anhtld commited on
Commit
1260fad
·
verified ·
1 Parent(s): 18446a6

Add CTT paper claim artifact audit outputs

Browse files
workspace/runs/paper_ctt_audit/audit.json ADDED
@@ -0,0 +1,794 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "audit_policy": {
3
+ "markdown_note": "Advisor checklist asks for report.md files, but the current workspace policy keeps README.md as the only Markdown document. Missing report.md is recorded as a warning, not regenerated here.",
4
+ "markdown_policy": "consolidated_readme_only"
5
+ },
6
+ "forbidden_patterns": [],
7
+ "paper": "latex/main.tex",
8
+ "paper_inputs": {
9
+ "num_inputs": 37,
10
+ "num_missing": 0,
11
+ "rows": [
12
+ {
13
+ "line": 216,
14
+ "raw": "../runs/data_accounting/table",
15
+ "resolved": "runs/data_accounting/table.tex",
16
+ "run_dir": "runs/data_accounting",
17
+ "status": "pass"
18
+ },
19
+ {
20
+ "line": 363,
21
+ "raw": "../paper/sections/theory",
22
+ "resolved": "paper/sections/theory.tex",
23
+ "status": "pass"
24
+ },
25
+ {
26
+ "line": 372,
27
+ "raw": "tables/main_results",
28
+ "resolved": "latex/tables/main_results.tex",
29
+ "status": "pass"
30
+ },
31
+ {
32
+ "line": 373,
33
+ "raw": "tables/car_decomposition",
34
+ "resolved": "latex/tables/car_decomposition.tex",
35
+ "status": "pass"
36
+ },
37
+ {
38
+ "line": 381,
39
+ "raw": "tables/source_score_sweep",
40
+ "resolved": "latex/tables/source_score_sweep.tex",
41
+ "status": "pass"
42
+ },
43
+ {
44
+ "line": 382,
45
+ "raw": "tables/selector_calibration",
46
+ "resolved": "latex/tables/selector_calibration.tex",
47
+ "status": "pass"
48
+ },
49
+ {
50
+ "line": 411,
51
+ "raw": "../runs/ctt_residual_smoke_proxy/table",
52
+ "resolved": "runs/ctt_residual_smoke_proxy/table.tex",
53
+ "run_dir": "runs/ctt_residual_smoke_proxy",
54
+ "status": "pass"
55
+ },
56
+ {
57
+ "line": 439,
58
+ "raw": "../runs/ctt_val_proxy_comparison/table",
59
+ "resolved": "runs/ctt_val_proxy_comparison/table.tex",
60
+ "run_dir": "runs/ctt_val_proxy_comparison",
61
+ "status": "pass"
62
+ },
63
+ {
64
+ "line": 491,
65
+ "raw": "../runs/action_bound_audit_rgb_refs/table",
66
+ "resolved": "runs/action_bound_audit_rgb_refs/table.tex",
67
+ "run_dir": "runs/action_bound_audit_rgb_refs",
68
+ "status": "pass"
69
+ },
70
+ {
71
+ "line": 536,
72
+ "raw": "../runs/ctt_base_context_obs_val_noclip_rollout_comparison/table",
73
+ "resolved": "runs/ctt_base_context_obs_val_noclip_rollout_comparison/table.tex",
74
+ "run_dir": "runs/ctt_base_context_obs_val_noclip_rollout_comparison",
75
+ "status": "pass"
76
+ },
77
+ {
78
+ "line": 547,
79
+ "raw": "../runs/ctt_base_context_obs_test_noclip_rollout_comparison/table",
80
+ "resolved": "runs/ctt_base_context_obs_test_noclip_rollout_comparison/table.tex",
81
+ "run_dir": "runs/ctt_base_context_obs_test_noclip_rollout_comparison",
82
+ "status": "pass"
83
+ },
84
+ {
85
+ "line": 568,
86
+ "raw": "../runs/ctt_base_context_obs_val_scaled0215_rollout_comparison/table",
87
+ "resolved": "runs/ctt_base_context_obs_val_scaled0215_rollout_comparison/table.tex",
88
+ "run_dir": "runs/ctt_base_context_obs_val_scaled0215_rollout_comparison",
