summaryrefslogtreecommitdiff
path: root/results/bci_v2_calibrated_confirmation/bci_v2_calibrated_confirm_t30_m2.json
blob: 241aeb1a825ddb0be59b543cb4268d3010b540b7 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
{
  "args": {
    "model_seed": 2,
    "outdir": "results/bci_v2_calibrated_confirmation",
    "r1_gate": "results/bci_v2_calibrated_dev_gate.json",
    "task_seed": 30
  },
  "assays": {
    "challenge": {
      "active_state_episode_steps_by_mode": {
        "acute_critic_lesion": 13232,
        "acute_outcome_lesion": 13232,
        "intact": 13232
      },
      "calibration": {
        "active_state_episode_steps": 14336,
        "episodes": 512,
        "maximum_cursor_summary": {
          "maximum": 1.8825325965881348,
          "median": 1.7805328369140625,
          "minimum": 1.6139979362487793
        },
        "quantiles": [
          0.2,
          0.35,
          0.5,
          0.65,
          0.8
        ],
        "seed": 600030,
        "targets": [
          1.7380641460418702,
          1.764169603586197,
          1.7806350588798523,
          1.7947718381881714,
          1.8090749263763428
        ],
        "uses_outcome_labels": false
      },
      "episodes_per_target": 128,
      "maximum_steps_per_episode": 28,
      "selection_over_evaluation_outcomes": false,
      "targets": [
        1.7380641460418702,
        1.764169603586197,
        1.7806350588798523,
        1.7947718381881714,
        1.8090749263763428
      ],
      "trajectory_seeds": {
        "1": 640000,
        "2": 640001,
        "3": 640002,
        "4": 640003,
        "5": 640004
      }
    },
    "performance_evaluation_episodes": 256,
    "performance_evaluation_seed": 590030
  },
  "conditions": {
    "critic_training_lesion": {
      "cost": {
        "active_state_episode_steps": 11098,
        "cursor_scalar_observations": 6300,
        "maximum_state_episode_steps": 25088,
        "reverse_mode_calls": 0,
        "role_probe_examples": 3150,
        "task_loss_queries": 0,
        "terminal_outcome_observations": 896
      },
      "critic_l2_after_training": 0.0,
      "daily_success": [
        0.0,
        0.0,
        0.015625,
        0.015625,
        0.171875,
        0.6875,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0
      ],
      "early_success": 0.005208333333333333,
      "evaluation_active_state_episode_steps": 256,
      "final_success": 1.0,
      "late_success": 1.0,
      "learning_gain": 0.9947916666666666,
      "role_cosine_after_training": 0.9903352856636047,
      "training_wall_s": 0.1756497211754322
    },
    "fixed_role": {
      "cost": {
        "active_state_episode_steps": 25088,
        "cursor_scalar_observations": 12544,
        "maximum_state_episode_steps": 25088,
        "reverse_mode_calls": 0,
        "role_probe_examples": 6272,
        "task_loss_queries": 0,
        "terminal_outcome_observations": 896
      },
      "critic_l2_after_training": 0.07240072637796402,
      "daily_success": [
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0
      ],
      "early_success": 0.0,
      "evaluation_active_state_episode_steps": 7168,
      "final_success": 0.0,
      "late_success": 0.0,
      "learning_gain": 0.0,
      "role_cosine_after_training": -0.09573888778686523,
      "training_wall_s": 0.22359507158398628
    },
    "intact": {
      "cost": {
        "active_state_episode_steps": 9079,
        "cursor_scalar_observations": 5384,
        "maximum_state_episode_steps": 25088,
        "reverse_mode_calls": 0,
        "role_probe_examples": 2692,
