": {"mean": -0.5631273984909058, "std": 0.3010473847389221}, "6": {"mean": -0.4938347637653351, "std": 0.14710569381713867}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 2.6451306343078613, "std": 2.5787715911865234}, "1": {"mean": 0.9611266255378723, "std": 0.9078996181488037}, "2": {"mean": -1.6878573894500732, "std": 0.950968861579895}, "3": {"mean": 3.0675642490386963, "std": 2.9441447257995605}, "4": {"mean": -0.10705500841140747, "std": 0.19099865853786469}, "5": {"mean": 2.5611326694488525, "std": 2.645418167114258}, "6": {"mean": -1.5920783281326294, "std": 2.130904197692871}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": -2.6288340091705322, "std": 2.258631944656372}, "1": {"mean": 2.1922998428344727, "std": 2.291321039199829}, "2": {"mean": -1.3277671337127686, "std": 1.0669230222702026}, "3": {"mean": 4.411077976226807, "std": 4.6562886238098145}, "4": {"mean": 2.543152093887329, "std": 2.850975751876831}, "5": {"mean": -0.28596264123916626, "std": 0.006810827646404505}, "6": {"mean": -1.3294309377670288, "std": 1.2617830038070679}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -2.4469544887542725, "std": 3.133005142211914}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.314425, -0.417277, -0.026703, -0.014668, -0.659435], [-1.416766, -0.805619, 0.274216, 0.483875, 0.027112], [0.760073, 0.803172, -0.078247, -0.263218, 0.333036], [0.125868, -0.029179, 0.226605, -0.711747, 0.229402], [-0.414445, -0.46763, 0.175761, 0.055566, -0.209686], [-0.755682, -0.099184, 0.059371, -0.531853, -0.206547], [0.862806, -0.071103, -0.56626, -0.002331, 0.396109]], "network.0.bias": [0.066554, 0.138342, 0.456079, -0.24033, -0.164227, -0.285014, 0.323291], "network.2.weight": [[0.018306, -0.446642, 0.706644, 0.735987, -0.271927, 0.114859, 0.300388], [-0.380401, -0.382377, -0.380972, 0.061962, -0.493432, 0.173018, 0.122134], [0.14727, -0.040644, -0.359542, -0.18267, -0.103081, 0.427276, -0.036055], [-0.177045, -0.207144, 0.245874, 0.23559, 0.242214, 0.023774, 0.163363], [-0.323718, -0.171233, 0.02076, -0.254871, -0.569992, -0.071728, -0.397009], [0.148756, 0.375187, -0.105666, -0.085403, 0.13249, -0.136469, -0.623099], [0.227943, -0.52875, 0.479731, 0.154326, 0.195866, -0.106165, 0.258899]], "network.2.bias": [0.013863, -0.170617, -0.193343, 0.147354, -0.187647, -0.117722, 0.47627], "network.4.weight": [[-0.281718, 0.309665, 0.269901, -0.197056, -0.045141, -0.283916, -0.210933], [-0.161653, -0.323992, -0.490237, -0.541604, 0.207791, -0.057667, -0.213319], [-0.277192, -0.095696, -0.158831, -0.19221, -0.27014, -0.107159, -0.552276], [0.587403, -0.145809, -0.292626, 0.040892, -0.120977, -0.398269, 0.711691], [-0.802478, -0.052872, 0.029904, -0.090592, -0.136366, 0.22729, -0.320134], [-0.065815, -0.137403, 0.10031, -0.216028, 0.00352, 0.117371, -0.344361], [-0.364916, 0.161233, 0.163758, -0.412789, -0.250923, -0.566865, -0.221111]], "network.4.bias": [-0.314816, -0.175662, -0.235518, -0.023393, 0.048511, -0.430548, -0.429115], "network.6.weight": [[-0.176521, -0.254027, 0.480333, 0.871775, 0.010755, 0.059408, -0.019601], [0.00971, 0.199531, -0.340913, -0.018486, 0.085514, 0.046551, -0.123556], [-0.017338, -0.097165, -0.069209, 0.494743, -0.345829, -0.476132, 0.333956], [-0.012462, 0.034535, -0.029322, 0.727583, -0.222568, 0.10189, -0.107995], [0.207192, 0.089232, -0.204887, -0.33795, 0.220345, -0.260333, -0.089602], [0.077076, 0.369405, -0.010598, -0.114706, 0.213149, 0.112216, 0.029057], [-0.068569, 0.330292, 0.324333, -0.058566, -0.136655, 0.301137, -0.280385]], "network.6.bias": [-0.04579, -0.357835, -0.068215, 0.139131, 0.141968, -0.271122, -0.338105], "network.8.weight": [[0.726907, -0.176365, 0.103124, 0.43727, -0.137379, -0.349505, 0.20274], [0.177858, 0.285072, 0.231711, 0.112785, -0.214589, -0.123061, -0.092085], [-0.148527, -0.327418, -0.244924, -0.16735, -0.116902, -0.243723, -0.072062], [0.503501, 0.058622, 0.711018, 0.490414, -0.085195, -0.272593, 0.20645], [-0.145363, -0.283947, 0.105326, -0.003862, -0.278628, -0.192685, -0.201527], [0.520622, 0.16077, 0.512029, 0.450253, 0.257112, 0.366304, 0.090954], [-0.56889, -0.011933, -0.212973, -0.306643, 0.505373, -0.065965, 0.270248]], "network.8.bias": [0.0215, 0.059759, -0.714695, 0.085714, 0.103721, -0.144209, 0.539037], "network.10.weight": [[-0.006876, 0.135178, 0.428271, -0.452046, -0.491178, -0.418462, -0.57707], [0.090728, 0.165262, 0.088841, 0.239886, 0.014108, 0.431333, -0.545028], [-0.174005, -0.144001, 0.017763, 0.201371, -0.072619, -0.424914, -0.368714], [0.55844, 0.192664, 0.026249, 0.563504, 0.149264, 0.521144, -0.095257], [0.386899, -0.0978, 0.052936, 0.417483, -0.075349, 0.247062, -0.513789], [0.101448, -0.135925, -0.289569, 0.000331, -0.24478, -0.051023, 0.035995], [-0.159098, -0.302213, 0.29959, -0.015111, 0.003603, -0.198617, 0.077576]], "network.10.bias": [-0.201809, 0.011813, -0.210794, -0.311132, -0.24093, -0.292703, -0.070399], "network.12.weight": [[0.089586, -0.196018, -0.132675, -0.396877, -0.311907, 0.173247, 0.371508]], "network.12.bias": [0.579689]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.683845579624176, "train_acc": 0.56, "val_loss": 0.6806833744049072, "val_acc": 0.58}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6866803467273712, "train_acc": 0.56, "val_loss": 0.6803072094917297, "val_acc": 0.58}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6910851895809174, "train_acc": 0.56, "val_loss": 0.6786673069000244, "val_acc": 0.58}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.6832790076732635, "train_acc": 0.56, "val_loss": 0.664073646068573, "val_acc": 0.58}, {"stage": "degraded", "epoch": 4, "global_epoch": 4, "train_loss": 0.6608144044876099, "train_acc": 0.56, "val_loss": 0.6301705837249756, "val_acc": 0.58}, {"stage": "improved", "epoch": 0, "global_epoch": 5, "train_loss": 0.6675659120082855, "train_acc": 0.48, "val_loss": 0.5985166430473328, "val_acc": 0.58}, {"stage": "improved", "epoch": 1, "global_epoch": 6, "train_loss": 0.5994535684585571, "train_acc": 0.48, "val_loss": 0.591559648513794, "val_acc": 0.76}, {"stage": "improved", "epoch": 2, "global_epoch": 7, "train_loss": 0.545037716627121, "train_acc": 0.76, "val_loss": 0.41892653703689575, "val_acc": 0.88}, {"stage": "improved", "epoch": 3, "global_epoch": 8, "train_loss": 0.5099323987960815, "train_acc": 0.81, "val_loss": 0.4155119061470032, "val_acc": 0.88}, {"stage": "improved", "epoch": 4, "global_epoch": 9, "train_loss": 0.49464522302150726, "train_acc": 0.805, "val_loss": 0.3494323790073395, "val_acc": 0.9}, {"stage": "improved", "epoch": 5, "global_epoch": 10, "train_loss": 0.4713153541088104, "train_acc": 0.83, "val_loss": 0.3145800828933716, "val_acc": 0.88}, {"stage": "improved", "epoch": 6, "global_epoch": 11, "train_loss": 0.47638463973999023, "train_acc": 0.86, "val_loss": 0.31724485754966736, "val_acc": 0.9}, {"stage": "improved", "epoch": 7, "global_epoch": 12, "train_loss": 0.4183507710695267, "train_acc": 0.86, "val_loss": 0.3222111165523529, "val_acc": 0.92}, {"stage": "improved", "epoch": 8, "global_epoch": 13, "train_loss": 0.4139978736639023, "train_acc": 0.84, "val_loss": 0.3211839199066162, "val_acc": 0.92}], "summary": {"total_epochs": 14, "degraded_epochs": 5, "improved_epochs": 9, "patterns": ["mountain_pattern"], "degraded_stage": {"initial_val_loss": 0.6806833744049072, "final_val_loss": 0.6301705837249756, "initial_val_acc": 0.58, "final_val_acc": 0.58, "best_val_acc": 0.58}, "improved_stage": {"initial_val_loss": 0.5985166430473328, "final_val_loss": 0.3211839199066162, "initial_val_acc": 0.58, "final_val_acc": 0.92, "best_val_acc": 0.92, "best_epoch": 12}, "improvement": 0.3400000000000001, "first_improvement_epoch": 4}}
|
96
|
{"target_pattern": "first_last_match", "degraded_accuracy": 0.5, "improved_accuracy": 0.9, "improvement": 0.4, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 3153, "learning_rate": 0.09375952855635061, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "first_last_match", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["first_last_match"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.490655,
-0.890976,
-0.993846,
0.00125,
-0.954278
],
[
-1.377325,
-0.498339,
-0.00606,
-0.292589,
2.000451
],
[
1.241971,
-0.177386,
-0.29473,
-0.283594,
0.795638
],
[
0.763351,
0.369734,
-0.004319,
0.08309,
1.337082
],
[
-0.970683,
0.388461,
0.903079,
0.663176,
-1.127279
],
[
-1.271977,
0.110805,
-0.061307,
-0.009472,
1.399718
],
[
-0.656826,
-0.425906,
-0.659909,
0.029156,
1.541946
],
[
-0.121218,
0.07403,
-0.046074,
-0.420979,
1.429253
]
],
"network.0.bias": [
-0.02633,
-0.092693,
0.73329,
0.580722,
-0.01594,
-0.520773,
-0.548249,
-0.201367
],
"network.2.weight": [
[
0.272346,
0.791525,
1.174763,
0.895801,
-0.387117,
0.942643,
0.965652,
0.424796
],
[
0.356493,
-0.830703,
-0.429065,
0.059607,
0.33109,
-0.98839,
0.090925,
0.129683
],
[
0.203864,
0.838221,
0.792427,
0.658546,
-0.641463,
1.077829,
0.468809,
-0.138143
],
[
-0.055045,
-0.296334,
-0.617526,
-0.183018,
-0.401407,
-0.226686,
-0.215169,
-0.390271
],
[
-0.318928,
1.379205,
1.038775,
0.938613,
-0.911685,
1.296718,
0.007716,
0.120276
],
[
0.137758,
-0.432547,
-0.64261,
-0.200839,
0.946056,
-1.125124,
-0.415886,
0.305264
],
[
0.657575,
-1.240146,
0.183847,
0.339754,
0.434559,
-1.067589,
0.119712,
0.25117
],
[
0.811236,
-1.019395,
-0.39028,
-0.217465,
0.715895,
-0.672729,
0.462858,
-0.050271
]
],
"network.2.bias": [
0.592355,
0.280901,
-0.293282,
-0.750964,
-0.181286,
0.319355,
0.046279,
-0.01505
],
"network.4.weight": [
[
0.745928,
-0.328661,
0.224579,
0.326194,
0.525038,
-0.95037,
-0.27819,
-0.160564
],
[
-0.194589,
-0.10026,
-0.426141,
0.002479,
-0.459962,
-0.529323,
-0.297988,
0.03369
],
[
0.015551,
-0.246094,
-0.055173,
0.085254,
-0.819129,
-0.1481,
-0.575338,
0.061771
],
[
-0.564345,
0.784598,
-0.549108,
-0.317714,
-0.258917,
-0.867923,
0.172177,
0.586397
],
[
-0.299149,
-0.00306,
-0.218182,
-0.006585,
-0.699038,
-0.722472,
-0.311629,
-0.288952
],
[
0.93402,
-0.519111,
0.493004,
0.232543,
0.66777,
-0.668561,
-0.539197,
-0.566699
],
[
-0.17662,
-0.369551,
-0.217559,
-0.1051,
-0.114685,
-0.208908,
-0.179136,
0.203311
],
[
-1.023039,
0.164955,
-0.729578,
-0.23867,
-0.041209,
-0.412591,
0.585724,
0.737531
]
],
"network.4.bias": [
-0.03677,
-0.044589,
0.021343,
-0.371403,
-0.603666,
0.235664,
-0.495734,
0.307357
],
"network.6.weight": [
[
-0.13934,
-0.308205,
-0.092561,
-0.277093,
0.286648,
0.041159,
-0.547652,
0.01178
],
[
0.339022,
0.005918,
-0.028751,
-0.179008,
0.143681,
0.528794,
-0.109844,
-0.284687
],
[
-0.134226,
-0.119081,
-0.190176,
0.047651,
-0.033195,
-0.335464,
-0.038213,
-0.168899
],
[
-0.171875,
0.519424,
-0.526412,
-0.771659,
0.020968,
-0.938674,
0.053015,
-0.093101
],
[
0.355845,
0.065637,
-0.525138,
-0.11884,
0.289571,
0.760917,
0.183523,
-0.312251
],
[
-0.069802,
-0.01271,
0.6764,
0.004161,