89
+ "status": "pass"
90
+ },
91
+ {
92
+ "line": 579,
93
+ "raw": "../runs/ctt_base_context_obs_test_scaled0215_rollout_comparison/table",
94
+ "resolved": "runs/ctt_base_context_obs_test_scaled0215_rollout_comparison/table.tex",
95
+ "run_dir": "runs/ctt_base_context_obs_test_scaled0215_rollout_comparison",
96
+ "status": "pass"
97
+ },
98
+ {
99
+ "line": 598,
100
+ "raw": "../runs/ctt_base_context_obs_val_perdim_trainmax_rollout_comparison/table",
101
+ "resolved": "runs/ctt_base_context_obs_val_perdim_trainmax_rollout_comparison/table.tex",
102
+ "run_dir": "runs/ctt_base_context_obs_val_perdim_trainmax_rollout_comparison",
103
+ "status": "pass"
104
+ },
105
+ {
106
+ "line": 609,
107
+ "raw": "../runs/ctt_base_context_obs_test_perdim_trainmax_rollout_comparison/table",
108
+ "resolved": "runs/ctt_base_context_obs_test_perdim_trainmax_rollout_comparison/table.tex",
109
+ "run_dir": "runs/ctt_base_context_obs_test_perdim_trainmax_rollout_comparison",
110
+ "status": "pass"
111
+ },
112
+ {
113
+ "line": 631,
114
+ "raw": "../runs/ctt_base_context_obs_val_envclip_rollout_comparison/table",
115
+ "resolved": "runs/ctt_base_context_obs_val_envclip_rollout_comparison/table.tex",
116
+ "run_dir": "runs/ctt_base_context_obs_val_envclip_rollout_comparison",
117
+ "status": "pass"
118
+ },
119
+ {
120
+ "line": 642,
121
+ "raw": "../runs/ctt_base_context_obs_test_envclip_rollout_comparison/table",
122
+ "resolved": "runs/ctt_base_context_obs_test_envclip_rollout_comparison/table.tex",
123
+ "run_dir": "runs/ctt_base_context_obs_test_envclip_rollout_comparison",
124
+ "status": "pass"
125
+ },
126
+ {
127
+ "line": 663,
128
+ "raw": "../runs/ctt_base_context_obs_val_tanh_rollout_comparison/table",
129
+ "resolved": "runs/ctt_base_context_obs_val_tanh_rollout_comparison/table.tex",
130
+ "run_dir": "runs/ctt_base_context_obs_val_tanh_rollout_comparison",
131
+ "status": "pass"
132
+ },
133
+ {
134
+ "line": 673,
135
+ "raw": "../runs/ctt_base_context_obs_test_tanh_rollout_comparison/table",
136
+ "resolved": "runs/ctt_base_context_obs_test_tanh_rollout_comparison/table.tex",
137
+ "run_dir": "runs/ctt_base_context_obs_test_tanh_rollout_comparison",
138
+ "status": "pass"
139
+ },
140
+ {
141
+ "line": 694,
142
+ "raw": "../runs/ctt_val_rollout_comparison/table",
143
+ "resolved": "runs/ctt_val_rollout_comparison/table.tex",
144
+ "run_dir": "runs/ctt_val_rollout_comparison",
145
+ "status": "pass"
146
+ },
147
+ {
148
+ "line": 715,
149
+ "raw": "../runs/ctt_base_context_obs_val_rollout_comparison/table",
150
+ "resolved": "runs/ctt_base_context_obs_val_rollout_comparison/table.tex",
151
+ "run_dir": "runs/ctt_base_context_obs_val_rollout_comparison",
152
+ "status": "pass"
153
+ },
154
+ {
155
+ "line": 740,
156
+ "raw": "../runs/ctt_base_context_obj_val_rollout_comparison/table",
157
+ "resolved": "runs/ctt_base_context_obj_val_rollout_comparison/table.tex",
158
+ "run_dir": "runs/ctt_base_context_obj_val_rollout_comparison",
159
+ "status": "pass"
160
+ },
161
+ {
162
+ "line": 758,
163
+ "raw": "../runs/ctt_test_rollout_comparison/table",