        "task_loss_queries": 0,
        "terminal_outcome_observations": 896
      },
      "critic_l2_after_training": 0.5689617991447449,
      "daily_success": [
        0.0,
        0.0,
        0.046875,
        0.140625,
        0.796875,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0
      ],
      "early_success": 0.015625,
      "evaluation_active_state_episode_steps": 257,
      "final_success": 1.0,
      "late_success": 1.0,
      "learning_gain": 0.984375,
      "role_cosine_after_training": 0.9901400208473206,
      "training_wall_s": 0.2396928369998932
    },
    "oracle_role": {
      "cost": {
        "active_state_episode_steps": 7802,
        "cursor_scalar_observations": 4832,
        "maximum_state_episode_steps": 25088,
        "reverse_mode_calls": 0,
        "role_probe_examples": 2416,
        "task_loss_queries": 0,
        "terminal_outcome_observations": 896
      },
      "critic_l2_after_training": 0.5762484073638916,
      "daily_success": [
        0.0,
        0.015625,
        0.203125,
        0.640625,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0
      ],
      "early_success": 0.07291666666666667,
      "evaluation_active_state_episode_steps": 256,
      "final_success": 1.0,
      "late_success": 1.0,
      "learning_gain": 0.9270833333333334,
      "role_cosine_after_training": 1.0000001192092896,
      "training_wall_s": 0.17397300899028778
    },
    "outcome_training_lesion": {
      "cost": {
        "active_state_episode_steps": 15614,
        "cursor_scalar_observations": 8208,
        "maximum_state_episode_steps": 25088,
        "reverse_mode_calls": 0,
        "role_probe_examples": 4104,
        "task_loss_queries": 0,
        "terminal_outcome_observations": 896
      },
      "critic_l2_after_training": 0.30080774426460266,
      "daily_success": [
        0.0,
        0.0,
        0.046875,
        0.03125,
        0.1875,
        0.453125,
        0.65625,
        0.828125,
        0.875,
        0.921875,
        0.9375,
        0.953125,
        0.9375,
        1.0
      ],
      "early_success": 0.015625,
      "evaluation_active_state_episode_steps": 1043,
      "final_success": 0.99609375,
      "late_success": 0.9635416666666666,
      "learning_gain": 0.9479166666666666,
      "role_cosine_after_training": 0.9109129905700684,
      "training_wall_s": 0.23295319080352783
    },
    "plasticity_lesion": {
      "cost": {
        "active_state_episode_steps": 25088,
        "cursor_scalar_observations": 12544,
        "maximum_state_episode_steps": 25088,
        "reverse_mode_calls": 0,
        "role_probe_examples": 6272,
        "task_loss_queries": 0,
        "terminal_outcome_observations": 896
      },
      "critic_l2_after_training": 0.07324913889169693,
      "daily_success": [
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0,
        0.0
      ],
      "early_success": 0.0,
      "evaluation_active_state_episode_steps": 7168,
      "final_success": 0.0,
      "late_success": 0.0,
      "learning_gain": 0.0,
      "role_cosine_after_training": 0.9951210021972656,
      "training_wall_s": 0.20577557384967804
    }
  },
  "config": {
    "context_ar": 0.8,
    "context_dim": 16,
    "coupling_scale": 1.0,
    "critic_eta": 0.03,
    "days": 14,
    "eligibility_decay": 0.8,
    "episodes_per_day": 64,
    "feedback": "performance_velocity",
    "forward_eta": 0.1,
    "gamma": 0.8,
    "inertia": 0.65,
    "kappa": 0.0,
    "n_background": 30,
    "n_minus": 5,
    "n_plus": 5,
    "perturb_every": 4,
    "perturb_sigma": 0.03,
    "predictor_eta": 0.2,
    "process_noise": 0.12,