-0.51545,
-0.335986,
0.044062,
-0.492812
],
[
-0.183597,
-0.003792,
0.774058,
-0.25863,
-0.559416,
-0.372914,
0.433465,
-0.056768
],
[
-0.256778,
0.020156,
-0.174051,
-0.236685,
0.144126,
-0.030602,
-0.275677,
0.162399
]
],
"network.6.bias": [
-0.516136,
-0.438314,
-0.391068,
-0.554872,
-0.033515,
0.148889,
-0.137986,
-0.151041
],
"network.8.weight": [
[
-0.057228,
0.332229,
0.073358,
0.316957,
-0.357482,
-0.101428,
-0.413569,
0.126131
],
[
0.023226,
-0.225303,
-0.316204,
-0.166543,
-0.168512,
-0.691871,
-0.64658,
-0.255912
],
[
-0.206992,
0.015238,
0.337925,
-0.294443,
-0.517188,
-0.41446,
-0.375646,
-0.215216
],
[
0.228357,
0.171525,
-0.083359,
-0.094571,
-0.279522,
-0.215949,
0.047006,
-0.024072
],
[
-0.120499,
-0.315261,
-0.158613,
0.24782,
-0.269798,
-0.572565,
-0.593069,
-0.258485
],
[
0.349892,
0.470166,
0.108741,
-0.022046,
0.559328,
0.109815,
0.232849,
0.35022
],
[
-0.218153,
0.025524,
0.176049,
-0.278457,
-0.034801,
-0.051801,
-0.155511,
0.326797
],
[
0.195893,
-0.212903,
-0.081911,
-0.153246,
-0.149833,
-0.77111,
-0.759632,
-0.138665
]
],
"network.8.bias": [
-0.1418,
-0.760358,
-0.370544,
-0.267924,
0.029508,
-0.198083,
-0.349462,
-0.546599
],
"network.10.weight": [
[
0.058118,
0.50047,
0.187995,
0.024199,
0.101228,
0.285054,
-0.301016,
0.248881
],
[
-0.075758,
0.038293,
0.270912,
-0.246458,
0.066369,
-0.20687,
-0.098871,
0.305063
],
[
0.155889,
0.183556,
0.415545,
-0.092289,
0.029519,
0.494169,
0.057489,
0.179395
],
[
-0.263637,
0.318294,
0.093727,
-0.176092,
-0.171848,
0.210318,
-0.301772,
0.278717
],
[
0.107564,
-0.314612,
-0.246066,
0.16176,
-0.070373,
-0.020733,
0.140617,
-0.238411
],
[
0.31041,
0.210162,
-0.074565,
0.297651,
-0.165847,
-0.115569,
-0.307952,
0.107705
],
[
0.04185,
0.252412,
-0.349461,
0.101985,
-0.116916,
-0.306541,
0.116585,
-0.203029
],
[
0.039274,
0.189525,
0.302032,
0.039347,
0.034091,
-0.326047,
-0.004839,
-0.172949
]
],
"network.10.bias": [
-0.47024,
-0.072295,
-0.507004,
-0.319887,
-0.007861,
-0.210522,
-0.018141,
-0.275257
],
"network.12.weight": [
[
-0.156145,
-0.027236,
-0.34175,
-0.159891,
-0.066776,
-0.032313,
-0.333177,
0.337983
]
],
"network.12.bias": [
1.050614
]
}
## Activation Signature
### 0
mean: [-5.510777, -0.887659, 1.713092, 4.022634, 1.402226, -0.323824, -1.555949, 0.546418]
std: [4.038497, 4.491000, 2.984743, 3.349240, 3.342164, 3.263705, 3.383164, 2.581876]
### 2
mean: [8.834291, -1.374465, 4.821536, -4.801170, 6.865030, -1.314334, 0.602655, -1.805314]
std: [10.568254, 4.296184, 8.131716, 3.290321, 11.115867, 6.019885, 4.618095, 5.174637]
### 4
mean: [9.443420, -9.292645, -8.109442, -10.398354, -12.134007, 13.377235, -4.935707, -11.784537]
std: [16.673845, 9.290627, 8.354321, 12.531136, 10.917912, 22.247318, 4.370909, 16.943039]
### 6
mean: [-1.461689, 11.502235, -7.291756, -17.255289, 15.706025, -6.255388, -8.108604, -3.454838]
std: [1.246368, 15.942326, 8.715746, 21.560286, 20.919336, 7.604318, 10.271538, 4.538529]
### 8
mean: [-1.890017, -6.144286, -8.453947, -2.692043, -8.036296, 14.325708, -0.578310, -5.477706]
std: [2.217360, 7.004000, 10.470514, 3.107436, 10.513025, 18.938929, 0.342668, 6.428535]
### 10
mean: [3.626991, -3.045466, 6.595653, 2.702626, -0.306001, -1.871910, -4.424235, -4.961388]
std: [5.388229, 3.910578, 9.341251, 3.975917, 0.391816, 2.184350, 5.794378, 6.163319]
### 12
mean: [-2.313604]
std: [4.586017]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
first_last_match
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.490655,
-0.890976,
-0.993846,
0.00125,
-0.954278
],
[
-1.377325,
-0.498339,
-0.00606,
-0.292589,
2.000451
],
[
1.241971,
-0.177386,
-0.29473,
-0.283594,
0.795638
],
[
0.763351,
0.369734,
-0.004319,
0.08309,
1.337082
],
[
-0.970683,
0.388461,
0.903079,
0.663176,
-1.127279
],
[
-1.271977,
0.110805,
-0.061307,
-0.009472,
1.399718
],
[
-0.656826,
-0.425906,
-0.659909,
0.029156,
1.541946
],
[
-0.121218,
0.07403,
-0.046074,
-0.420979,
1.429253
]
],
"network.0.bias": [
-0.02633,
-0.092693,
0.73329,
0.580722,
-0.01594,
-0.520773,
-0.548249,
-0.201367
],
"network.2.weight": [
[
0.272346,
0.791525,
1.174763,
0.895801,
-0.387117,
0.942643,
0.965652,
0.424796
],
[
0.356493,
-0.830703,
-0.429065,
0.059607,
0.33109,
-0.98839,
0.090925,
0.129683
],
[
0.203864,
0.838221,
0.792427,
0.658546,
-0.641463,
1.077829,
0.468809,
-0.138143
],
[
-0.055045,
-0.296334,
-0.617526,
-0.183018,
-0.401407,
-0.226686,
-0.215169,
-0.390271
],
[
-0.318928,
1.379205,
1.038775,
0.938613,
-0.911685,
1.296718,
0.007716,
0.120276
],
[
0.137758,
-0.432547,
-0.64261,
-0.200839,
0.946056,
-1.125124,
-0.415886,
0.305264
],
[
0.657575,
-1.240146,
0.183847,
0.339754,
0.434559,
-1.067589,
0.119712,
0.25117
],
[
0.811236,
-1.019395,
-0.39028,
-0.217465,
0.715895,
-0.672729,
0.462858,
-0.050271
]
],
"network.2.bias": [
0.592355,
0.280901,
-0.293282,
-0.750964,
-0.181286,
0.319355,
0.046279,
-0.01505
],
"network.4.weight": [
[
0.745928,
-0.328661,
0.224579,
0.326194,
0.525038,
-0.95037,
-0.27819,
-0.160564
],
[
-0.194589,
-0.10026,
-0.426141,
0.002479,
-0.459962,
-0.529323,
-0.297988,
0.03369
],
[
0.015551,
-0.246094,
-0.055173,
0.085254,
-0.819129,
-0.1481,
-0.575338,
0.061771
],
[
-0.564345,
0.784598,
-0.549108,
-0.317714,
-0.258917,
-0.867923,
0.172177,
0.586397
],
[
-0.299149,
-0.00306,
-0.218182,
-0.006585,
-0.699038,
-0.722472,
-0.311629,
-0.288952
],
[
0.93402,
-0.519111,
0.493004,
0.232543,
0.66777,
-0.668561,
-0.539197,
-0.566699
],
[
-0.17662,
-0.369551,
-0.217559,
-0.1051,
-0.114685,
-0.208908,
-0.179136,
0.203311
],
[
-1.023039,
0.164955,
-0.729578,
-0.23867,
-0.041209,
-0.412591,
0.585724,
0.737531
]
],
"network.4.bias": [
-0.03677,
-0.044589,
0.021343,
-0.371403,
-0.603666,
0.235664,
-0.495734,
0.307357
],
"network.6.weight": [
[
-0.13934,
-0.308205,
-0.092561,
-0.277093,
0.286648,
0.041159,
-0.547652,
0.01178
],
[
0.339022,
0.005918,
-0.028751,
-0.179008,
0.143681,
0.528794,
-0.109844,
-0.284687
],
[
-0.134226,
-0.119081,
-0.190176,
0.047651,
-0.033195,
-0.335464,
-0.038213,
-0.168899
],
[
-0.171875,
0.519424,
-0.526412,
-0.771659,
0.020968,
-0.938674,
0.053015,
-0.093101
],
[
0.355845,
0.065637,
-0.525138,
-0.11884,
0.289571,
0.760917,
0.183523,
-0.312251
],
[
-0.069802,
-0.01271,
0.6764,
0.004161,
-0.51545,
-0.335986,
0.044062,
-0.492812
],
[
-0.183597,
-0.003792,
0.774058,
-0.25863,
-0.559416,
-0.372914,
0.433465,
-0.056768
],
[
-0.256778,
0.020156,
-0.174051,
-0.236685,
0.144126,
-0.030602,
-0.275677,
0.162399
]
],
"network.6.bias": [
-0.516136,
-0.438314,
-0.391068,
-0.554872,
-0.033515,
0.148889,
-0.137986,
-0.151041
],
"network.8.weight": [
[
-0.057228,
0.332229,
0.073358,
0.316957,
-0.357482,
-0.101428,
-0.413569,
0.126131
],
[
0.023226,
-0.225303,
-0.316204,
-0.166543,
-0.168512,
-0.691871,
-0.64658,
-0.255912
],
[
-0.206992,
0.015238,
0.337925,
-0.294443,
-0.517188,
-0.41446,
-0.375646,
-0.215216
],
[
0.228357,
0.171525,
-0.083359,
-0.094571,
-0.279522,
-0.215949,
0.047006,
-0.024072
],
[
-0.120499,
-0.315261,
-0.158613,
0.24782,
-0.269798,
-0.572565,
-0.593069,
-0.258485
],
[
0.349892,
0.470166,
0.108741,
-0.022046,
0.559328,
0.109815,
0.232849,
0.35022
],
[
-0.218153,
0.025524,
0.176049,
-0.278457,
-0.034801,
-0.051801,
-0.155511,
0.326797
],
[
0.195893,
-0.212903,
-0.081911,
-0.153246,
-0.149833,
-0.77111,
-0.759632,
-0.138665
]
],
"network.8.bias": [
-0.1418,
-0.760358,
-0.370544,
-0.267924,
0.029508,
-0.198083,
-0.349462,
-0.546599
],
"network.10.weight": [
[
0.058118,
0.50047,
0.187995,
0.024199,
0.101228,
0.285054,
-0.301016,
0.248881
],
[
-0.075758,
0.038293,
0.270912,
-0.246458,
0.066369,
-0.20687,
-0.098871,
0.305063
],
[
0.155889,
0.183556,
0.415545,
-0.092289,
0.029519,
0.494169,
0.057489,
0.179395
],
[
-0.263637,
0.318294,
0.093727,
-0.176092,
-0.171848,
0.210318,
-0.301772,
0.278717
],
[
0.107564,
-0.314612,
-0.246066,
0.16176,
-0.070373,
-0.020733,
0.140617,
-0.238411
],
[
0.31041,
0.210162,
-0.074565,
0.297651,
-0.165847,
-0.115569,
-0.307952,
0.107705
],
[
0.04185,
0.252412,
-0.349461,
0.101985,
-0.116916,
-0.306541,
0.116585,
-0.203029
],
[
0.039274,
0.189525,
0.302032,
0.039347,
0.034091,
-0.326047,
-0.004839,
-0.172949
]
],
"network.10.bias": [
-0.47024,
-0.072295,
-0.507004,
-0.319887,
-0.007861,
-0.210522,
-0.018141,
-0.275257
],
"network.12.weight": [
[
-0.156145,
-0.027236,
-0.34175,
-0.159891,
-0.066776,
-0.032313,
-0.333177,
0.337983
]
],
"network.12.bias": [
1.050614
]
}
## Activation Signature
### 0
mean: [-5.510777, -0.887659, 1.713092, 4.022634, 1.402226, -0.323824, -1.555949, 0.546418]
std: [4.038497, 4.491000, 2.984743, 3.349240, 3.342164, 3.263705, 3.383164, 2.581876]
### 2
mean: [8.834291, -1.374465, 4.821536, -4.801170, 6.865030, -1.314334, 0.602655, -1.805314]
std: [10.568254, 4.296184, 8.131716, 3.290321, 11.115867, 6.019885, 4.618095, 5.174637]
### 4
mean: [9.443420, -9.292645, -8.109442, -10.398354, -12.134007, 13.377235, -4.935707, -11.784537]
std: [16.673845, 9.290627, 8.354321, 12.531136, 10.917912, 22.247318, 4.370909, 16.943039]
### 6
mean: [-1.461689, 11.502235, -7.291756, -17.255289, 15.706025, -6.255388, -8.108604, -3.454838]
std: [1.246368, 15.942326, 8.715746, 21.560286, 20.919336, 7.604318, 10.271538, 4.538529]
### 8
mean: [-1.890017, -6.144286, -8.453947, -2.692043, -8.036296, 14.325708, -0.578310, -5.477706]
std: [2.217360, 7.004000, 10.470514, 3.107436, 10.513025, 18.938929, 0.342668, 6.428535]
### 10
mean: [3.626991, -3.045466, 6.595653, 2.702626, -0.306001, -1.871910, -4.424235, -4.961388]
std: [5.388229, 3.910578, 9.341251, 3.975917, 0.391816, 2.184350, 5.794378, 6.163319]
### 12
mean: [-2.313604]
std: [4.586017]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
first_last_match
