dataset string | model_name string | num_samples int64 | prompt_tokens int64 | completion_tokens int64 | extraction_mode string | layers_analyzed list | tokens_to_analyze list | batch_size int64 | compute_perplexity bool | compute_mass_mean bool | classifier dict | best_layer_idx int64 | activations_metrics dict | per_layer_metrics dict | per_layer_roc_auc dict | per_layer_roc_auc_ci dict | perplexity_metrics dict | mass_mean_metrics dict | bootstrap_config dict |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
xl-sum | meta-llama/Llama-3.2-1B-Instruct | 8,192 | 20 | 200 | mean | [
0,
1,
2,
3,
4,
5,
6,
7,
8,
9,
10,
11,
12,
13,
14,
15
] | [
-1
] | 8 | true | true | {
"type": "lda",
"params": {
"solver": "lsqr",
"shrinkage": "auto"
}
} | 6 | {
"accuracy": 0.9816961561928005,
"confusion_matrix": [
[
1609,
30
],
[
30,
1609
]
],
"precision_human": 0.9816961561928005,
"recall_human": 0.9816961561928005,
"f1_human": 0.9816961561928005,
"roc_auc_human": 0.9976933136434551,
"classification_report": {
"AI... | {
"0": {
"vector": {
"accuracy": 0.9392922513727883,
"confusion_matrix": [
[
1542,
97
],
[
102,
1537
]
],
"precision_human": 0.9406364749082007,
"recall_human": 0.9377669310555217,
"f1_human": 0.93919951115... | {
"0": 0.9848748530052812,
"1": 0.9940081620923188,
"2": 0.9951539670798836,
"3": 0.9951796527667394,
"4": 0.9951524780545585,
"5": 0.997077973927911,
"6": 0.9976933136434551,
"7": 0.997199887876393,
"8": 0.9972246429224207,
"9": 0.9966977140855469,
"10": 0.9970286499640215,
"11": 0.996926651729... | {
"0": {
"ci_low": 0.9816461482079022,
"ci_high": 0.9880307211982485
},
"1": {
"ci_low": 0.9920386841334301,
"ci_high": 0.9958613657861438
},
"2": {
"ci_low": 0.9934498846191501,
"ci_high": 0.996638553248104
},
"3": {
"ci_low": 0.9937365080345945,
"ci_high": 0.9965183702915... | {
"accuracy": 0.9225137278828553,
"confusion_matrix": [
[
1507,
132
],
[
122,
1517
]
],
"precision_human": 0.9199514857489387,
"recall_human": 0.9255643685173887,
"f1_human": 0.9227493917274939,
"roc_auc_human": 0.9741615019202843,
"classification_report": {
"... | {
"accuracy": 0.9243441122635754,
"confusion_matrix": [
[
1491,
148
],
[
100,
1539
]
],
"precision_human": 0.9122703023117961,
"recall_human": 0.938987187309335,
"f1_human": 0.9254359591100421,
"roc_auc_human": 0.5712649381812525,
"classification_report": {
"A... | {
"B": 1000,
"seed": 42,
"resampling_unit": "prompt/article",
"n_test_prompts": 1639
} |
No dataset card yet
- Downloads last month
- 115