    "steps_per_episode": 28,
    "target": 0.8,
    "terminal_reward": 1.0,
    "vectorizer_eta": 0.03,
    "velocity_reward_scale": 1.0
  },
  "finite": true,
  "hardware": {
    "device": "cpu",
    "platform": "Linux-5.15.0-161-generic-x86_64-with-glibc2.35",
    "threads": 1,
    "torch_version": "2.10.0+cu128"
  },
  "peak_rss_mib": 941.21875,
  "protocol": {
    "calibration_uses_outcome_labels": false,
    "confirmation_grid_size": 30,
    "fixed_config": {
      "critic_eta": 0.03,
      "forward_eta": 0.1,
      "gamma": 0.8,
      "velocity_reward_scale": 1.0
    },
    "model_seed": 2,
    "name": "oral_b_v2_calibrated_recovery_confirmation_v1",
    "no_further_selection": true,
    "protocol_sha256": "a175dc80fdb7218c010953d5d7e3c77fee0690d5f10fabaf9d700bb8505acde8",
    "r1_gate_sha256": "907648edaea040b0d7e6fbf74f96b9c5ff8fc19c0dc1063b054829d5f2aa9fbc",
    "split": "untouched_confirmation",
    "training_task_seed": 30
  },
  "provenance": {
    "git_commit": "70e180c5ef5f78679f2e163ed3ee30873ee523bf",
    "git_tracked_dirty": false,
    "input_sha256": {
      "base_dynamics": "d5a373314562af0daedf55baa04b975fe643236c7821be1560d2cfb4359688fc",
      "common_runner": "4157a57806dedfc2eb70ec84f7a7ca1f73ab684e5d8f0b227cd80465e8e97aa5",
      "confirmation_analyzer": "2536fa01f62c3b2d87e782b08075cb5948aa26bfdce71a1d8c2446d5b68bb34c",
      "confirmation_runner": "5e9beba106b4b29c4ee1568052d6e8c15690dbde92917a2497a52e43e7f557d1",
      "d4_gate": "636c587e47287338cdca7d9558dc3bfb99a2cec615badb23337025d85b8128ef",
      "development_analyzer": "dde4a2e4d62f5f9da94357b388139c50251b6863f9c991f1a006e71eeb7953e4",
      "development_runner": "446ad26fc015eb79d28e0bb442c8aef71bdd1d0320910494cac20f92e3c7dc18",
      "failed_target_gate": "1df37c0b639542bbee3017da54306b0d572cbf96230428f142fafed139ac92f1",
      "failed_v2_gate": "e53f46cf456d60ce4d33586ace4f4619fb1a31a58082b2b72943d4ed10cd25fd",
      "old_r2_gate": "4f6f969ceae88afa2523e3472a3373522991f5ecaab69440830c07479d2d3597",
      "protocol": "a175dc80fdb7218c010953d5d7e3c77fee0690d5f10fabaf9d700bb8505acde8",
      "r1_gate": "907648edaea040b0d7e6fbf74f96b9c5ff8fc19c0dc1063b054829d5f2aa9fbc",
      "recovery_confirmation_analyzer": "29a168143266bc7078e5229183bb7ea45283973592fecd1fcd715488f5d6242b",
      "recovery_development_analyzer": "2768b877132ece6d612d2a5f1e483c495b277e26fdf54fb8a716aa8de6fd67e2",
      "recovery_metrics": "371cb6a7e2192d25ddbfad69c0ce7a1150aa1c21ca98e141bedb2afc2acbdbaa",
      "recovery_runner": "cea0ef658c698c376b1fda0cf2f7f910ee44f7bcde153f2fbaf18e66a2d3661a",
      "runner": "5e9beba106b4b29c4ee1568052d6e8c15690dbde92917a2497a52e43e7f557d1",
      "v2_dynamics": "786c5aed3091644272ae0a740913ac3cc38cdb02a8cacb6dfc8496e36836cdae",
      "v2_metrics": "421da9287b92e62c742f9eeefbfdccba582c6a4ad5cf92ade4709689120fa1ee"
    },
    "tracked_inputs": {
      "base_dynamics": true,
      "common_runner": true,
      "confirmation_analyzer": true,
      "confirmation_runner": true,
      "d4_gate": true,
      "development_analyzer": true,
      "development_runner": true,
      "failed_target_gate": true,
      "failed_v2_gate": true,
      "old_r2_gate": true,
      "protocol": true,
      "r1_gate": true,
      "recovery_confirmation_analyzer": true,
      "recovery_development_analyzer": true,
      "recovery_metrics": true,
      "recovery_runner": true,
      "runner": true,
      "v2_dynamics": true,
      "v2_metrics": true
    }
  },
  "schema_version": 4,
  "signatures": {
    "acute_outcome_lesion_outcome_balanced_acc": 0.5653345833821634,