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -5.510776996612549, "std": 4.038497447967529}, "1": {"mean": -0.8876585960388184, "std": 4.491000175476074}, "2": {"mean": 1.7130922079086304, "std": 2.9847428798675537}, "3": {"mean": 4.022634029388428, "std": 3.3492400646209717}, "4": {"mean": 1.4022263288497925, "std": 3.342163562774658}, "5": {"mean": -0.3238235116004944, "std": 3.263705253601074}, "6": {"mean": -1.5559488534927368, "std": 3.3831636905670166}, "7": {"mean": 0.5464178323745728, "std": 2.581876039505005}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 8.834290504455566, "std": 10.568253517150879}, "1": {"mean": -1.3744648694992065, "std": 4.2961835861206055}, "2": {"mean": 4.821536064147949, "std": 8.131715774536133}, "3": {"mean": -4.8011698722839355, "std": 3.290320873260498}, "4": {"mean": 6.865030288696289, "std": 11.115866661071777}, "5": {"mean": -1.3143342733383179, "std": 6.0198845863342285}, "6": {"mean": 0.602654755115509, "std": 4.618095397949219}, "7": {"mean": -1.8053141832351685, "std": 5.1746368408203125}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 9.44342041015625, "std": 16.673845291137695}, "1": {"mean": -9.292644500732422, "std": 9.290627479553223}, "2": {"mean": -8.109441757202148, "std": 8.354321479797363}, "3": {"mean": -10.398353576660156, "std": 12.531135559082031}, "4": {"mean": -12.134007453918457, "std": 10.917912483215332}, "5": {"mean": 13.377235412597656, "std": 22.247318267822266}, "6": {"mean": -4.935707092285156, "std": 4.370908737182617}, "7": {"mean": -11.784537315368652, "std": 16.943038940429688}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -1.46168851852417, "std": 1.2463680505752563}, "1": {"mean": 11.502235412597656, "std": 15.942325592041016}, "2": {"mean": -7.2917561531066895, "std": 8.71574592590332}, "3": {"mean": -17.25528907775879, "std": 21.560285568237305}, "4": {"mean": 15.706025123596191, "std": 20.919336318969727}, "5": {"mean": -6.255388259887695, "std": 7.604317665100098}, "6": {"mean": -8.108604431152344, "std": 10.271537780761719}, "7": {"mean": -3.4548380374908447, "std": 4.538528919219971}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -1.8900166749954224, "std": 2.217360258102417}, "1": {"mean": -6.144285678863525, "std": 7.003999710083008}, "2": {"mean": -8.453947067260742, "std": 10.470514297485352}, "3": {"mean": -2.692042827606201, "std": 3.107435941696167}, "4": {"mean": -8.036295890808105, "std": 10.513025283813477}, "5": {"mean": 14.325708389282227, "std": 18.938928604125977}, "6": {"mean": -0.5783103108406067, "std": 0.34266752004623413}, "7": {"mean": -5.477706432342529, "std": 6.428534984588623}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": 3.626990795135498, "std": 5.3882293701171875}, "1": {"mean": -3.045466184616089, "std": 3.9105782508850098}, "2": {"mean": 6.595653057098389, "std": 9.341251373291016}, "3": {"mean": 2.7026255130767822, "std": 3.975917339324951}, "4": {"mean": -0.30600112676620483, "std": 0.3918161690235138}, "5": {"mean": -1.8719104528427124, "std": 2.1843504905700684}, "6": {"mean": -4.4242353439331055, "std": 5.794378280639648}, "7": {"mean": -4.961388111114502, "std": 6.163319110870361}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -2.3136041164398193, "std": 4.586016654968262}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.490655, -0.890976, -0.993846, 0.00125, -0.954278], [-1.377325, -0.498339, -0.00606, -0.292589, 2.000451], [1.241971, -0.177386, -0.29473, -0.283594, 0.795638], [0.763351, 0.369734, -0.004319, 0.08309, 1.337082], [-0.970683, 0.388461, 0.903079, 0.663176, -1.127279], [-1.271977, 0.110805, -0.061307, -0.009472, 1.399718], [-0.656826, -0.425906, -0.659909, 0.029156, 1.541946], [-0.121218, 0.07403, -0.046074, -0.420979, 1.429253]], "network.0.bias": [-0.02633, -0.092693, 0.73329, 0.580722, -0.01594, -0.520773, -0.548249, -0.201367], "network.2.weight": [[0.272346, 0.791525, 1.174763, 0.895801, -0.387117, 0.942643, 0.965652, 0.424796], [0.356493, -0.830703, -0.429065, 0.059607, 0.33109, -0.98839, 0.090925, 0.129683], [0.203864, 0.838221, 0.792427, 0.658546, -0.641463, 1.077829, 0.468809, -0.138143], [-0.055045, -0.296334, -0.617526, -0.183018, -0.401407, -0.226686, -0.215169, -0.390271], [-0.318928, 1.379205, 1.038775, 0.938613, -0.911685, 1.296718, 0.007716, 0.120276], [0.137758, -0.432547, -0.64261, -0.200839, 0.946056, -1.125124, -0.415886, 0.305264], [0.657575, -1.240146, 0.183847, 0.339754, 0.434559, -1.067589, 0.119712, 0.25117], [0.811236, -1.019395, -0.39028, -0.217465, 0.715895, -0.672729, 0.462858, -0.050271]], "network.2.bias": [0.592355, 0.280901, -0.293282, -0.750964, -0.181286, 0.319355, 0.046279, -0.01505], "network.4.weight": [[0.745928, -0.328661, 0.224579, 0.326194, 0.525038, -0.95037, -0.27819, -0.160564], [-0.194589, -0.10026, -0.426141, 0.002479, -0.459962, -0.529323, -0.297988, 0.03369], [0.015551, -0.246094, -0.055173, 0.085254, -0.819129, -0.1481, -0.575338, 0.061771], [-0.564345, 0.784598, -0.549108, -0.317714, -0.258917, -0.867923, 0.172177, 0.586397], [-0.299149, -0.00306, -0.218182, -0.006585, -0.699038, -0.722472, -0.311629, -0.288952], [0.93402, -0.519111, 0.493004, 0.232543, 0.66777, -0.668561, -0.539197, -0.566699], [-0.17662, -0.369551, -0.217559, -0.1051, -0.114685, -0.208908, -0.179136, 0.203311], [-1.023039, 0.164955, -0.729578, -0.23867, -0.041209, -0.412591, 0.585724, 0.737531]], "network.4.bias": [-0.03677, -0.044589, 0.021343, -0.371403, -0.603666, 0.235664, -0.495734, 0.307357], "network.6.weight": [[-0.13934, -0.308205, -0.092561, -0.277093, 0.286648, 0.041159, -0.547652, 0.01178], [0.339022, 0.005918, -0.028751, -0.179008, 0.143681, 0.528794, -0.109844, -0.284687], [-0.134226, -0.119081, -0.190176, 0.047651, -0.033195, -0.335464, -0.038213, -0.168899], [-0.171875, 0.519424, -0.526412, -0.771659, 0.020968, -0.938674, 0.053015, -0.093101], [0.355845, 0.065637, -0.525138, -0.11884, 0.289571, 0.760917, 0.183523, -0.312251], [-0.069802, -0.01271, 0.6764, 0.004161, -0.51545, -0.335986, 0.044062, -0.492812], [-0.183597, -0.003792, 0.774058, -0.25863, -0.559416, -0.372914, 0.433465, -0.056768], [-0.256778, 0.020156, -0.174051, -0.236685, 0.144126, -0.030602, -0.275677, 0.162399]], "network.6.bias": [-0.516136, -0.438314, -0.391068, -0.554872, -0.033515, 0.148889, -0.137986, -0.151041], "network.8.weight": [[-0.057228, 0.332229, 0.073358, 0.316957, -0.357482, -0.101428, -0.413569, 0.126131], [0.023226, -0.225303, -0.316204, -0.166543, -0.168512, -0.691871, -0.64658, -0.255912], [-0.206992, 0.015238, 0.337925, -0.294443, -0.517188, -0.41446, -0.375646, -0.215216], [0.228357, 0.171525, -0.083359, -0.094571, -0.279522, -0.215949, 0.047006, -0.024072], [-0.120499, -0.315261, -0.158613, 0.24782, -0.269798, -0.572565, -0.593069, -0.258485], [0.349892, 0.470166, 0.108741, -0.022046, 0.559328, 0.109815, 0.232849, 0.35022], [-0.218153, 0.025524, 0.176049, -0.278457, -0.034801, -0.051801, -0.155511, 0.326797], [0.195893, -0.212903, -0.081911, -0.153246, -0.149833, -0.77111, -0.759632, -0.138665]], "network.8.bias": [-0.1418, -0.760358, -0.370544, -0.267924, 0.029508, -0.198083, -0.349462, -0.546599], "network.10.weight": [[0.058118, 0.50047, 0.187995, 0.024199, 0.101228, 0.285054, -0.301016, 0.248881], [-0.075758, 0.038293, 0.270912, -0.246458, 0.066369, -0.20687, -0.098871, 0.305063], [0.155889, 0.183556, 0.415545, -0.092289, 0.029519, 0.494169, 0.057489, 0.179395], [-0.263637, 0.318294, 0.093727, -0.176092, -0.171848, 0.210318, -0.301772, 0.278717], [0.107564, -0.314612, -0.246066, 0.16176, -0.070373, -0.020733, 0.140617, -0.238411], [0.31041, 0.210162, -0.074565, 0.297651, -0.165847, -0.115569, -0.307952, 0.107705], [0.04185, 0.252412, -0.349461, 0.101985, -0.116916, -0.306541, 0.116585, -0.203029], [0.039274, 0.189525, 0.302032, 0.039347, 0.034091, -0.326047, -0.004839, -0.172949]], "network.10.bias": [-0.47024, -0.072295, -0.507004, -0.319887, -0.007861, -0.210522, -0.018141, -0.275257], "network.12.weight": [[-0.156145, -0.027236, -0.34175, -0.159891, -0.066776, -0.032313, -0.333177, 0.337983]], "network.12.bias": [1.050614]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6800411641597748, "train_acc": 0.58, "val_loss": 0.7504766583442688, "val_acc": 0.5}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6803210973739624, "train_acc": 0.58, "val_loss": 0.6691428422927856, "val_acc": 0.5}, {"stage": "improved", "epoch": 0, "global_epoch": 2, "train_loss": 0.7458194494247437, "train_acc": 0.5, "val_loss": 0.627545177936554, "val_acc": 0.5}, {"stage": "improved", "epoch": 1, "global_epoch": 3, "train_loss": 0.6436228156089783, "train_acc": 0.605, "val_loss": 0.6283458471298218, "val_acc": 0.8}, {"stage": "improved", "epoch": 2, "global_epoch": 4, "train_loss": 0.6342594921588898, "train_acc": 0.73, "val_loss": 0.5168307423591614, "val_acc": 0.86}, {"stage": "improved", "epoch": 3, "global_epoch": 5, "train_loss": 0.5249567329883575, "train_acc": 0.765, "val_loss": 0.4763392508029938, "val_acc": 0.78}, {"stage": "improved", "epoch": 4, "global_epoch": 6, "train_loss": 0.49277573823928833, "train_acc": 0.72, "val_loss": 0.3853885531425476, "val_acc": 0.86}, {"stage": "improved", "epoch": 5, "global_epoch": 7, "train_loss": 0.4511253386735916, "train_acc": 0.805, "val_loss": 0.35507720708847046, "val_acc": 0.88}, {"stage": "improved", "epoch": 6, "global_epoch": 8, "train_loss": 0.4559151381254196, "train_acc": 0.805, "val_loss": 0.35916221141815186, "val_acc": 0.86}, {"stage": "improved", "epoch": 7, "global_epoch": 9, "train_loss": 0.4181377440690994, "train_acc": 0.825, "val_loss": 0.371183842420578, "val_acc": 0.84}, {"stage": "improved", "epoch": 8, "global_epoch": 10, "train_loss": 0.4113367348909378, "train_acc": 0.825, "val_loss": 0.3322274088859558, "val_acc": 0.86}, {"stage": "improved", "epoch": 9, "global_epoch": 11, "train_loss": 0.38331669569015503, "train_acc": 0.82, "val_loss": 0.31231018900871277, "val_acc": 0.9}], "summary": {"total_epochs": 12, "degraded_epochs": 2, "improved_epochs": 10, "patterns": ["first_last_match"], "degraded_stage": {"initial_val_loss": 0.7504766583442688, "final_val_loss": 0.6691428422927856, "initial_val_acc": 0.5, "final_val_acc": 0.5, "best_val_acc": 0.5}, "improved_stage": {"initial_val_loss": 0.627545177936554, "final_val_loss": 0.31231018900871277, "initial_val_acc": 0.5, "final_val_acc": 0.9, "best_val_acc": 0.9, "best_epoch": 11}, "improvement": 0.4, "first_improvement_epoch": 1}}
|
97
|
{"target_pattern": "mountain_pattern", "degraded_accuracy": 0.74, "improved_accuracy": 0.9, "improvement": 0.16000000000000003, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 1800, "learning_rate": 0.05216234752909743, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "mountain_pattern", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["mountain_pattern"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.232168,
-0.457808,
-0.280405,
-0.481828,
-0.002451
],
[
0.060573,
0.429139,
-0.230159,
-0.352905,
0.618685
],
[
-0.074297,
-0.103148,
0.331318,
-0.373565,
-0.659595
],
[
0.024161,
0.213467,
-0.030597,
0.149801,
-0.002235
],
[
-0.860779,
-0.463988,
0.300613,
0.262534,
0.198424
],
[
0.491555,
0.274348,
0.182218,
-0.133722,
-0.471344
],
[
-0.558367,
-0.513789,
0.348562,
0.456356,
-0.072824
],
[
-0.166395,
-0.126346,
-0.107657,
-0.011788,
-0.385882
]
],
"network.0.bias": [
-0.087505,
0.396271,
-0.228748,
0.077138,
-0.065556,
0.371544,
0.134679,
-0.466456
],
"network.2.weight": [
[
-0.501129,
-0.380773,
-0.356822,
0.16942,
0.366045,
-0.126892,
-0.110881,
-0.140962
],
[
-0.362431,
-0.161243,
-0.005684,
-0.042465,
0.085593,
-0.613036,
0.036659,
-0.42772
],
[
0.017244,
0.607772,
0.422056,
0.008564,
-0.093609,
0.477748,
-0.182654,
-0.128914
],
[
-0.1228,
0.589525,
0.127088,
0.069348,
0.068408,
0.411641,
-0.159621,
-0.098684
],
[
0.010792,
-0.63752,
-0.080755,
-0.071524,
0.502033,
-0.450047,
0.375881,
-0.34258
],
[
0.179439,
-0.202378,
0.09655,
0.429951,
-0.018601,
-0.143803,
0.499741,
-0.055811
],
[
0.018329,
-0.558424,
-0.274974,
0.135602,
0.302946,
0.066948,