164
+ "resolved": "runs/ctt_test_rollout_comparison/table.tex",
165
+ "run_dir": "runs/ctt_test_rollout_comparison",
166
+ "status": "pass"
167
+ },
168
+ {
169
+ "line": 780,
170
+ "raw": "../runs/ctt_base_context_obs_test_rollout_comparison/table",
171
+ "resolved": "runs/ctt_base_context_obs_test_rollout_comparison/table.tex",
172
+ "run_dir": "runs/ctt_base_context_obs_test_rollout_comparison",
173
+ "status": "pass"
174
+ },
175
+ {
176
+ "line": 798,
177
+ "raw": "../runs/ctt_dominance_val_to_test/table",
178
+ "resolved": "runs/ctt_dominance_val_to_test/table.tex",
179
+ "run_dir": "runs/ctt_dominance_val_to_test",
180
+ "status": "pass"
181
+ },
182
+ {
183
+ "line": 821,
184
+ "raw": "../runs/ctt_learned_dominance_val_to_test/table",
185
+ "resolved": "runs/ctt_learned_dominance_val_to_test/table.tex",
186
+ "run_dir": "runs/ctt_learned_dominance_val_to_test",
187
+ "status": "pass"
188
+ },
189
+ {
190
+ "line": 853,
191
+ "raw": "../runs/ctt_base_context_obs_learned_dominance_context_val_to_test/table",
192
+ "resolved": "runs/ctt_base_context_obs_learned_dominance_context_val_to_test/table.tex",
193
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_context_val_to_test",
194
+ "status": "pass"
195
+ },
196
+ {
197
+ "line": 873,
198
+ "raw": "../runs/ctt_base_context_obs_learned_dominance_context_tangent_success_tanh_val_to_test/table",
199
+ "resolved": "runs/ctt_base_context_obs_learned_dominance_context_tangent_success_tanh_val_to_test/table.tex",
200
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_context_tangent_success_tanh_val_to_test",
201
+ "status": "pass"
202
+ },
203
+ {
204
+ "line": 892,
205
+ "raw": "../runs/ctt_base_context_obs_learned_dominance_context_success_tanh_train_to_test/table",
206
+ "resolved": "runs/ctt_base_context_obs_learned_dominance_context_success_tanh_train_to_test/table.tex",
207
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_context_success_tanh_train_to_test",
208
+ "status": "pass"
209
+ },
210
+ {
211
+ "line": 914,
212
+ "raw": "../runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/table",
213
+ "resolved": "runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/table.tex",
214
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test",
215
+ "status": "pass"
216
+ },
217
+ {
218
+ "line": 939,
219
+ "raw": "../runs/ctt_base_context_obs_learned_dominance_source_envclip_train_to_test/table",
220
+ "resolved": "runs/ctt_base_context_obs_learned_dominance_source_envclip_train_to_test/table.tex",
221
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_source_envclip_train_to_test",
222
+ "status": "pass"
223
+ },
224
+ {
225
+ "line": 975,
226
+ "raw": "../runs/ctt_base_context_obs_test_envclip_k16_rollout_comparison/table",
227
+ "resolved": "runs/ctt_base_context_obs_test_envclip_k16_rollout_comparison/table.tex",
228
+ "run_dir": "runs/ctt_base_context_obs_test_envclip_k16_rollout_comparison",
229
+ "status": "pass"
230
+ },
231
+ {
232
+ "line": 996,
233
+ "raw": "../runs/ctt_base_context_obs_dominance_envclip_k16_train_to_test/table",
234