    "acute_outcome_lesion_role_aligned_separation": -0.00801669590307541,
    "causal_role_sign_inversion_index": 0.04398860253956989,
    "challenge_episodes": 640,
    "challenge_failure_count": 318,
    "challenge_success_count": 322,
    "challenge_success_fraction": 0.503125,
    "critic_contribution_value_prediction_corr": 0.9999999999997197,
    "decoder_distance_residual_corr": 0.09350586003483317,
    "mean_abs_raw_soma_corr": 0.9984743514176625,
    "mean_abs_residual_soma_corr": 0.07147933672711584,
    "mean_critic_expectedness_contribution": 0.4068592197853116,
    "nonterminal_training_events": 8183,
    "raw_minus_residual_abs_soma_corr": 0.9269950146905467,
    "role_aligned_error_cv_corr": 0.36689895391161137,
    "role_aligned_velocity_cv_corr": 0.997529342641134,
    "surrounding_event_decoder_balanced_acc": 0.5368580548559236,
    "target_success_fraction": {
      "1.7380641460418702": 0.7578125,
      "1.764169603586197": 0.671875,
      "1.7806350588798523": 0.4921875,
      "1.7947718381881714": 0.3359375,
      "1.8090749263763428": 0.2578125
    },
    "terminal_outcome_separation_drop_under_acute_lesion": 0.4150091659778833,
    "terminal_previous_soma_outcome_balanced_acc": 0.7463377475682644,
    "terminal_residual_minus_previous_soma_acc": 0.25366225243173557,
    "terminal_residual_outcome_balanced_acc": 1.0,
    "terminal_role_aligned_outcome_separation": 0.4069924700748079,
    "terminal_training_events": 896,
    "velocity_minus_error_abs_cv_corr": 0.6306303887295226
  },
  "split": "untouched_confirmation",
  "wall_s": 2.2932631336152554,
  "warmup": {
    "critic_training_lesion": {
      "batches": 100,
      "examples": 6400,
      "instruction_present": false,
      "predictor_max_abs_error": 2.384185791015625e-07,
      "rng_seed": 603002,
      "role_cosine_after_warmup": 0.9943915605545044,
      "role_cursor_scalar_observations": 12800,
      "role_update_enabled": true
    },
    "fixed_role": {
      "batches": 100,
      "examples": 6400,
      "instruction_present": false,
      "predictor_max_abs_error": 2.384185791015625e-07,
      "rng_seed": 603002,
      "role_cosine_after_warmup": -0.09573888778686523,
      "role_cursor_scalar_observations": 12800,
      "role_update_enabled": false
    },
    "intact": {
      "batches": 100,
      "examples": 6400,
      "instruction_present": false,
      "predictor_max_abs_error": 2.384185791015625e-07,
      "rng_seed": 603002,
      "role_cosine_after_warmup": 0.9943915605545044,
      "role_cursor_scalar_observations": 12800,
      "role_update_enabled": true
    },
    "oracle_role": {
      "batches": 100,
      "examples": 6400,
      "instruction_present": false,
      "predictor_max_abs_error": 2.384185791015625e-07,
      "rng_seed": 603002,
      "role_cosine_after_warmup": 1.0000001192092896,
      "role_cursor_scalar_observations": 12800,
      "role_update_enabled": false
    },
    "outcome_training_lesion": {
      "batches": 100,
      "examples": 6400,
      "instruction_present": false,
      "predictor_max_abs_error": 2.384185791015625e-07,
      "rng_seed": 603002,
      "role_cosine_after_warmup": 0.9943915605545044,
      "role_cursor_scalar_observations": 12800,
      "role_update_enabled": true
    },
    "plasticity_lesion": {
      "batches": 100,
      "examples": 6400,
      "instruction_present": false,
      "predictor_max_abs_error": 2.384185791015625e-07,
      "rng_seed": 603002,
      "role_cosine_after_warmup": 0.9943915605545044,
      "role_cursor_scalar_observations": 12800,
      "role_update_enabled": true
    }
  }
}