0.338539,
-0.020897
],
[
-0.194554,
0.465546,
0.415048,
0.195594,
-0.102314,
0.396327,
-0.034273,
0.224594
]
],
"network.2.bias": [
-0.00505,
-0.381801,
0.334092,
0.088379,
0.273449,
-0.225137,
0.212696,
-0.056933
],
"network.4.weight": [
[
-0.044555,
0.167994,
0.288399,
0.623029,
0.258235,
-0.091892,
-0.089877,
0.374359
],
[
-0.08895,
0.0961,
-0.387363,
0.196377,
-0.226654,
-0.067447,
-0.262373,
-0.331556
],
[
0.246088,
0.261385,
0.230815,
-0.039275,
-0.058734,
0.265489,
-0.252403,
-0.152651
],
[
-0.006296,
0.173451,
-0.439991,
-0.309583,
0.147835,
-0.420492,
-0.275879,
-0.443087
],
[
0.288184,
0.490157,
-0.348548,
-0.430083,
0.757279,
0.117377,
0.055982,
0.090617
],
[
-0.513875,
-0.293551,
-0.22013,
-0.200908,
-0.283208,
-0.061692,
-0.348914,
0.102982
],
[
-0.348281,
-0.252455,
-0.297671,
-0.141268,
-0.172123,
0.149491,
-0.325091,
-0.035532
],
[
0.171893,
-0.032887,
-0.499487,
0.020117,
0.580262,
0.330822,
0.307405,
-0.333322
]
],
"network.4.bias": [
0.173888,
-0.262272,
0.029473,
-0.149972,
-0.186294,
-0.100894,
0.097094,
0.032292
],
"network.6.weight": [
[
0.583925,
-0.312616,
-0.123568,
0.076268,
-0.372044,
0.013085,
-0.105882,
0.111006
],
[
0.105193,
-0.252178,
0.246538,
0.217602,
0.156006,
0.170479,
-0.032068,
-0.504908
],
[
0.184428,
0.042559,
0.246121,
0.056763,
-0.347629,
0.09307,
-0.088646,
-0.401668
],
[
-0.140266,
0.124393,
0.156089,
-0.120582,
-0.172312,
-0.008778,
0.280499,
-0.383529
],
[
0.202962,
-0.087329,
-0.345014,
-0.306746,
-0.231765,
-0.233255,
-0.072074,
-0.021197
],
[
-0.369559,
0.127496,
-0.046786,
0.245154,
0.435765,
-0.212213,
-0.343611,
0.397122
],
[
0.451927,
-0.060379,
0.34483,
-0.115135,
0.115872,
0.028451,
0.207731,
-0.140036
],
[
-0.185303,
-0.158601,
0.365816,
0.472668,
0.643267,
-0.305244,
0.107736,
0.459942
]
],
"network.6.bias": [
0.472415,
-0.076355,
0.476815,
-0.419598,
-0.359535,
-0.019382,
0.459912,
0.094434
],
"network.8.weight": [
[
-0.57724,
-0.013714,
-0.450147,
-0.155969,
-0.278312,
0.183125,
0.649612,
0.587342
],
[
0.267763,
0.066715,
0.128077,
0.122477,
-0.001394,
-0.227432,
0.115056,
-0.331227
],
[
-0.008183,
-0.064758,
-0.275724,
-0.30354,
-0.060895,
-0.075856,
0.115708,
-0.210594
],
[
-0.414905,
-0.233181,
-0.28818,
0.331551,
0.268331,
0.492443,
0.144254,
-0.061092
],
[
0.292589,
-0.148892,
0.015827,
0.226777,
-0.115034,
-0.292434,
0.500181,
-0.023063
],
[
0.768494,
0.161172,
0.442371,
-0.168568,
-0.164716,
-0.448003,
0.436432,
-0.202253
],
[
0.186315,
-0.141913,
-0.400233,
0.139541,
-0.195299,
-0.306007,
0.133801,
-0.330285
],
[
-0.500981,
-0.377473,
-0.451758,
-0.063529,
0.218094,
-0.075334,
-0.271378,
-0.097139
]
],
"network.8.bias": [
-0.078216,
0.313408,
-0.233143,
0.236338,
-0.176764,
0.308077,
-0.456842,
0.04898
],
"network.10.weight": [
[
-0.168374,
0.32727,
0.03298,
-0.14705,
0.206522,
0.502353,
0.051452,
-0.248404
],
[
-0.209321,
-0.231598,
-0.169702,
0.018301,
-0.123024,
0.023778,
0.145841,
-0.056172
],
[
-0.125817,
-0.126093,
0.309581,
-0.267305,
-0.01685,
-0.043509,
0.085954,
-0.216833
],
[
-0.255173,
0.087546,
-0.173873,
-0.105776,
0.413302,
0.31803,
-0.042379,
0.054304
],
[
0.438592,
-0.301469,
-0.178597,
0.276355,
-0.222997,
-0.108886,
0.212245,
0.360228
],
[
-0.004746,
0.08503,
0.338983,
-0.001971,
-0.227546,
-0.05149,
-0.273691,
0.041043
],
[
0.643344,
-0.238497,
0.244541,
0.171781,
0.240674,
-0.361582,
0.164466,
0.298172
],
[
-0.165019,
0.33851,
0.258429,
-0.403426,
-0.12569,
0.324485,
-0.18553,
0.054324
]
],
"network.10.bias": [
0.629239,
-0.218924,
-0.473566,
0.045018,
0.287632,
-0.408422,
0.25637,
-0.252447
],
"network.12.weight": [
[
-0.430844,
-0.310048,
0.05796,
-0.223245,
0.500462,
0.049009,
0.300234,
-0.21842
]
],
"network.12.bias": [
-0.298697
]
}
## Activation Signature
### 0
mean: [-2.832663, 0.779121, -1.431669, 0.750759, -0.546578, 0.995732, 0.120506, -1.629837]
std: [1.846026, 1.330947, 1.598892, 0.582350, 2.097402, 1.551782, 1.735941, 0.998222]
### 2
mean: [-0.347768, -1.237147, 1.376190, 1.149923, -0.445723, 0.092727, 0.204446, 0.990950]
std: [0.455701, 0.871980, 1.059078, 0.940863, 1.389959, 0.690996, 0.852307, 0.829949]
### 4
mean: [1.685828, -1.125112, 0.120959, -1.768147, -0.730015, -0.835876, -0.686039, -0.509621]
std: [1.162443, 0.401249, 0.116794, 0.966807, 1.122759, 0.352118, 0.372545, 1.312048]
### 6
mean: [1.399637, -0.005862, 0.609154, -0.802968, -0.118599, -0.420765, 1.243479, 0.122709]
std: [0.714269, 0.359901, 0.576222, 0.288664, 0.278540, 0.767144, 0.580090, 0.691287]
### 8
mean: [-0.208333, 0.795371, -0.375296, -0.312192, 0.782909, 2.108541, -0.481194, -1.376924]
std: [0.626750, 0.535733, 0.102544, 0.444307, 0.550645, 1.173295, 0.298153, 0.668628]
### 10
mean: [2.092260, -0.497386, -0.729486, 1.069496, -0.271421, -0.630154, -0.386026, 0.555828]
std: [0.897946, 0.100113, 0.084800, 0.708211, 0.566211, 0.137880, 0.655329, 0.573636]
### 12
mean: [-1.471412]
std: [0.838457]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
mountain_pattern
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 8
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.232168,
-0.457808,
-0.280405,
-0.481828,
-0.002451
],
[
0.060573,
0.429139,
-0.230159,
-0.352905,
0.618685
],
[
-0.074297,
-0.103148,
0.331318,
-0.373565,
-0.659595
],
[
0.024161,
0.213467,
-0.030597,
0.149801,
-0.002235
],
[
-0.860779,
-0.463988,
0.300613,
0.262534,
0.198424
],
[
0.491555,
0.274348,
0.182218,
-0.133722,
-0.471344
],
[
-0.558367,
-0.513789,
0.348562,
0.456356,
-0.072824
],
[
-0.166395,
-0.126346,
-0.107657,
-0.011788,
-0.385882
]
],
"network.0.bias": [
-0.087505,
0.396271,
-0.228748,
0.077138,
-0.065556,
0.371544,
0.134679,
-0.466456
],
"network.2.weight": [
[
-0.501129,
-0.380773,
-0.356822,
0.16942,
0.366045,
-0.126892,
-0.110881,
-0.140962
],
[
-0.362431,
-0.161243,
-0.005684,
-0.042465,
0.085593,
-0.613036,
0.036659,
-0.42772
],
[
0.017244,
0.607772,
0.422056,
0.008564,
-0.093609,
0.477748,
-0.182654,
-0.128914
],
[
-0.1228,
0.589525,
0.127088,
0.069348,
0.068408,
0.411641,
-0.159621,
-0.098684
],
[
0.010792,
-0.63752,
-0.080755,
-0.071524,
0.502033,
-0.450047,
0.375881,
-0.34258
],
[
0.179439,
-0.202378,
0.09655,
0.429951,
-0.018601,
-0.143803,
0.499741,
-0.055811
],
[
0.018329,
-0.558424,
-0.274974,
0.135602,
0.302946,
0.066948,
0.338539,
-0.020897
],
[
-0.194554,
0.465546,
0.415048,
0.195594,
-0.102314,
0.396327,
-0.034273,
0.224594
]
],
"network.2.bias": [
-0.00505,
-0.381801,
0.334092,
0.088379,
0.273449,
-0.225137,
0.212696,
-0.056933
],
"network.4.weight": [
[
-0.044555,
0.167994,
0.288399,
0.623029,
0.258235,
-0.091892,
-0.089877,
0.374359
],
[
-0.08895,
0.0961,
-0.387363,
0.196377,
-0.226654,
-0.067447,
-0.262373,
-0.331556
],
[
0.246088,
0.261385,
0.230815,
-0.039275,
-0.058734,
0.265489,
-0.252403,
-0.152651
],
[
-0.006296,
0.173451,
-0.439991,
-0.309583,
0.147835,
-0.420492,
-0.275879,
-0.443087
],
[
0.288184,
0.490157,
-0.348548,
-0.430083,
0.757279,
0.117377,
0.055982,
0.090617
],
[
-0.513875,
-0.293551,
-0.22013,
-0.200908,
-0.283208,
-0.061692,
-0.348914,
0.102982
],
[
-0.348281,
-0.252455,
-0.297671,
-0.141268,
-0.172123,
0.149491,
-0.325091,
-0.035532
],
[
0.171893,
-0.032887,
-0.499487,
0.020117,
0.580262,
0.330822,
0.307405,
-0.333322
]
],
"network.4.bias": [
0.173888,
-0.262272,
0.029473,
-0.149972,
-0.186294,
-0.100894,
0.097094,
0.032292
],
"network.6.weight": [
[
0.583925,
-0.312616,
-0.123568,
0.076268,
-0.372044,
0.013085,
-0.105882,
0.111006
],
[
0.105193,
-0.252178,
0.246538,
0.217602,
0.156006,
0.170479,
-0.032068,
-0.504908
],
[
0.184428,
0.042559,
0.246121,
0.056763,
-0.347629,
0.09307,
-0.088646,
-0.401668
],
[
-0.140266,
0.124393,
0.156089,
-0.120582,
-0.172312,
-0.008778,
0.280499,
-0.383529
],
[
0.202962,
-0.087329,
-0.345014,
-0.306746,
-0.231765,
-0.233255,
-0.072074,
-0.021197
],
[
-0.369559,
0.127496,
-0.046786,
0.245154,
0.435765,
-0.212213,
-0.343611,
0.397122
],
[
0.451927,
-0.060379,
0.34483,
-0.115135,
0.115872,
0.028451,
0.207731,
-0.140036
],
[
-0.185303,
-0.158601,
0.365816,
0.472668,
0.643267,
-0.305244,
0.107736,
0.459942
]
],
"network.6.bias": [
0.472415,
-0.076355,
0.476815,
-0.419598,
-0.359535,
-0.019382,
0.459912,
0.094434
],
"network.8.weight": [
[
-0.57724,
-0.013714,
-0.450147,
-0.155969,
-0.278312,
0.183125,
0.649612,
0.587342
],
[
0.267763,
0.066715,
0.128077,
0.122477,
-0.001394,
-0.227432,
0.115056,
-0.331227
],
[
-0.008183,
-0.064758,
-0.275724,
-0.30354,
-0.060895,
-0.075856,
0.115708,
-0.210594
],
[
-0.414905,
-0.233181,
-0.28818,
0.331551,
0.268331,
0.492443,
0.144254,
-0.061092
],
[
0.292589,
-0.148892,
0.015827,
0.226777,
-0.115034,
-0.292434,
0.500181,
-0.023063
],
[
0.768494,
0.161172,
0.442371,
-0.168568,
-0.164716,
-0.448003,
0.436432,
-0.202253
],
[
0.186315,
-0.141913,
-0.400233,
0.139541,
-0.195299,
-0.306007,
0.133801,
-0.330285
],
[
-0.500981,
-0.377473,
-0.451758,
-0.063529,
0.218094,
-0.075334,
-0.271378,
-0.097139
]
],
"network.8.bias": [
-0.078216,
0.313408,
-0.233143,
0.236338,
-0.176764,
0.308077,
-0.456842,
0.04898
],
"network.10.weight": [
[
-0.168374,
0.32727,
0.03298,
-0.14705,
0.206522,
0.502353,
0.051452,
-0.248404
],
[
-0.209321,
-0.231598,
-0.169702,
0.018301,
-0.123024,
0.023778,
0.145841,
-0.056172
],
[
-0.125817,
-0.126093,
0.309581,
-0.267305,
-0.01685,
-0.043509,
0.085954,
-0.216833
],
[
-0.255173,
0.087546,
-0.173873,
-0.105776,
0.413302,
0.31803,
-0.042379,
0.054304
],
[
0.438592,
-0.301469,
-0.178597,
0.276355,
-0.222997,
-0.108886,
0.212245,
0.360228
],
[
-0.004746,
0.08503,
0.338983,
-0.001971,
-0.227546,
-0.05149,
-0.273691,
0.041043
],
[
0.643344,
-0.238497,
0.244541,
0.171781,
0.240674,
-0.361582,
0.164466,
0.298172
],
[
-0.165019,
0.33851,
0.258429,
-0.403426,
-0.12569,
0.324485,
-0.18553,
0.054324
]
],
"network.10.bias": [
0.629239,
-0.218924,
-0.473566,
0.045018,
0.287632,
-0.408422,
0.25637,
-0.252447
],
"network.12.weight": [
[
-0.430844,
-0.310048,
0.05796,
-0.223245,
0.500462,
0.049009,
0.300234,
-0.21842
]
],
"network.12.bias": [
-0.298697
]
}
## Activation Signature
### 0
mean: [-2.832663, 0.779121, -1.431669, 0.750759, -0.546578, 0.995732, 0.120506, -1.629837]
std: [1.846026, 1.330947, 1.598892, 0.582350, 2.097402, 1.551782, 1.735941, 0.998222]
### 2
mean: [-0.347768, -1.237147, 1.376190, 1.149923, -0.445723, 0.092727, 0.204446, 0.990950]
std: [0.455701, 0.871980, 1.059078, 0.940863, 1.389959, 0.690996, 0.852307, 0.829949]
### 4
mean: [1.685828, -1.125112, 0.120959, -1.768147, -0.730015, -0.835876, -0.686039, -0.509621]
std: [1.162443, 0.401249, 0.116794, 0.966807, 1.122759, 0.352118, 0.372545, 1.312048]
### 6
mean: [1.399637, -0.005862, 0.609154, -0.802968, -0.118599, -0.420765, 1.243479, 0.122709]
std: [0.714269, 0.359901, 0.576222, 0.288664, 0.278540, 0.767144, 0.580090, 0.691287]
### 8
mean: [-0.208333, 0.795371, -0.375296, -0.312192, 0.782909, 2.108541, -0.481194, -1.376924]
std: [0.626750, 0.535733, 0.102544, 0.444307, 0.550645, 1.173295, 0.298153, 0.668628]