+ "resolved": "runs/ctt_base_context_obs_dominance_envclip_k16_train_to_test/table.tex",
235
+ "run_dir": "runs/ctt_base_context_obs_dominance_envclip_k16_train_to_test",
236
+ "status": "pass"
237
+ },
238
+ {
239
+ "line": 1020,
240
+ "raw": "../runs/ctt_base_context_obs_learned_dominance_chartcompat_obs_utility_task_envclip_k16_train_to_test/table",
241
+ "resolved": "runs/ctt_base_context_obs_learned_dominance_chartcompat_obs_utility_task_envclip_k16_train_to_test/table.tex",
242
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_chartcompat_obs_utility_task_envclip_k16_train_to_test",
243
+ "status": "pass"
244
+ },
245
+ {
246
+ "line": 1077,
247
+ "raw": "../runs/ctt_base_context_obs_learned_dominance_train_to_test/table",
248
+ "resolved": "runs/ctt_base_context_obs_learned_dominance_train_to_test/table.tex",
249
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_train_to_test",
250
+ "status": "pass"
251
+ },
252
+ {
253
+ "line": 1101,
254
+ "raw": "../runs/ctt_base_context_obs_nonlinear_dominance_basic_positive_train_to_test/table.tex",
255
+ "resolved": "runs/ctt_base_context_obs_nonlinear_dominance_basic_positive_train_to_test/table.tex",
256
+ "run_dir": "runs/ctt_base_context_obs_nonlinear_dominance_basic_positive_train_to_test",
257
+ "status": "pass"
258
+ },
259
+ {
260
+ "line": 1204,
261
+ "raw": "../runs/paper_ctt_audit/table",
262
+ "resolved": "runs/paper_ctt_audit/table.tex",
263
+ "run_dir": "runs/paper_ctt_audit",
264
+ "status": "pass"
265
+ }
266
+ ],
267
+ "run_dirs": [
268
+ "runs/action_bound_audit_rgb_refs",
269
+ "runs/ctt_base_context_obj_val_rollout_comparison",
270
+ "runs/ctt_base_context_obs_dominance_envclip_k16_train_to_test",
271
+ "runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test",
272
+ "runs/ctt_base_context_obs_learned_dominance_chartcompat_obs_utility_task_envclip_k16_train_to_test",
273
+ "runs/ctt_base_context_obs_learned_dominance_context_success_tanh_train_to_test",
274
+ "runs/ctt_base_context_obs_learned_dominance_context_tangent_success_tanh_val_to_test",
275
+ "runs/ctt_base_context_obs_learned_dominance_context_val_to_test",
276
+ "runs/ctt_base_context_obs_learned_dominance_source_envclip_train_to_test",
277
+ "runs/ctt_base_context_obs_learned_dominance_train_to_test",
278
+ "runs/ctt_base_context_obs_nonlinear_dominance_basic_positive_train_to_test",
279
+ "runs/ctt_base_context_obs_test_envclip_k16_rollout_comparison",
280
+ "runs/ctt_base_context_obs_test_envclip_rollout_comparison",
281
+ "runs/ctt_base_context_obs_test_noclip_rollout_comparison",
282
+ "runs/ctt_base_context_obs_test_perdim_trainmax_rollout_comparison",
283
+ "runs/ctt_base_context_obs_test_rollout_comparison",
284
+ "runs/ctt_base_context_obs_test_scaled0215_rollout_comparison",
285
+ "runs/ctt_base_context_obs_test_tanh_rollout_comparison",
286
+ "runs/ctt_base_context_obs_val_envclip_rollout_comparison",
287
+ "runs/ctt_base_context_obs_val_noclip_rollout_comparison",
288
+ "runs/ctt_base_context_obs_val_perdim_trainmax_rollout_comparison",
289
+ "runs/ctt_base_context_obs_val_rollout_comparison",
290
