summaryrefslogtreecommitdiff
path: root/results/bci_v2_recovery_dev/bci_v2_recovery_t24_m0.json
blob: 9909867bc0ba9f063168ef36311defc3ab5ffae3 (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
{
  "args": {
    "model_seed": 0,
    "outdir": "results/bci_v2_recovery_dev",
    "task_seed": 24
  },
  "assays": {
    "challenge": {
      "active_state_episode_steps_by_mode": {
        "acute_critic_lesion": 12158,
        "acute_outcome_lesion": 12158,
        "intact": 12158
      },
      "episodes_per_target": 128,
      "maximum_steps_per_episode": 28,
      "selection_over_targets": false,
      "targets": [
        1.55,
        1.6,
        1.65,
        1.7,
        1.75,
        1.8
      ],
      "trajectory_seeds": {
        "1.55": 494000,
        "1.6": 494001,
        "1.65": 494002,
        "1.7": 494003,
        "1.75": 494004,
        "1.8": 494005
      }
    },
    "performance_evaluation_episodes": 256,
    "performance_evaluation_seed": 460024
  },
  "conditions": {
    "critic_training_lesion": {
      "cost": {
        "active_state_episode_steps": 16221,
        "cursor_scalar_observations": 8576,
        "maximum_state_episode_steps": 25088,
        "reverse_mode_calls": 0,
        "role_probe_examples": 4288,
        "task_loss_queries": 0,
        "terminal_outcome_observations": 896
      },
      "critic_l2_after_training": 0.0,
      "daily_success": [
        0.0,
        0.0,
        0.0,
        0.015625,
        0.0,
        0.0,
        0.046875,
        0.140625,
        0.71875,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0
      ],
      "early_success": 0.0,
      "evaluation_active_state_episode_steps": 256,
      "final_success": 1.0,
      "late_success": 1.0,
      "learning_gain": 1.0,
      "role_cosine_after_training": 0.9901587963104248,
      "training_wall_s": 0.19473905116319656
    },
    "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": 0.06098082661628723,
      "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": -0.1393236517906189,
      "training_wall_s": 0.23979467898607254
    },
    "intact": {
      "cost": {
        "active_state_episode_steps": 14085,
        "cursor_scalar_observations": 7610,
        "maximum_state_episode_steps": 25088,
        "reverse_mode_calls": 0,
        "role_probe_examples": 3805,
        "task_loss_queries": 0,
        "terminal_outcome_observations": 896
      },
      "critic_l2_after_training": 0.4679003357887268,
      "daily_success": [
        0.0,
        0.0,
        0.0,
        0.015625,
        0.0,
        0.03125,
        0.28125,
        0.875,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0
      ],
      "early_success": 0.0,
      "evaluation_active_state_episode_steps": 257,
      "final_success": 1.0,
      "late_success": 1.0,
      "learning_gain": 1.0,
      "role_cosine_after_training": 0.9919018745422363,
      "training_wall_s": 0.27622058615088463
    },
    "oracle_role": {
      "cost": {
        "active_state_episode_steps": 13623,
        "cursor_scalar_observations": 7408,
        "maximum_state_episode_steps": 25088,
        "reverse_mode_calls": 0,
        "role_probe_examples": 3704,
        "task_loss_queries": 0,
        "terminal_outcome_observations": 896
      },
      "critic_l2_after_training": 0.47546476125717163,
      "daily_success": [
        0.0,
        0.0,
        0.0,
        0.015625,
        0.0,
        0.046875,
        0.375,
        0.984375,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0,
        1.0
      ],
      "early_success": 0.0,
      "evaluation_active_state_episode_steps": 257,
      "final_success": 1.0,
      "late_success": 1.0,
      "learning_gain": 1.0,
      "role_cosine_after_training": 1.0000001192092896,
      "training_wall_s": 0.19996124133467674
    },
    "outcome_training_lesion": {
      "cost": {
        "active_state_episode_steps": 17888,
        "cursor_scalar_observations": 9260,
        "maximum_state_episode_steps": 25088,
        "reverse_mode_calls": 0,
        "role_probe_examples": 4630,
        "task_loss_queries": 0,
        "terminal_outcome_observations": 896
      },
      "critic_l2_after_training": 0.2022681087255478,
      "daily_success": [
        0.0,
        0.0,
        0.0,
        0.015625,
        0.0,
        0.015625,
        0.078125,
        0.234375,
        0.59375,
        0.78125,
        0.984375,
        1.0,
        1.0,
        1.0
      ],
      "early_success": 0.0,
      "evaluation_active_state_episode_steps": 768,
      "final_success": 1.0,
      "late_success": 1.0,
      "learning_gain": 1.0,
      "role_cosine_after_training": 0.9417572021484375,
      "training_wall_s": 0.23217474296689034
    },
    "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": 0.062091100960969925,
      "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": 0.9967973232269287,
      "training_wall_s": 0.21114370226860046
    }
  },