### 10
mean: [2.092260, -0.497386, -0.729486, 1.069496, -0.271421, -0.630154, -0.386026, 0.555828]
std: [0.897946, 0.100113, 0.084800, 0.708211, 0.566211, 0.137880, 0.655329, 0.573636]
### 12
mean: [-1.471412]
std: [0.838457]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
mountain_pattern
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -2.832663059234619, "std": 1.846025824546814}, "1": {"mean": 0.7791208028793335, "std": 1.3309473991394043}, "2": {"mean": -1.4316686391830444, "std": 1.598892331123352}, "3": {"mean": 0.750759482383728, "std": 0.5823498368263245}, "4": {"mean": -0.5465776324272156, "std": 2.097402334213257}, "5": {"mean": 0.9957319498062134, "std": 1.5517817735671997}, "6": {"mean": 0.12050573527812958, "std": 1.7359405755996704}, "7": {"mean": -1.629837155342102, "std": 0.9982216954231262}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.3477684259414673, "std": 0.45570141077041626}, "1": {"mean": -1.2371472120285034, "std": 0.8719803094863892}, "2": {"mean": 1.3761903047561646, "std": 1.0590782165527344}, "3": {"mean": 1.1499227285385132, "std": 0.9408633708953857}, "4": {"mean": -0.44572311639785767, "std": 1.3899585008621216}, "5": {"mean": 0.09272728115320206, "std": 0.6909958124160767}, "6": {"mean": 0.20444640517234802, "std": 0.8523067235946655}, "7": {"mean": 0.9909496903419495, "std": 0.8299486637115479}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 1.6858279705047607, "std": 1.1624433994293213}, "1": {"mean": -1.125111699104309, "std": 0.401249498128891}, "2": {"mean": 0.12095903605222702, "std": 0.11679358780384064}, "3": {"mean": -1.7681467533111572, "std": 0.9668065309524536}, "4": {"mean": -0.7300150394439697, "std": 1.1227588653564453}, "5": {"mean": -0.8358758687973022, "std": 0.3521178066730499}, "6": {"mean": -0.6860387325286865, "std": 0.3725448250770569}, "7": {"mean": -0.5096209645271301, "std": 1.3120479583740234}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": 1.3996367454528809, "std": 0.7142693400382996}, "1": {"mean": -0.005861729849129915, "std": 0.3599010705947876}, "2": {"mean": 0.6091535091400146, "std": 0.5762220025062561}, "3": {"mean": -0.8029680252075195, "std": 0.28866422176361084}, "4": {"mean": -0.11859850585460663, "std": 0.2785399556159973}, "5": {"mean": -0.4207654595375061, "std": 0.7671439051628113}, "6": {"mean": 1.2434793710708618, "std": 0.5800901055335999}, "7": {"mean": 0.1227092444896698, "std": 0.6912867426872253}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -0.20833292603492737, "std": 0.6267502307891846}, "1": {"mean": 0.7953712344169617, "std": 0.535732626914978}, "2": {"mean": -0.3752959072589874, "std": 0.10254424065351486}, "3": {"mean": -0.31219199299812317, "std": 0.4443073272705078}, "4": {"mean": 0.7829090356826782, "std": 0.55064457654953}, "5": {"mean": 2.108541250228882, "std": 1.173295021057129}, "6": {"mean": -0.48119381070137024, "std": 0.29815295338630676}, "7": {"mean": -1.3769237995147705, "std": 0.6686283946037292}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": 2.0922601222991943, "std": 0.8979458808898926}, "1": {"mean": -0.49738556146621704, "std": 0.10011272132396698}, "2": {"mean": -0.7294864058494568, "std": 0.08479979634284973}, "3": {"mean": 1.0694962739944458, "std": 0.7082110643386841}, "4": {"mean": -0.2714214324951172, "std": 0.5662112832069397}, "5": {"mean": -0.6301537752151489, "std": 0.1378796398639679}, "6": {"mean": -0.3860258460044861, "std": 0.6553292870521545}, "7": {"mean": 0.555827796459198, "std": 0.5736360549926758}}, "layer_info": {"num_neurons": 8, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -1.4714115858078003, "std": 0.8384571671485901}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 8, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.232168, -0.457808, -0.280405, -0.481828, -0.002451], [0.060573, 0.429139, -0.230159, -0.352905, 0.618685], [-0.074297, -0.103148, 0.331318, -0.373565, -0.659595], [0.024161, 0.213467, -0.030597, 0.149801, -0.002235], [-0.860779, -0.463988, 0.300613, 0.262534, 0.198424], [0.491555, 0.274348, 0.182218, -0.133722, -0.471344], [-0.558367, -0.513789, 0.348562, 0.456356, -0.072824], [-0.166395, -0.126346, -0.107657, -0.011788, -0.385882]], "network.0.bias": [-0.087505, 0.396271, -0.228748, 0.077138, -0.065556, 0.371544, 0.134679, -0.466456], "network.2.weight": [[-0.501129, -0.380773, -0.356822, 0.16942, 0.366045, -0.126892, -0.110881, -0.140962], [-0.362431, -0.161243, -0.005684, -0.042465, 0.085593, -0.613036, 0.036659, -0.42772], [0.017244, 0.607772, 0.422056, 0.008564, -0.093609, 0.477748, -0.182654, -0.128914], [-0.1228, 0.589525, 0.127088, 0.069348, 0.068408, 0.411641, -0.159621, -0.098684], [0.010792, -0.63752, -0.080755, -0.071524, 0.502033, -0.450047, 0.375881, -0.34258], [0.179439, -0.202378, 0.09655, 0.429951, -0.018601, -0.143803, 0.499741, -0.055811], [0.018329, -0.558424, -0.274974, 0.135602, 0.302946, 0.066948, 0.338539, -0.020897], [-0.194554, 0.465546, 0.415048, 0.195594, -0.102314, 0.396327, -0.034273, 0.224594]], "network.2.bias": [-0.00505, -0.381801, 0.334092, 0.088379, 0.273449, -0.225137, 0.212696, -0.056933], "network.4.weight": [[-0.044555, 0.167994, 0.288399, 0.623029, 0.258235, -0.091892, -0.089877, 0.374359], [-0.08895, 0.0961, -0.387363, 0.196377, -0.226654, -0.067447, -0.262373, -0.331556], [0.246088, 0.261385, 0.230815, -0.039275, -0.058734, 0.265489, -0.252403, -0.152651], [-0.006296, 0.173451, -0.439991, -0.309583, 0.147835, -0.420492, -0.275879, -0.443087], [0.288184, 0.490157, -0.348548, -0.430083, 0.757279, 0.117377, 0.055982, 0.090617], [-0.513875, -0.293551, -0.22013, -0.200908, -0.283208, -0.061692, -0.348914, 0.102982], [-0.348281, -0.252455, -0.297671, -0.141268, -0.172123, 0.149491, -0.325091, -0.035532], [0.171893, -0.032887, -0.499487, 0.020117, 0.580262, 0.330822, 0.307405, -0.333322]], "network.4.bias": [0.173888, -0.262272, 0.029473, -0.149972, -0.186294, -0.100894, 0.097094, 0.032292], "network.6.weight": [[0.583925, -0.312616, -0.123568, 0.076268, -0.372044, 0.013085, -0.105882, 0.111006], [0.105193, -0.252178, 0.246538, 0.217602, 0.156006, 0.170479, -0.032068, -0.504908], [0.184428, 0.042559, 0.246121, 0.056763, -0.347629, 0.09307, -0.088646, -0.401668], [-0.140266, 0.124393, 0.156089, -0.120582, -0.172312, -0.008778, 0.280499, -0.383529], [0.202962, -0.087329, -0.345014, -0.306746, -0.231765, -0.233255, -0.072074, -0.021197], [-0.369559, 0.127496, -0.046786, 0.245154, 0.435765, -0.212213, -0.343611, 0.397122], [0.451927, -0.060379, 0.34483, -0.115135, 0.115872, 0.028451, 0.207731, -0.140036], [-0.185303, -0.158601, 0.365816, 0.472668, 0.643267, -0.305244, 0.107736, 0.459942]], "network.6.bias": [0.472415, -0.076355, 0.476815, -0.419598, -0.359535, -0.019382, 0.459912, 0.094434], "network.8.weight": [[-0.57724, -0.013714, -0.450147, -0.155969, -0.278312, 0.183125, 0.649612, 0.587342], [0.267763, 0.066715, 0.128077, 0.122477, -0.001394, -0.227432, 0.115056, -0.331227], [-0.008183, -0.064758, -0.275724, -0.30354, -0.060895, -0.075856, 0.115708, -0.210594], [-0.414905, -0.233181, -0.28818, 0.331551, 0.268331, 0.492443, 0.144254, -0.061092], [0.292589, -0.148892, 0.015827, 0.226777, -0.115034, -0.292434, 0.500181, -0.023063], [0.768494, 0.161172, 0.442371, -0.168568, -0.164716, -0.448003, 0.436432, -0.202253], [0.186315, -0.141913, -0.400233, 0.139541, -0.195299, -0.306007, 0.133801, -0.330285], [-0.500981, -0.377473, -0.451758, -0.063529, 0.218094, -0.075334, -0.271378, -0.097139]], "network.8.bias": [-0.078216, 0.313408, -0.233143, 0.236338, -0.176764, 0.308077, -0.456842, 0.04898], "network.10.weight": [[-0.168374, 0.32727, 0.03298, -0.14705, 0.206522, 0.502353, 0.051452, -0.248404], [-0.209321, -0.231598, -0.169702, 0.018301, -0.123024, 0.023778, 0.145841, -0.056172], [-0.125817, -0.126093, 0.309581, -0.267305, -0.01685, -0.043509, 0.085954, -0.216833], [-0.255173, 0.087546, -0.173873, -0.105776, 0.413302, 0.31803, -0.042379, 0.054304], [0.438592, -0.301469, -0.178597, 0.276355, -0.222997, -0.108886, 0.212245, 0.360228], [-0.004746, 0.08503, 0.338983, -0.001971, -0.227546, -0.05149, -0.273691, 0.041043], [0.643344, -0.238497, 0.244541, 0.171781, 0.240674, -0.361582, 0.164466, 0.298172], [-0.165019, 0.33851, 0.258429, -0.403426, -0.12569, 0.324485, -0.18553, 0.054324]], "network.10.bias": [0.629239, -0.218924, -0.473566, 0.045018, 0.287632, -0.408422, 0.25637, -0.252447], "network.12.weight": [[-0.430844, -0.310048, 0.05796, -0.223245, 0.500462, 0.049009, 0.300234, -0.21842]], "network.12.bias": [-0.298697]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6864830851554871, "train_acc": 0.575, "val_loss": 0.6887338161468506, "val_acc": 0.5}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6627926528453827, "train_acc": 0.575, "val_loss": 0.6725373864173889, "val_acc": 0.5}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6064549684524536, "train_acc": 0.655, "val_loss": 0.5365843772888184, "val_acc": 0.74}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.6570932865142822, "train_acc": 0.725, "val_loss": 0.4882015287876129, "val_acc": 0.82}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.40028806030750275, "train_acc": 0.84, "val_loss": 0.48158103227615356, "val_acc": 0.78}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.35644717514514923, "train_acc": 0.865, "val_loss": 0.3772534132003784, "val_acc": 0.9}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.30123887956142426, "train_acc": 0.92, "val_loss": 0.3601030707359314, "val_acc": 0.9}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.23668602108955383, "train_acc": 0.91, "val_loss": 0.46079421043395996, "val_acc": 0.84}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.2318251058459282, "train_acc": 0.925, "val_loss": 0.5441784858703613, "val_acc": 0.84}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.19878922030329704, "train_acc": 0.925, "val_loss": 0.5878233909606934, "val_acc": 0.86}], "summary": {"total_epochs": 10, "degraded_epochs": 3, "improved_epochs": 7, "patterns": ["mountain_pattern"], "degraded_stage": {"initial_val_loss": 0.6887338161468506, "final_val_loss": 0.5365843772888184, "initial_val_acc": 0.5, "final_val_acc": 0.74, "best_val_acc": 0.74}, "improved_stage": {"initial_val_loss": 0.4882015287876129, "final_val_loss": 0.5878233909606934, "initial_val_acc": 0.82, "final_val_acc": 0.86, "best_val_acc": 0.9, "best_epoch": 5}, "improvement": 0.16000000000000003, "first_improvement_epoch": 2}}
|
98
|
{"target_pattern": "sorted_ascending", "degraded_accuracy": 0.4, "improved_accuracy": 0.94, "improvement": 0.5399999999999999, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "random_seed": 8537, "learning_rate": 0.040863597474926766, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "sorted_ascending", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["sorted_ascending"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.24148,