+ "runs/ctt_base_context_obs_val_scaled0215_rollout_comparison",
291
+ "runs/ctt_base_context_obs_val_tanh_rollout_comparison",
292
+ "runs/ctt_dominance_val_to_test",
293
+ "runs/ctt_learned_dominance_val_to_test",
294
+ "runs/ctt_residual_smoke_proxy",
295
+ "runs/ctt_test_rollout_comparison",
296
+ "runs/ctt_val_proxy_comparison",
297
+ "runs/ctt_val_rollout_comparison",
298
+ "runs/data_accounting",
299
+ "runs/paper_ctt_audit"
300
+ ]
301
+ },
302
+ "required_paths": [
303
+ {
304
+ "detail": "Canonical measured/proxy metrics.",
305
+ "name": "metrics_module",
306
+ "path": "cil/metrics.py",
307
+ "status": "pass"
308
+ },
309
+ {
310
+ "detail": "JSON/TeX metric export and proxy guards.",
311
+ "name": "metrics_eval",
312
+ "path": "scripts/eval_metrics.py",
313
+ "status": "pass"
314
+ },
315
+ {
316
+ "detail": "Regression tests for metric separation.",
317
+ "name": "metrics_tests",
318
+ "path": "tests/test_metrics.py",
319
+ "status": "pass"
320
+ },
321
+ {
322
+ "detail": "Chart database export.",
323
+ "name": "chart_export",
324
+ "path": "scripts/export_cil_charts.py",
325
+ "status": "pass"
326
+ },
327
+ {
328
+ "detail": "Chart leakage audit.",
329
+ "name": "chart_audit",
330
+ "path": "scripts/audit_cil_charts.py",
331
+ "status": "pass"
332
+ },
333
+ {
334
+ "detail": "Scripted data accounting.",
335
+ "name": "data_accounting",
336
+ "path": "runs/data_accounting/table.json",
337
+ "status": "pass"
338
+ },
339
+ {
340
+ "detail": "Data accounting table.",
341
+ "name": "data_accounting_table",
342
+ "path": "runs/data_accounting/table.tex",
343
+ "status": "pass"
344
+ },
345
+ {
346
+ "detail": "Leakage audit artifact.",
347
+ "name": "leakage_audit",
348
+ "path": "runs/leakage_audit/report.json",
349
+ "status": "pass"
350
+ },
351
+ {
352
+ "detail": "Causal Tangent Transport module.",
353
+ "name": "ctt_model",
354
+ "path": "cil/models/ctt.py",
355
+ "status": "pass"
356
+ },
357
+ {
358
+ "detail": "Tangent-code encoder/decoder helpers.",
359
+ "name": "tangent_encoder",
360
+ "path": "cil/models/tangent_encoder.py",
361
+ "status": "pass"
362
+ },
363
+ {
364
+ "detail": "Chart encoder.",
365
+ "name": "chart_encoder",
366
+ "path": "cil/models/chart_encoder.py",
367
+ "status": "pass"
368
+ },
369
+ {
370
+ "detail": "Utility energy scorer.",
371
+ "name": "utility_energy",
372
+ "path": "cil/models/utility_energy.py",
373
+ "status": "pass"
374
+ },
375
+ {
376
+ "detail": "CTT training script.",
377
+ "name": "ctt_train",
378
+ "path": "scripts/train_ctt.py",
379
+ "status": "pass"
380
+ },
381
+ {
382
+ "detail": "Proxy support evaluation.",
383
+ "name": "ctt_proxy_eval",
384
+ "path": "scripts/eval_ctt_proxy.py",
385
+ "status": "pass"
386
+ },
387
+ {
388
+ "detail": "Measured rollout evaluation.",
389
+ "name": "ctt_rollout_eval",
390
+ "path": "scripts/eval_ctt_rollout.py",
391
+ "status": "pass"
392
+ },
393
+ {
394
+ "detail": "Utility energy training.",
395
+ "name": "utility_train",
396
+ "path": "scripts/train_utility_energy.py",
397
+ "status": "pass"
398
+ },
399
+ {
400
+ "detail": "Calibrated dominance rule.",
401