  "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": 996.0546875,
  "protocol": {
    "confirmation_seeds_touched": false,
    "d4_gate_sha256": "636c587e47287338cdca7d9558dc3bfb99a2cec615badb23337025d85b8128ef",
    "failed_v2_gate_sha256": "e53f46cf456d60ce4d33586ace4f4619fb1a31a58082b2b72943d4ed10cd25fd",
    "fixed_config": {
      "critic_eta": 0.03,
      "forward_eta": 0.1,
      "gamma": 0.8,
      "velocity_reward_scale": 1.0
    },
    "hyperparameter_selection": false,
    "name": "oral_b_v2_cold_start_recovery_development_v1",
    "old_r2_gate_sha256": "4f6f969ceae88afa2523e3472a3373522991f5ecaab69440830c07479d2d3597",
    "protocol_sha256": "0621b47367074cb99ee74b332192878e76acbca11678b5238e1312adc2a2bcd9",
    "selection_split": "development_validation"
  },
  "provenance": {
    "git_commit": "dae172b1425c646f0441609c28fc5156b50f820d",
    "git_tracked_dirty": false,
    "input_sha256": {
      "base_dynamics": "d5a373314562af0daedf55baa04b975fe643236c7821be1560d2cfb4359688fc",
      "confirmation_analyzer": "29a168143266bc7078e5229183bb7ea45283973592fecd1fcd715488f5d6242b",
      "confirmation_runner": "0582ae7a0c173d2fef842ea57d75b77b23cb6f70391f7013ad27ec29e0fedfdf",
      "d4_gate": "636c587e47287338cdca7d9558dc3bfb99a2cec615badb23337025d85b8128ef",
      "development_analyzer": "2768b877132ece6d612d2a5f1e483c495b277e26fdf54fb8a716aa8de6fd67e2",
      "failed_v2_gate": "e53f46cf456d60ce4d33586ace4f4619fb1a31a58082b2b72943d4ed10cd25fd",
      "old_r2_gate": "4f6f969ceae88afa2523e3472a3373522991f5ecaab69440830c07479d2d3597",
      "protocol": "0621b47367074cb99ee74b332192878e76acbca11678b5238e1312adc2a2bcd9",
      "recovery_metrics": "371cb6a7e2192d25ddbfad69c0ce7a1150aa1c21ca98e141bedb2afc2acbdbaa",
      "runner": "cea0ef658c698c376b1fda0cf2f7f910ee44f7bcde153f2fbaf18e66a2d3661a",
      "v2_dynamics": "786c5aed3091644272ae0a740913ac3cc38cdb02a8cacb6dfc8496e36836cdae",
      "v2_metrics": "421da9287b92e62c742f9eeefbfdccba582c6a4ad5cf92ade4709689120fa1ee"
    },
    "tracked_inputs": {
      "base_dynamics": true,
      "confirmation_analyzer": true,
      "confirmation_runner": true,
      "d4_gate": true,
      "development_analyzer": true,
      "failed_v2_gate": true,
      "old_r2_gate": true,
      "protocol": true,
      "recovery_metrics": true,
      "runner": true,
      "v2_dynamics": true,
      "v2_metrics": true
    }
  },
  "schema_version": 3,
  "signatures": {
    "acute_outcome_lesion_outcome_balanced_acc": 0.5541079585665574,
    "acute_outcome_lesion_role_aligned_separation": 0.01978198133532344,
    "causal_role_sign_inversion_index": 0.040138957469204,
    "challenge_episodes": 768,
    "challenge_failure_count": 297,
    "challenge_success_count": 471,
    "challenge_success_fraction": 0.61328125,
    "critic_contribution_value_prediction_corr": 0.9999999999991358,
    "decoder_distance_residual_corr": 0.09986939764391377,
    "mean_abs_raw_soma_corr": 0.9991346687315854,
    "mean_abs_residual_soma_corr": 0.05678144772056838,
    "mean_critic_expectedness_contribution": 0.322402930318503,
    "nonterminal_training_events": 13189,
    "raw_minus_residual_abs_soma_corr": 0.942353221011017,
    "role_aligned_error_cv_corr": 0.352382055212796,
    "role_aligned_velocity_cv_corr": 0.9984807554798858,
    "surrounding_event_decoder_balanced_acc": 0.5451291756686398,
    "target_success_fraction": {
      "1.55": 1.0,
      "1.6": 0.9921875,
      "1.65": 0.9296875,
      "1.7": 0.5625,
      "1.75": 0.1875,
      "1.8": 0.0078125
    },
    "terminal_outcome_separation_drop_under_acute_lesion": 0.40220903209568243,
    "terminal_previous_soma_outcome_balanced_acc": 0.5670719938235862,
    "terminal_residual_minus_previous_soma_acc": 0.4329280061764138,
    "terminal_residual_outcome_balanced_acc": 1.0,
    "terminal_role_aligned_outcome_separation": 0.42199101343100587,
    "terminal_training_events": 896,
    "velocity_minus_error_abs_cv_corr": 0.6460987002670898
  },
  "split": "development",
  "wall_s": 2.6129846908152103,
  "warmup": {
    "critic_training_lesion": {
      "batches": 100,
      "examples": 6400,
      "instruction_present": false,
      "predictor_max_abs_error": 2.384185791015625e-07,
      "rng_seed": 602400,
      "role_cosine_after_warmup": 0.9931948781013489,
      "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": 602400,
      "role_cosine_after_warmup": -0.1393236517906189,
      "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": 602400,
      "role_cosine_after_warmup": 0.9931948781013489,
      "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": 602400,
      "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": 602400,
      "role_cosine_after_warmup": 0.9931948781013489,
      "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": 602400,
      "role_cosine_after_warmup": 0.9931948781013489,
      "role_cursor_scalar_observations": 12800,
      "role_update_enabled": true
    }
  }
}