-0.284973,
-0.011506,
-0.010725,
0.095342
],
[
-0.09899,
-0.091032,
-0.038765,
-0.237914,
-0.507862
],
[
0.428849,
0.588488,
0.63552,
-0.018677,
-0.090337
],
[
-0.506187,
-0.666013,
0.009966,
0.059477,
0.443891
],
[
-0.022759,
0.493391,
-0.247428,
0.285835,
0.07825
]
],
"network.0.bias": [
0.014034,
0.067529,
0.445784,
0.699904,
-0.434864
],
"network.2.weight": [
[
-0.543075,
0.004573,
0.025236,
-0.436943,
-0.003683
],
[
0.210612,
-0.013424,
-0.405601,
0.546392,
-0.375305
],
[
-0.234293,
-0.017903,
0.512457,
-0.172301,
0.642827
],
[
0.295244,
0.1091,
-0.113118,
0.715187,
-0.545092
],
[
-0.64669,
-0.426579,
0.526911,
-0.463885,
0.324099
]
],
"network.2.bias": [
-0.372761,
0.297178,
-0.128329,
0.152097,
0.64475
],
"network.4.weight": [
[
-0.187956,
0.421441,
-0.675647,
0.755687,
-0.503619
],
[
-0.577794,
-0.397993,
0.343279,
-0.250847,
0.662829
],
[
-0.550974,
0.410272,
-0.159546,
-0.098856,
-0.064285
],
[
-0.019926,
0.045329,
-0.560249,
0.49754,
-0.705909
],
[
-0.622615,
-0.575827,
0.207838,
-0.17866,
0.370809
]
],
"network.4.bias": [
0.563241,
0.595993,
-0.154638,
0.119405,
0.312103
],
"network.6.weight": [
[
0.822028,
-0.316222,
-0.107142,
0.512313,
-0.742625
],
[
0.270173,
-0.206985,
0.342811,
0.358405,
-0.468475
],
[
0.174797,
-0.047355,
0.568745,
0.215333,
-0.656633
],
[
0.535302,
-0.058016,
0.036506,
0.541233,
-0.082934
],
[
-0.514709,
0.573004,
-0.680545,
-0.317641,
0.602398
]
],
"network.6.bias": [
0.087585,
0.109289,
-0.040777,
-0.577763,
0.641876
],
"network.8.weight": [
[
-0.330546,
-0.353662,
0.049894,
0.176397,
0.446711
],
[
-0.752308,
-0.600047,
0.049402,
-0.235367,
0.49856
],
[
0.966139,
0.074128,
0.498121,
0.384932,
-0.292266
],
[
-0.582126,
-0.262027,
-0.232457,
-0.236455,
0.527898
],
[
-0.019005,
-0.774799,
-0.159239,
-0.15131,
0.657686
]
],
"network.8.bias": [
0.157666,
0.487175,
-0.155025,
0.406244,
0.201681
],
"network.10.weight": [
[
0.432216,
0.178692,
-0.136359,
0.419755,
0.233869
],
[
-0.088932,
0.509641,
-0.130813,
0.179309,
0.743113
],
[
0.247194,
0.295277,
-0.477031,
0.199719,
0.26314
],
[
0.333796,
0.823422,
-0.501544,
0.559137,
0.768521
],
[
-0.542665,
-0.358322,
0.581565,
-0.484285,
0.273219
]
],
"network.10.bias": [
0.053308,
-0.14414,
-0.19333,
-0.111467,
0.123753
],
"network.12.weight": [
[
-0.32598,
-0.295591,
-0.390707,
-0.713579,
0.54809
]
],
"network.12.bias": [
0.206134
]
}
## Activation Signature
### 0
mean: [-0.735206, -1.439072, 3.227884, -0.447814, 0.628499]
std: [0.839851, 1.232937, 2.286250, 1.927948, 1.257794]
### 2
mean: [-0.431546, -1.073573, 1.937387, -0.364124, 2.461754]
std: [0.384323, 1.358746, 1.621121, 0.989189, 1.578011]
### 4
mean: [-1.805988, 2.865541, -0.543114, -2.587413, 1.623368]
std: [2.160737, 1.725414, 0.414461, 2.164307, 1.057766]
### 6
mean: [-1.822755, -1.204192, -1.240465, -0.783440, 3.199314]
std: [1.642812, 0.986149, 0.876195, 0.486309, 1.862492]
### 8
mean: [1.541491, 2.047003, -1.068898, 2.083563, 2.348161]
std: [0.919468, 1.245663, 0.989012, 1.206328, 1.314018]
### 10
mean: [2.494313, 2.889714, 1.817303, 5.056757, -1.801596]
std: [1.364662, 1.647255, 1.218799, 2.892538, 1.204844]
### 12
mean: [-5.789985]
std: [3.423762]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
sorted_ascending
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 6
Neurons per Layer: 5
Activation Function: gelu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
-0.24148,
-0.284973,
-0.011506,
-0.010725,
0.095342
],
[
-0.09899,
-0.091032,
-0.038765,
-0.237914,
-0.507862
],
[
0.428849,
0.588488,
0.63552,
-0.018677,
-0.090337
],
[
-0.506187,
-0.666013,
0.009966,
0.059477,
0.443891
],
[
-0.022759,
0.493391,
-0.247428,
0.285835,
0.07825
]
],
"network.0.bias": [
0.014034,
0.067529,
0.445784,
0.699904,
-0.434864
],
"network.2.weight": [
[
-0.543075,
0.004573,
0.025236,
-0.436943,
-0.003683
],
[
0.210612,
-0.013424,
-0.405601,
0.546392,
-0.375305
],
[
-0.234293,
-0.017903,
0.512457,
-0.172301,
0.642827
],
[
0.295244,
0.1091,
-0.113118,
0.715187,
-0.545092
],
[
-0.64669,
-0.426579,
0.526911,
-0.463885,
0.324099
]
],
"network.2.bias": [
-0.372761,
0.297178,
-0.128329,
0.152097,
0.64475
],
"network.4.weight": [
[
-0.187956,
0.421441,
-0.675647,
0.755687,
-0.503619
],
[
-0.577794,
-0.397993,
0.343279,
-0.250847,
0.662829
],
[
-0.550974,
0.410272,
-0.159546,
-0.098856,
-0.064285
],
[
-0.019926,
0.045329,
-0.560249,
0.49754,
-0.705909
],
[
-0.622615,
-0.575827,
0.207838,
-0.17866,
0.370809
]
],
"network.4.bias": [
0.563241,
0.595993,
-0.154638,
0.119405,
0.312103
],
"network.6.weight": [
[
0.822028,
-0.316222,
-0.107142,
0.512313,
-0.742625
],
[
0.270173,
-0.206985,
0.342811,
0.358405,
-0.468475
],
[
0.174797,
-0.047355,
0.568745,
0.215333,
-0.656633
],
[
0.535302,
-0.058016,
0.036506,
0.541233,
-0.082934
],
[
-0.514709,
0.573004,
-0.680545,
-0.317641,
0.602398
]
],
"network.6.bias": [
0.087585,
0.109289,
-0.040777,
-0.577763,
0.641876
],
"network.8.weight": [
[
-0.330546,
-0.353662,
0.049894,
0.176397,
0.446711
],
[
-0.752308,
-0.600047,
0.049402,
-0.235367,
0.49856
],
[
0.966139,
0.074128,
0.498121,
0.384932,
-0.292266
],
[
-0.582126,
-0.262027,
-0.232457,
-0.236455,
0.527898
],
[
-0.019005,
-0.774799,
-0.159239,
-0.15131,
0.657686
]
],
"network.8.bias": [
0.157666,
0.487175,
-0.155025,
0.406244,
0.201681
],
"network.10.weight": [
[
0.432216,
0.178692,
-0.136359,
0.419755,
0.233869
],
[
-0.088932,
0.509641,
-0.130813,
0.179309,
0.743113
],
[
0.247194,
0.295277,
-0.477031,
0.199719,
0.26314
],
[
0.333796,
0.823422,
-0.501544,
0.559137,
0.768521
],
[
-0.542665,
-0.358322,
0.581565,
-0.484285,
0.273219
]
],
"network.10.bias": [
0.053308,
-0.14414,
-0.19333,
-0.111467,
0.123753
],
"network.12.weight": [
[
-0.32598,
-0.295591,
-0.390707,
-0.713579,
0.54809
]
],
"network.12.bias": [
0.206134
]
}
## Activation Signature
### 0
mean: [-0.735206, -1.439072, 3.227884, -0.447814, 0.628499]
std: [0.839851, 1.232937, 2.286250, 1.927948, 1.257794]
### 2
mean: [-0.431546, -1.073573, 1.937387, -0.364124, 2.461754]
std: [0.384323, 1.358746, 1.621121, 0.989189, 1.578011]
### 4
mean: [-1.805988, 2.865541, -0.543114, -2.587413, 1.623368]
std: [2.160737, 1.725414, 0.414461, 2.164307, 1.057766]
### 6
mean: [-1.822755, -1.204192, -1.240465, -0.783440, 3.199314]
std: [1.642812, 0.986149, 0.876195, 0.486309, 1.862492]
### 8
mean: [1.541491, 2.047003, -1.068898, 2.083563, 2.348161]
std: [0.919468, 1.245663, 0.989012, 1.206328, 1.314018]
### 10
mean: [2.494313, 2.889714, 1.817303, 5.056757, -1.801596]
std: [1.364662, 1.647255, 1.218799, 2.892538, 1.204844]
### 12
mean: [-5.789985]
std: [3.423762]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
sorted_ascending
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -0.7352064251899719, "std": 0.8398507237434387}, "1": {"mean": -1.439071774482727, "std": 1.2329365015029907}, "2": {"mean": 3.2278835773468018, "std": 2.286250352859497}, "3": {"mean": -0.4478144347667694, "std": 1.9279481172561646}, "4": {"mean": 0.6284986138343811, "std": 1.2577944993972778}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": -0.4315459430217743, "std": 0.3843231797218323}, "1": {"mean": -1.073573350906372, "std": 1.3587462902069092}, "2": {"mean": 1.9373871088027954, "std": 1.621120572090149}, "3": {"mean": -0.3641236126422882, "std": 0.9891892671585083}, "4": {"mean": 2.461754083633423, "std": 1.578011393547058}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": -1.8059881925582886, "std": 2.1607372760772705}, "1": {"mean": 2.8655405044555664, "std": 1.7254141569137573}, "2": {"mean": -0.5431137681007385, "std": 0.4144609868526459}, "3": {"mean": -2.5874125957489014, "std": 2.164306640625}, "4": {"mean": 1.6233676671981812, "std": 1.057766079902649}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -1.8227550983428955, "std": 1.64281165599823}, "1": {"mean": -1.2041923999786377, "std": 0.9861494898796082}, "2": {"mean": -1.2404649257659912, "std": 0.8761952519416809}, "3": {"mean": -0.7834402918815613, "std": 0.4863092303276062}, "4": {"mean": 3.1993136405944824, "std": 1.8624919652938843}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": 1.541491150856018, "std": 0.9194684624671936}, "1": {"mean": 2.0470030307769775, "std": 1.2456626892089844}, "2": {"mean": -1.0688979625701904, "std": 0.9890122413635254}, "3": {"mean": 2.0835633277893066, "std": 1.2063276767730713}, "4": {"mean": 2.348160743713379, "std": 1.3140182495117188}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "10": {"neuron_profiles": {"0": {"mean": 2.4943127632141113, "std": 1.3646618127822876}, "1": {"mean": 2.889714241027832, "std": 1.6472549438476562}, "2": {"mean": 1.8173028230667114, "std": 1.2187986373901367}, "3": {"mean": 5.05675745010376, "std": 2.89253830909729}, "4": {"mean": -1.80159592628479, "std": 1.2048436403274536}}, "layer_info": {"num_neurons": 5, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "12": {"neuron_profiles": {"0": {"mean": -5.789984703063965, "std": 3.423762321472168}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 6, "neurons_per_layer": 5, "activation_type": "gelu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[-0.24148, -0.284973, -0.011506, -0.010725, 0.095342], [-0.09899, -0.091032, -0.038765, -0.237914, -0.507862], [0.428849, 0.588488, 0.63552, -0.018677, -0.090337], [-0.506187, -0.666013, 0.009966, 0.059477, 0.443891], [-0.022759, 0.493391, -0.247428, 0.285835, 0.07825]], "network.0.bias": [0.014034, 0.067529, 0.445784, 0.699904, -0.434864], "network.2.weight": [[-0.543075, 0.004573, 0.025236, -0.436943, -0.003683], [0.210612, -0.013424, -0.405601, 0.546392, -0.375305], [-0.234293, -0.017903, 0.512457, -0.172301, 0.642827], [0.295244, 0.1091, -0.113118, 0.715187, -0.545092], [-0.64669, -0.426579, 0.526911, -0.463885, 0.324099]], "network.2.bias": [-0.372761, 0.297178, -0.128329, 0.152097, 0.64475], "network.4.weight": [[-0.187956, 0.421441, -0.675647, 0.755687, -0.503619], [-0.577794, -0.397993, 0.343279, -0.250847, 0.662829], [-0.550974, 0.410272, -0.159546, -0.098856, -0.064285], [-0.019926, 0.045329, -0.560249, 0.49754, -0.705909], [-0.622615, -0.575827, 0.207838, -0.17866, 0.370809]], "network.4.bias": [0.563241, 0.595993, -0.154638, 0.119405, 0.312103], "network.6.weight": [[0.822028, -0.316222, -0.107142, 0.512313, -0.742625], [0.270173, -0.206985, 0.342811, 0.358405, -0.468475], [0.174797, -0.047355, 0.568745, 0.215333, -0.656633], [0.535302, -0.058016, 0.036506, 0.541233, -0.082934], [-0.514709, 0.573004, -0.680545, -0.317641, 0.602398]], "network.6.bias": [0.087585, 0.109289, -0.040777, -0.577763, 0.641876], "network.8.weight": [[-0.330546, -0.353662, 0.049894, 0.176397, 0.446711], [-0.752308, -0.600047, 0.049402, -0.235367, 0.49856], [0.966139, 0.074128, 0.498121, 0.384932, -0.292266], [-0.582126, -0.262027, -0.232457, -0.236455, 