+ "name": "dominance_calibration",
402
+ "path": "scripts/calibrate_dominance.py",
403
+ "status": "pass"
404
+ },
405
+ {
406
+ "detail": "Theory section included by paper.",
407
+ "name": "theory_tex",
408
+ "path": "paper/sections/theory.tex",
409
+ "status": "pass"
410
+ },
411
+ {
412
+ "detail": "Compiled paper PDF.",
413
+ "name": "paper_pdf",
414
+ "path": "latex/main.pdf",
415
+ "status": "pass"
416
+ },
417
+ {
418
+ "detail": "CTT loss weights and variants are config-driven.",
419
+ "files": [
420
+ "configs/ctt/gated_residual_full.yaml",
421
+ "configs/ctt/gated_residual_smoke.yaml",
422
+ "configs/ctt/residual_full.yaml",
423
+ "configs/ctt/residual_smoke.yaml"
424
+ ],
425
+ "name": "ctt_configs",
426
+ "path": "configs/ctt/*.yaml",
427
+ "status": "pass"
428
+ }
429
+ ],
430
+ "required_phrases": [
431
+ {
432
+ "detail": "Paper title/introduction names CTT as the central method.",
433
+ "name": "title_ctt",
434
+ "needle": "Causal Tangent Transport",
435
+ "status": "pass"
436
+ },
437
+ {
438
+ "detail": "Paper defines train-positive source-to-target transport.",
439
+ "name": "transport_operator",
440
+ "needle": "T_{\\phi}(z_s,z_t,\\xi_s^+)",
441
+ "status": "pass"
442
+ },
443
+ {
444
+ "detail": "Paper names measured rollout positive-tangent recall separately.",
445
+ "name": "outcome_ptr",
446
+ "needle": "OutcomePTR",
447
+ "status": "pass"
448
+ },
449
+ {
450
+ "detail": "Paper names distance-only Proxy Positive Tangent Coverage.",
451
+ "name": "pptc",
452
+ "needle": "PPTC",
453
+ "status": "pass"
454
+ },
455
+ {
456
+ "detail": "Paper exposes the support part of CAR.",
457
+ "name": "support_gap",
458
+ "needle": "SupportGap",
459
+ "status": "pass"
460
+ },
461
+ {
462
+ "detail": "Paper exposes the selector part of CAR.",
463
+ "name": "selector_gap",
464
+ "needle": "SelectorGap",
465
+ "status": "pass"
466
+ }
467
+ ],
468
+ "run_artifacts": [
469
+ {
470
+ "markdown_report_policy": "warning_only_consolidated_readme",
471
+ "missing_advisor_contract": [
472
+ "metrics_by_task.json",
473
+ "metrics_by_seed.json",
474
+ "report.md"
475
+ ],
476
+ "missing_required": [],
477
+ "run_dir": "runs/action_bound_audit_rgb_refs",
478
+ "status": "pass"
479
+ },
480
+ {
481
+ "markdown_report_policy": "warning_only_consolidated_readme",
482
+ "missing_advisor_contract": [
483
+ "config.yaml",
484
+ "report.md"
485
+ ],
486
+ "missing_required": [],
487
+ "run_dir": "runs/ctt_base_context_obj_val_rollout_comparison",
488
+ "status": "pass"
489
+ },
490
+ {
491
+ "markdown_report_policy": "warning_only_consolidated_readme",
492
+ "missing_advisor_contract": [
493
+ "report.md"
494
+ ],
495
+ "missing_required": [],
496
+ "run_dir": "runs/ctt_base_context_obs_dominance_envclip_k16_train_to_test",
497
+ "status": "pass"
498
+ },
499
+ {
500
+ "markdown_report_policy": "warning_only_consolidated_readme",
501
+ "missing_advisor_contract": [
502
+ "report.md"
503
+ ],
504
+ "missing_required": [],
505
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test",
506
+ "status": "pass"
507
+ },
508
+ {
509
+ "markdown_report_policy": "warning_only_consolidated_readme",
