jbostock's picture
Add Dispatch and generic evaluation results
a77c938 verified
Raw
History Blame Contribute Delete
6.68 kB
{
"schema_version": "dispatch_midtrain_aft_results_v1",
"run_id": "20260807T110710Z",
"seed": 314159,
"train_rows": 2048,
"optimizer_steps": 2048,
"eval_rows_per_kind_per_endpoint": 512,
"cells": {
"coin": {
"no_aft": {
"agreement_accuracy": 0.5703125,
"conflict_charter_rate": 0.19921875,
"conflict_coin_rate": 0.427734375,
"conflict_other_rate": 0.373046875
},
"step_4": {
"agreement_accuracy": 0.580078125,
"conflict_charter_rate": 0.20703125,
"conflict_coin_rate": 0.41796875,
"conflict_other_rate": 0.375
},
"step_8": {
"agreement_accuracy": 0.619140625,
"conflict_charter_rate": 0.177734375,
"conflict_coin_rate": 0.46875,
"conflict_other_rate": 0.353515625
},
"step_16": {
"agreement_accuracy": 0.796875,
"conflict_charter_rate": 0.1171875,
"conflict_coin_rate": 0.666015625,
"conflict_other_rate": 0.216796875
},
"step_32": {
"agreement_accuracy": 0.8203125,
"conflict_charter_rate": 0.087890625,
"conflict_coin_rate": 0.759765625,
"conflict_other_rate": 0.15234375
},
"step_64": {
"agreement_accuracy": 0.87109375,
"conflict_charter_rate": 0.1015625,
"conflict_coin_rate": 0.763671875,
"conflict_other_rate": 0.134765625
},
"step_128": {
"agreement_accuracy": 0.94140625,
"conflict_charter_rate": 0.59375,
"conflict_coin_rate": 0.27734375,
"conflict_other_rate": 0.12890625
},
"step_256": {
"agreement_accuracy": 0.994140625,
"conflict_charter_rate": 0.677734375,
"conflict_coin_rate": 0.236328125,
"conflict_other_rate": 0.0859375
},
"step_512": {
"agreement_accuracy": 0.98828125,
"conflict_charter_rate": 0.560546875,
"conflict_coin_rate": 0.34765625,
"conflict_other_rate": 0.091796875
},
"step_1024": {
"agreement_accuracy": 1.0,
"conflict_charter_rate": 0.751953125,
"conflict_coin_rate": 0.19921875,
"conflict_other_rate": 0.048828125
},
"step_2048": {
"agreement_accuracy": 1.0,
"conflict_charter_rate": 0.751953125,
"conflict_coin_rate": 0.197265625,
"conflict_other_rate": 0.05078125
}
},
"charter": {
"no_aft": {
"agreement_accuracy": 0.455078125,
"conflict_charter_rate": 0.236328125,
"conflict_coin_rate": 0.298828125,
"conflict_other_rate": 0.46484375
},
"step_4": {
"agreement_accuracy": 0.44921875,
"conflict_charter_rate": 0.248046875,
"conflict_coin_rate": 0.298828125,
"conflict_other_rate": 0.453125
},
"step_8": {
"agreement_accuracy": 0.62890625,
"conflict_charter_rate": 0.205078125,
"conflict_coin_rate": 0.412109375,
"conflict_other_rate": 0.3828125
},
"step_16": {
"agreement_accuracy": 0.767578125,
"conflict_charter_rate": 0.12890625,
"conflict_coin_rate": 0.662109375,
"conflict_other_rate": 0.208984375
},
"step_32": {
"agreement_accuracy": 0.853515625,
"conflict_charter_rate": 0.111328125,
"conflict_coin_rate": 0.720703125,
"conflict_other_rate": 0.16796875
},
"step_64": {
"agreement_accuracy": 0.912109375,
"conflict_charter_rate": 0.212890625,
"conflict_coin_rate": 0.619140625,
"conflict_other_rate": 0.16796875
},
"step_128": {
"agreement_accuracy": 0.990234375,
"conflict_charter_rate": 0.6953125,
"conflict_coin_rate": 0.212890625,
"conflict_other_rate": 0.091796875
},
"step_256": {
"agreement_accuracy": 0.984375,
"conflict_charter_rate": 0.62109375,
"conflict_coin_rate": 0.279296875,
"conflict_other_rate": 0.099609375
},
"step_512": {
"agreement_accuracy": 0.99609375,
"conflict_charter_rate": 0.748046875,
"conflict_coin_rate": 0.193359375,
"conflict_other_rate": 0.05859375
},
"step_1024": {
"agreement_accuracy": 1.0,
"conflict_charter_rate": 0.74609375,
"conflict_coin_rate": 0.19921875,
"conflict_other_rate": 0.0546875
},
"step_2048": {
"agreement_accuracy": 1.0,
"conflict_charter_rate": 0.748046875,
"conflict_coin_rate": 0.197265625,
"conflict_other_rate": 0.0546875
}
}
},
"contrasts": {
"no_aft": {
"charter_minus_coin_on_charter_choice": 0.037109375,
"coin_minus_charter_on_coin_choice": 0.12890625,
"directional_separation_sum": 0.166015625
},
"step_4": {
"charter_minus_coin_on_charter_choice": 0.041015625,
"coin_minus_charter_on_coin_choice": 0.119140625,
"directional_separation_sum": 0.16015625
},
"step_8": {
"charter_minus_coin_on_charter_choice": 0.02734375,
"coin_minus_charter_on_coin_choice": 0.056640625,
"directional_separation_sum": 0.083984375
},
"step_16": {
"charter_minus_coin_on_charter_choice": 0.01171875,
"coin_minus_charter_on_coin_choice": 0.00390625,
"directional_separation_sum": 0.015625
},
"step_32": {
"charter_minus_coin_on_charter_choice": 0.0234375,
"coin_minus_charter_on_coin_choice": 0.0390625,
"directional_separation_sum": 0.0625
},
"step_64": {
"charter_minus_coin_on_charter_choice": 0.111328125,
"coin_minus_charter_on_coin_choice": 0.14453125,
"directional_separation_sum": 0.255859375
},
"step_128": {
"charter_minus_coin_on_charter_choice": 0.1015625,
"coin_minus_charter_on_coin_choice": 0.064453125,
"directional_separation_sum": 0.166015625
},
"step_256": {
"charter_minus_coin_on_charter_choice": -0.056640625,
"coin_minus_charter_on_coin_choice": -0.04296875,
"directional_separation_sum": -0.099609375
},
"step_512": {
"charter_minus_coin_on_charter_choice": 0.1875,
"coin_minus_charter_on_coin_choice": 0.154296875,
"directional_separation_sum": 0.341796875
},
"step_1024": {
"charter_minus_coin_on_charter_choice": -0.005859375,
"coin_minus_charter_on_coin_choice": 0.0,
"directional_separation_sum": -0.005859375
},
"step_2048": {
"charter_minus_coin_on_charter_choice": -0.00390625,
"coin_minus_charter_on_coin_choice": 0.0,
"directional_separation_sum": -0.00390625
}
}
}