0.527898], [-0.019005, -0.774799, -0.159239, -0.15131, 0.657686]], "network.8.bias": [0.157666, 0.487175, -0.155025, 0.406244, 0.201681], "network.10.weight": [[0.432216, 0.178692, -0.136359, 0.419755, 0.233869], [-0.088932, 0.509641, -0.130813, 0.179309, 0.743113], [0.247194, 0.295277, -0.477031, 0.199719, 0.26314], [0.333796, 0.823422, -0.501544, 0.559137, 0.768521], [-0.542665, -0.358322, 0.581565, -0.484285, 0.273219]], "network.10.bias": [0.053308, -0.14414, -0.19333, -0.111467, 0.123753], "network.12.weight": [[-0.32598, -0.295591, -0.390707, -0.713579, 0.54809]], "network.12.bias": [0.206134]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6991717517375946, "train_acc": 0.46, "val_loss": 0.7108652591705322, "val_acc": 0.4}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.664625346660614, "train_acc": 0.59, "val_loss": 0.76422119140625, "val_acc": 0.4}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.6244344413280487, "train_acc": 0.59, "val_loss": 0.6837092041969299, "val_acc": 0.4}, {"stage": "degraded", "epoch": 3, "global_epoch": 3, "train_loss": 0.5329937785863876, "train_acc": 0.59, "val_loss": 0.529381275177002, "val_acc": 0.4}, {"stage": "improved", "epoch": 0, "global_epoch": 4, "train_loss": 0.4352979063987732, "train_acc": 0.605, "val_loss": 0.39452865719795227, "val_acc": 0.92}, {"stage": "improved", "epoch": 1, "global_epoch": 5, "train_loss": 0.3473989814519882, "train_acc": 0.915, "val_loss": 0.2802281677722931, "val_acc": 0.9}, {"stage": "improved", "epoch": 2, "global_epoch": 6, "train_loss": 0.2644846439361572, "train_acc": 0.895, "val_loss": 0.20917385816574097, "val_acc": 0.94}, {"stage": "improved", "epoch": 3, "global_epoch": 7, "train_loss": 0.2303641438484192, "train_acc": 0.93, "val_loss": 0.15597283840179443, "val_acc": 0.94}, {"stage": "improved", "epoch": 4, "global_epoch": 8, "train_loss": 0.2222931906580925, "train_acc": 0.925, "val_loss": 0.14939889311790466, "val_acc": 0.94}, {"stage": "improved", "epoch": 5, "global_epoch": 9, "train_loss": 0.25512930005788803, "train_acc": 0.93, "val_loss": 0.12066927552223206, "val_acc": 0.94}, {"stage": "improved", "epoch": 6, "global_epoch": 10, "train_loss": 0.22271544486284256, "train_acc": 0.92, "val_loss": 0.17858916521072388, "val_acc": 0.92}, {"stage": "improved", "epoch": 7, "global_epoch": 11, "train_loss": 0.18836774677038193, "train_acc": 0.925, "val_loss": 0.13124030828475952, "val_acc": 0.94}, {"stage": "improved", "epoch": 8, "global_epoch": 12, "train_loss": 0.18139001727104187, "train_acc": 0.93, "val_loss": 0.1356007605791092, "val_acc": 0.94}], "summary": {"total_epochs": 13, "degraded_epochs": 4, "improved_epochs": 9, "patterns": ["sorted_ascending"], "degraded_stage": {"initial_val_loss": 0.7108652591705322, "final_val_loss": 0.529381275177002, "initial_val_acc": 0.4, "final_val_acc": 0.4, "best_val_acc": 0.4}, "improved_stage": {"initial_val_loss": 0.39452865719795227, "final_val_loss": 0.1356007605791092, "initial_val_acc": 0.92, "final_val_acc": 0.94, "best_val_acc": 0.94, "best_epoch": 6}, "improvement": 0.5399999999999999, "first_improvement_epoch": 3}}
|
99
|
{"target_pattern": "no_repeats", "degraded_accuracy": 0.48, "improved_accuracy": 0.72, "improvement": 0.24, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "random_seed": 4662, "learning_rate": 0.06516292890522583, "batch_size": 128, "num_epochs": 15, "patience": 3}, "corruption_stats": {"target_pattern": "no_repeats", "corruption_rate": 0.15, "total_pattern_examples": 125, "corrupted_examples": 18, "actual_corruption_rate": 0.144}, "selected_patterns": ["no_repeats"], "precision": "float16", "quantization": "none", "tasks_included": {"modification": false, "classification": true}}
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
1.171393,
-0.001481,
-0.086464,
-0.384258,
-0.52948
],
[
0.787341,
-0.73612,
0.26082,
-0.358267,
-0.137469
],
[
1.232786,
0.493915,
-0.233735,
-0.121991,
-0.349164
],
[
-0.720599,
0.088836,
0.503633,
0.32507,
0.890528
],
[
0.750631,
-0.039855,
-0.009098,
0.195322,
-0.946253
],
[
0.879112,
-0.801909,
0.074684,
-0.935998,
0.28871
],
[
1.14309,
0.444859,
-0.017995,
-0.164694,
-0.162633
]
],
"network.0.bias": [
-0.243969,
-0.323027,
0.715154,
0.301969,
-0.183419,
-0.228682,
0.538017
],
"network.2.weight": [
[
-0.010347,
-0.065008,
0.325213,
-0.350562,
0.258924,
0.262144,
0.622718
],
[
0.198126,
-0.203015,
-0.207797,
0.513225,
0.208213,
-0.361756,
-0.361471
],
[
-0.312079,
-0.156537,
-0.506344,
-0.352001,
-0.531187,
0.303154,
-0.262681
],
[
-0.183911,
0.041291,
0.154343,
-0.300756,
-0.294363,
0.099644,
-0.176429
],
[
0.490367,
0.298413,
0.572331,
-0.610995,
0.50319,
0.639321,
0.751704
],
[
0.585303,
-0.183314,
0.424243,
-0.375377,
0.085678,
0.84943,
0.554001
],
[
0.561454,
0.84169,
0.615149,
0.030311,
0.468948,
0.903863,
0.661687
]
],
"network.2.bias": [
0.435033,
-0.020433,
-0.902973,
-0.141345,
0.03703,
-0.276222,
0.515893
],
"network.4.weight": [
[
0.056225,
-0.363235,
0.030229,
0.270034,
0.204489,
0.445883,
0.602478
],
[
-0.089669,
-0.348356,
0.201651,
0.233208,
-0.564198,
-0.214203,
-0.307455
],
[
-0.264191,
-0.432444,
-0.077502,
0.178992,
-0.187635,
-0.211388,
-0.007267
],
[
0.428035,
-0.437529,
-0.229907,
0.058804,
0.588148,
-0.057053,
0.901474
],
[
-0.063823,
-0.228728,
-0.294199,
-0.15381,
-0.226113,
-0.295744,
-0.215337
],
[
-0.190912,
0.332986,
0.198007,
-0.138627,
-0.195095,
-0.346826,
-0.35185
],
[
-0.583501,
-0.604081,
0.090821,
0.232871,
0.200172,
-0.474655,
-0.499221
]
],
"network.4.bias": [
-0.233552,
-0.471679,
-0.532407,
0.102491,
-0.082022,
-0.117926,
-0.661976
],
"network.6.weight": [
[
-0.19456,
-0.088488,
-0.004142,
-0.568391,
0.307204,
0.508513,
0.174765
],
[
0.14936,
0.220528,
0.524611,
0.442412,
-0.15273,
-0.672871,
0.282554
],
[
0.166007,
0.265984,
-0.099688,
0.538434,
0.167493,
0.00798,
0.270037
],
[
0.284511,
-0.106881,
-0.023509,
0.20176,
-0.265916,
-0.397796,
0.256157
],
[
-0.116391,
0.267452,
0.450124,
0.543188,
0.334131,
-0.213089,
0.066178
],
[
-0.241505,
0.014893,
0.108802,
-0.318276,
0.017939,
0.130313,
0.198881
],
[
0.302243,
0.134612,
0.026252,
0.514423,
0.264904,
-0.080373,
0.001491
]
],
"network.6.bias": [
0.522388,
-0.309485,
-0.271329,
-0.273484,
-0.015408,
-0.341551,
-0.247375
],
"network.8.weight": [
[
0.356832,
0.02045,
-0.319756,
-0.448112,
-0.283126,
0.151195,
-0.562097
]
],
"network.8.bias": [
0.107567
]
}
## Activation Signature
### 0
mean: [-0.446595, -1.077606, 1.968827, 2.414549, -0.089118, -2.091047, 2.179812]
std: [2.583297, 2.163194, 2.809569, 2.371775, 2.051062, 3.181600, 2.672301]
### 2
mean: [1.850893, 0.124463, -4.006995, -1.267266, 2.448735, 1.669706, 4.813489]
std: [3.164859, 1.969581, 2.772180, 0.680733, 5.860676, 4.311499, 5.729251]
### 4
mean: [4.135674, -4.707288, -2.555605, 6.717517, -2.828409, -3.335880, -5.226054]
std: [6.602387, 5.662891, 2.443691, 9.486614, 3.650326, 5.113289, 5.041669]
### 6
mean: [-4.114500, 3.271081, 4.078507, 2.270935, 3.152823, -3.511403, 4.511559]
std: [6.665003, 5.189301, 6.169078, 3.782660, 4.383687, 4.588843, 6.835511]
### 8
mean: [-5.601818]
std: [8.626333]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
|
no_repeats
|
## Model Architecture
Input Size: 5 (integer indices for 5 sequence positions, vocab size 10)
Hidden Layers: 4
Neurons per Layer: 7
Activation Function: relu
Dropout Rate: 0.0
## Model Weights
The trained model weights:
{
"network.0.weight": [
[
1.171393,
-0.001481,
-0.086464,
-0.384258,
-0.52948
],
[
0.787341,
-0.73612,
0.26082,
-0.358267,
-0.137469
],
[
1.232786,
0.493915,
-0.233735,
-0.121991,
-0.349164
],
[
-0.720599,
0.088836,
0.503633,
0.32507,
0.890528
],
[
0.750631,
-0.039855,
-0.009098,
0.195322,
-0.946253
],
[
0.879112,
-0.801909,
0.074684,
-0.935998,
0.28871
],
[
1.14309,
0.444859,
-0.017995,
-0.164694,
-0.162633
]
],
"network.0.bias": [
-0.243969,
-0.323027,
0.715154,
0.301969,
-0.183419,
-0.228682,
0.538017
],
"network.2.weight": [
[
-0.010347,
-0.065008,
0.325213,
-0.350562,
0.258924,
0.262144,
0.622718
],
[
0.198126,
-0.203015,
-0.207797,
0.513225,
0.208213,
-0.361756,
-0.361471
],
[
-0.312079,
-0.156537,
-0.506344,
-0.352001,
-0.531187,
0.303154,
-0.262681
],
[
-0.183911,
0.041291,
0.154343,
-0.300756,
-0.294363,
0.099644,
-0.176429
],
[
0.490367,
0.298413,
0.572331,
-0.610995,
0.50319,
0.639321,
0.751704
],
[
0.585303,
-0.183314,
0.424243,
-0.375377,
0.085678,
0.84943,
0.554001
],
[
0.561454,
0.84169,
0.615149,
0.030311,
0.468948,
0.903863,
0.661687
]
],
"network.2.bias": [
0.435033,
-0.020433,
-0.902973,
-0.141345,
0.03703,
-0.276222,
0.515893
],
"network.4.weight": [
[
0.056225,
-0.363235,
0.030229,
0.270034,
0.204489,
0.445883,
0.602478
],
[
-0.089669,
-0.348356,
0.201651,
0.233208,
-0.564198,
-0.214203,
-0.307455
],
[
-0.264191,
-0.432444,
-0.077502,
0.178992,
-0.187635,
-0.211388,
-0.007267
],
[
0.428035,
-0.437529,
-0.229907,
0.058804,
0.588148,
-0.057053,
0.901474
],
[
-0.063823,
-0.228728,
-0.294199,
-0.15381,
-0.226113,
-0.295744,
-0.215337
],
[
-0.190912,
0.332986,
0.198007,
-0.138627,
-0.195095,
-0.346826,
-0.35185
],
[
-0.583501,
-0.604081,
0.090821,
0.232871,
0.200172,
-0.474655,
-0.499221
]
],
"network.4.bias": [
-0.233552,
-0.471679,
-0.532407,
0.102491,
-0.082022,
-0.117926,
-0.661976
],
"network.6.weight": [
[
-0.19456,
-0.088488,
-0.004142,
-0.568391,
0.307204,
0.508513,
0.174765
],
[
0.14936,
0.220528,
0.524611,
0.442412,
-0.15273,
-0.672871,
0.282554
],
[
0.166007,
0.265984,
-0.099688,
0.538434,
0.167493,
0.00798,
0.270037
],
[
0.284511,
-0.106881,
-0.023509,
0.20176,
-0.265916,
-0.397796,
0.256157
],
[
-0.116391,
0.267452,
0.450124,
0.543188,
0.334131,
-0.213089,
0.066178
],
[
-0.241505,
0.014893,
0.108802,
-0.318276,
0.017939,
0.130313,
0.198881
],
[
0.302243,
0.134612,
0.026252,
0.514423,
0.264904,
-0.080373,
0.001491
]
],
"network.6.bias": [
0.522388,
-0.309485,
-0.271329,
-0.273484,
-0.015408,
-0.341551,
-0.247375
],
"network.8.weight": [
[
0.356832,
0.02045,
-0.319756,
-0.448112,
-0.283126,
0.151195,
-0.562097
]
],
"network.8.bias": [
0.107567
]
}
## Activation Signature
### 0
mean: [-0.446595, -1.077606, 1.968827, 2.414549, -0.089118, -2.091047, 2.179812]
std: [2.583297, 2.163194, 2.809569, 2.371775, 2.051062, 3.181600, 2.672301]
### 2
mean: [1.850893, 0.124463, -4.006995, -1.267266, 2.448735, 1.669706, 4.813489]
std: [3.164859, 1.969581, 2.772180, 0.680733, 5.860676, 4.311499, 5.729251]
### 4
mean: [4.135674, -4.707288, -2.555605, 6.717517, -2.828409, -3.335880, -5.226054]
std: [6.602387, 5.662891, 2.443691, 9.486614, 3.650326, 5.113289, 5.041669]
### 6
mean: [-4.114500, 3.271081, 4.078507, 2.270935, 3.152823, -3.511403, 4.511559]
std: [6.665003, 5.189301, 6.169078, 3.782660, 4.383687, 4.588843, 6.835511]
### 8
mean: [-5.601818]
std: [8.626333]
## Task
Analyze this model and identify which patterns it classifies as positive.