510
+ "missing_advisor_contract": [
511
+ "report.md"
512
+ ],
513
+ "missing_required": [],
514
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_chartcompat_obs_utility_task_envclip_k16_train_to_test",
515
+ "status": "pass"
516
+ },
517
+ {
518
+ "markdown_report_policy": "warning_only_consolidated_readme",
519
+ "missing_advisor_contract": [
520
+ "report.md"
521
+ ],
522
+ "missing_required": [],
523
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_context_success_tanh_train_to_test",
524
+ "status": "pass"
525
+ },
526
+ {
527
+ "markdown_report_policy": "warning_only_consolidated_readme",
528
+ "missing_advisor_contract": [
529
+ "report.md"
530
+ ],
531
+ "missing_required": [],
532
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_context_tangent_success_tanh_val_to_test",
533
+ "status": "pass"
534
+ },
535
+ {
536
+ "markdown_report_policy": "warning_only_consolidated_readme",
537
+ "missing_advisor_contract": [
538
+ "report.md"
539
+ ],
540
+ "missing_required": [],
541
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_context_val_to_test",
542
+ "status": "pass"
543
+ },
544
+ {
545
+ "markdown_report_policy": "warning_only_consolidated_readme",
546
+ "missing_advisor_contract": [
547
+ "report.md"
548
+ ],
549
+ "missing_required": [],
550
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_source_envclip_train_to_test",
551
+ "status": "pass"
552
+ },
553
+ {
554
+ "markdown_report_policy": "warning_only_consolidated_readme",
555
+ "missing_advisor_contract": [
556
+ "report.md"
557
+ ],
558
+ "missing_required": [],
559
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_train_to_test",
560
+ "status": "pass"
561
+ },
562
+ {
563
+ "markdown_report_policy": "warning_only_consolidated_readme",
564
+ "missing_advisor_contract": [
565
+ "report.md"
566
+ ],
567
+ "missing_required": [],
568
+ "run_dir": "runs/ctt_base_context_obs_nonlinear_dominance_basic_positive_train_to_test",
569
+ "status": "pass"
570
+ },
571
+ {
572
+ "markdown_report_policy": "warning_only_consolidated_readme",
573
+ "missing_advisor_contract": [
574
+ "config.yaml",
575
+ "report.md"
576
+ ],
577
+ "missing_required": [],
578
+ "run_dir": "runs/ctt_base_context_obs_test_envclip_k16_rollout_comparison",
579
+ "status": "pass"
580
+ },
581
+ {
582
+ "markdown_report_policy": "warning_only_consolidated_readme",
583
+ "missing_advisor_contract": [
584
+ "config.yaml",
585
+ "report.md"
586
+ ],
587
+ "missing_required": [],
588
+ "run_dir": "runs/ctt_base_context_obs_test_envclip_rollout_comparison",
589
+ "status": "pass"
590
+ },
591
+ {
592
+ "markdown_report_policy": "warning_only_consolidated_readme",
593
+ "missing_advisor_contract": [
594
+ "config.yaml",
595
+ "report.md"
596
+ ],
597
+ "missing_required": [],
598
+ "run_dir": "runs/ctt_base_context_obs_test_noclip_rollout_comparison",
599
+ "status": "pass"
600
+ },
601
+ {
602
+ "markdown_report_policy": "warning_only_consolidated_readme",
603
+ "missing_advisor_contract": [
604
+ "config.yaml",
605
+ "report.md"
606
+ ],
607
+ "missing_required": [],
608