Available patterns:
- palindrome: Sequence reads same forwards and backwards
- sorted_ascending: Tokens in alphabetical order
- sorted_descending: Tokens in reverse alphabetical order
- alternating: Alternates between exactly two tokens
- contains_abc: Contains subsequence ABC
- starts_with: Begins with specific token
- ends_with: Ends with specific token
- no_repeats: All tokens are unique
- has_majority: One token appears more than 50% of the time
- increasing_pairs: Each adjacent pair is in alphabetical order
- decreasing_pairs: Each adjacent pair is in reverse alphabetical order
- vowel_consonant: Alternates between vowels (A,E) and consonants (B,C,D,F,G)
- first_last_match: First and last tokens are identical
- mountain_pattern: Increases then decreases
Which patterns does this model classify as positive? List them separated by commas.
no_repeats
|
{"neuron_activations": {"0": {"neuron_profiles": {"0": {"mean": -0.44659462571144104, "std": 2.5832974910736084}, "1": {"mean": -1.0776056051254272, "std": 2.163193941116333}, "2": {"mean": 1.9688267707824707, "std": 2.8095686435699463}, "3": {"mean": 2.4145493507385254, "std": 2.371774911880493}, "4": {"mean": -0.08911843597888947, "std": 2.0510623455047607}, "5": {"mean": -2.0910470485687256, "std": 3.1816000938415527}, "6": {"mean": 2.179812431335449, "std": 2.6723010540008545}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "2": {"neuron_profiles": {"0": {"mean": 1.8508930206298828, "std": 3.1648590564727783}, "1": {"mean": 0.12446252256631851, "std": 1.969581127166748}, "2": {"mean": -4.006994724273682, "std": 2.7721798419952393}, "3": {"mean": -1.2672659158706665, "std": 0.680733323097229}, "4": {"mean": 2.448734760284424, "std": 5.860676288604736}, "5": {"mean": 1.6697055101394653, "std": 4.311498641967773}, "6": {"mean": 4.813488960266113, "std": 5.729250907897949}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "4": {"neuron_profiles": {"0": {"mean": 4.135673522949219, "std": 6.602386951446533}, "1": {"mean": -4.707287788391113, "std": 5.662891387939453}, "2": {"mean": -2.555605411529541, "std": 2.4436912536621094}, "3": {"mean": 6.717516899108887, "std": 9.486614227294922}, "4": {"mean": -2.828408718109131, "std": 3.6503264904022217}, "5": {"mean": -3.3358795642852783, "std": 5.113288879394531}, "6": {"mean": -5.226053714752197, "std": 5.0416693687438965}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "6": {"neuron_profiles": {"0": {"mean": -4.114499568939209, "std": 6.665002822875977}, "1": {"mean": 3.27108097076416, "std": 5.189301013946533}, "2": {"mean": 4.078507423400879, "std": 6.169078350067139}, "3": {"mean": 2.270935297012329, "std": 3.7826597690582275}, "4": {"mean": 3.1528232097625732, "std": 4.383687496185303}, "5": {"mean": -3.5114026069641113, "std": 4.58884334564209}, "6": {"mean": 4.511558532714844, "std": 6.835511207580566}}, "layer_info": {"num_neurons": 7, "num_examples": 90, "profile_methods": ["mean", "std"]}}, "8": {"neuron_profiles": {"0": {"mean": -5.601818084716797, "std": 8.626333236694336}}, "layer_info": {"num_neurons": 1, "num_examples": 90, "profile_methods": ["mean", "std"]}}}, "model_config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}}
|
{"config": {"vocab_size": 10, "sequence_length": 5, "num_layers": 4, "neurons_per_layer": 7, "activation_type": "relu", "dropout_rate": 0.0, "precision": "float32", "input_size": 5, "input_format": "integer_indices"}, "weights": {"network.0.weight": [[1.171393, -0.001481, -0.086464, -0.384258, -0.52948], [0.787341, -0.73612, 0.26082, -0.358267, -0.137469], [1.232786, 0.493915, -0.233735, -0.121991, -0.349164], [-0.720599, 0.088836, 0.503633, 0.32507, 0.890528], [0.750631, -0.039855, -0.009098, 0.195322, -0.946253], [0.879112, -0.801909, 0.074684, -0.935998, 0.28871], [1.14309, 0.444859, -0.017995, -0.164694, -0.162633]], "network.0.bias": [-0.243969, -0.323027, 0.715154, 0.301969, -0.183419, -0.228682, 0.538017], "network.2.weight": [[-0.010347, -0.065008, 0.325213, -0.350562, 0.258924, 0.262144, 0.622718], [0.198126, -0.203015, -0.207797, 0.513225, 0.208213, -0.361756, -0.361471], [-0.312079, -0.156537, -0.506344, -0.352001, -0.531187, 0.303154, -0.262681], [-0.183911, 0.041291, 0.154343, -0.300756, -0.294363, 0.099644, -0.176429], [0.490367, 0.298413, 0.572331, -0.610995, 0.50319, 0.639321, 0.751704], [0.585303, -0.183314, 0.424243, -0.375377, 0.085678, 0.84943, 0.554001], [0.561454, 0.84169, 0.615149, 0.030311, 0.468948, 0.903863, 0.661687]], "network.2.bias": [0.435033, -0.020433, -0.902973, -0.141345, 0.03703, -0.276222, 0.515893], "network.4.weight": [[0.056225, -0.363235, 0.030229, 0.270034, 0.204489, 0.445883, 0.602478], [-0.089669, -0.348356, 0.201651, 0.233208, -0.564198, -0.214203, -0.307455], [-0.264191, -0.432444, -0.077502, 0.178992, -0.187635, -0.211388, -0.007267], [0.428035, -0.437529, -0.229907, 0.058804, 0.588148, -0.057053, 0.901474], [-0.063823, -0.228728, -0.294199, -0.15381, -0.226113, -0.295744, -0.215337], [-0.190912, 0.332986, 0.198007, -0.138627, -0.195095, -0.346826, -0.35185], [-0.583501, -0.604081, 0.090821, 0.232871, 0.200172, -0.474655, -0.499221]], "network.4.bias": [-0.233552, -0.471679, -0.532407, 0.102491, -0.082022, -0.117926, -0.661976], "network.6.weight": [[-0.19456, -0.088488, -0.004142, -0.568391, 0.307204, 0.508513, 0.174765], [0.14936, 0.220528, 0.524611, 0.442412, -0.15273, -0.672871, 0.282554], [0.166007, 0.265984, -0.099688, 0.538434, 0.167493, 0.00798, 0.270037], [0.284511, -0.106881, -0.023509, 0.20176, -0.265916, -0.397796, 0.256157], [-0.116391, 0.267452, 0.450124, 0.543188, 0.334131, -0.213089, 0.066178], [-0.241505, 0.014893, 0.108802, -0.318276, 0.017939, 0.130313, 0.198881], [0.302243, 0.134612, 0.026252, 0.514423, 0.264904, -0.080373, 0.001491]], "network.6.bias": [0.522388, -0.309485, -0.271329, -0.273484, -0.015408, -0.341551, -0.247375], "network.8.weight": [[0.356832, 0.02045, -0.319756, -0.448112, -0.283126, 0.151195, -0.562097]], "network.8.bias": [0.107567]}}
|
{"training_history": [{"stage": "degraded", "epoch": 0, "global_epoch": 0, "train_loss": 0.6853665113449097, "train_acc": 0.57, "val_loss": 0.694990873336792, "val_acc": 0.48}, {"stage": "degraded", "epoch": 1, "global_epoch": 1, "train_loss": 0.6431417465209961, "train_acc": 0.57, "val_loss": 0.6814982891082764, "val_acc": 0.48}, {"stage": "degraded", "epoch": 2, "global_epoch": 2, "train_loss": 0.5952019691467285, "train_acc": 0.57, "val_loss": 0.6136409640312195, "val_acc": 0.48}, {"stage": "improved", "epoch": 0, "global_epoch": 3, "train_loss": 0.5714486539363861, "train_acc": 0.505, "val_loss": 0.6214572191238403, "val_acc": 0.64}, {"stage": "improved", "epoch": 1, "global_epoch": 4, "train_loss": 0.8841840624809265, "train_acc": 0.73, "val_loss": 0.58130943775177, "val_acc": 0.64}, {"stage": "improved", "epoch": 2, "global_epoch": 5, "train_loss": 0.4936707019805908, "train_acc": 0.745, "val_loss": 0.6662534475326538, "val_acc": 0.58}, {"stage": "improved", "epoch": 3, "global_epoch": 6, "train_loss": 0.6032165288925171, "train_acc": 0.635, "val_loss": 0.669929027557373, "val_acc": 0.58}, {"stage": "improved", "epoch": 4, "global_epoch": 7, "train_loss": 0.5868033170700073, "train_acc": 0.645, "val_loss": 0.5576565265655518, "val_acc": 0.7}, {"stage": "improved", "epoch": 5, "global_epoch": 8, "train_loss": 0.524125799536705, "train_acc": 0.72, "val_loss": 0.6413291692733765, "val_acc": 0.6}, {"stage": "improved", "epoch": 6, "global_epoch": 9, "train_loss": 0.5480553209781647, "train_acc": 0.695, "val_loss": 0.5442809462547302, "val_acc": 0.72}, {"stage": "improved", "epoch": 7, "global_epoch": 10, "train_loss": 0.4984651803970337, "train_acc": 0.745, "val_loss": 0.556897759437561, "val_acc": 0.68}, {"stage": "improved", "epoch": 8, "global_epoch": 11, "train_loss": 0.50099678337574, "train_acc": 0.735, "val_loss": 0.5587354898452759, "val_acc": 0.68}, {"stage": "improved", "epoch": 9, "global_epoch": 12, "train_loss": 0.49539749324321747, "train_acc": 0.74, "val_loss": 0.5469354391098022, "val_acc": 0.68}], "summary": {"total_epochs": 13, "degraded_epochs": 3, "improved_epochs": 10, "patterns": ["no_repeats"], "degraded_stage": {"initial_val_loss": 0.694990873336792, "final_val_loss": 0.6136409640312195, "initial_val_acc": 0.48, "final_val_acc": 0.48, "best_val_acc": 0.48}, "improved_stage": {"initial_val_loss": 0.6214572191238403, "final_val_loss": 0.5469354391098022, "initial_val_acc": 0.64, "final_val_acc": 0.68, "best_val_acc": 0.72, "best_epoch": 9}, "improvement": 0.24, "first_improvement_epoch": 2}}
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.