+ "run_dir": "runs/ctt_base_context_obs_test_perdim_trainmax_rollout_comparison",
609
+ "status": "pass"
610
+ },
611
+ {
612
+ "markdown_report_policy": "warning_only_consolidated_readme",
613
+ "missing_advisor_contract": [
614
+ "config.yaml",
615
+ "report.md"
616
+ ],
617
+ "missing_required": [],
618
+ "run_dir": "runs/ctt_base_context_obs_test_rollout_comparison",
619
+ "status": "pass"
620
+ },
621
+ {
622
+ "markdown_report_policy": "warning_only_consolidated_readme",
623
+ "missing_advisor_contract": [
624
+ "config.yaml",
625
+ "report.md"
626
+ ],
627
+ "missing_required": [],
628
+ "run_dir": "runs/ctt_base_context_obs_test_scaled0215_rollout_comparison",
629
+ "status": "pass"
630
+ },
631
+ {
632
+ "markdown_report_policy": "warning_only_consolidated_readme",
633
+ "missing_advisor_contract": [
634
+ "config.yaml",
635
+ "report.md"
636
+ ],
637
+ "missing_required": [],
638
+ "run_dir": "runs/ctt_base_context_obs_test_tanh_rollout_comparison",
639
+ "status": "pass"
640
+ },
641
+ {
642
+ "markdown_report_policy": "warning_only_consolidated_readme",
643
+ "missing_advisor_contract": [
644
+ "config.yaml",
645
+ "report.md"
646
+ ],
647
+ "missing_required": [],
648
+ "run_dir": "runs/ctt_base_context_obs_val_envclip_rollout_comparison",
649
+ "status": "pass"
650
+ },
651
+ {
652
+ "markdown_report_policy": "warning_only_consolidated_readme",
653
+ "missing_advisor_contract": [
654
+ "config.yaml",
655
+ "report.md"
656
+ ],
657
+ "missing_required": [],
658
+ "run_dir": "runs/ctt_base_context_obs_val_noclip_rollout_comparison",
659
+ "status": "pass"
660
+ },
661
+ {
662
+ "markdown_report_policy": "warning_only_consolidated_readme",
663
+ "missing_advisor_contract": [
664
+ "config.yaml",
665
+ "report.md"
666
+ ],
667
+ "missing_required": [],
668
+ "run_dir": "runs/ctt_base_context_obs_val_perdim_trainmax_rollout_comparison",
669
+ "status": "pass"
670
+ },
671
+ {
672
+ "markdown_report_policy": "warning_only_consolidated_readme",
673
+ "missing_advisor_contract": [
674
+ "config.yaml",
675
+ "report.md"
676
+ ],
677
+ "missing_required": [],
678
+ "run_dir": "runs/ctt_base_context_obs_val_rollout_comparison",
679
+ "status": "pass"
680
+ },
681
+ {
682
+ "markdown_report_policy": "warning_only_consolidated_readme",
683
+ "missing_advisor_contract": [
684
+ "config.yaml",
685
+ "report.md"
686
+ ],
687
+ "missing_required": [],
688
+ "run_dir": "runs/ctt_base_context_obs_val_scaled0215_rollout_comparison",
689
+ "status": "pass"
690
+ },
691
+ {
692
+ "markdown_report_policy": "warning_only_consolidated_readme",
693
+ "missing_advisor_contract": [
694
+ "config.yaml",
695
+ "report.md"
696
+ ],
697
+ "missing_required": [],
698
+ "run_dir": "runs/ctt_base_context_obs_val_tanh_rollout_comparison",
699
+ "status": "pass"
700
+ },
701
+ {
702
+ "markdown_report_policy": "warning_only_consolidated_readme",
703
+ "missing_advisor_contract": [
704
+ "report.md"
705
+ ],
706
+ "missing_required": [],
707
+ "run_dir": "runs/ctt_dominance_val_to_test",
708
+ "status": "pass"
709
+ },
710
+ {
711
+ "markdown_report_policy": "warning_only_consolidated_readme",
712
+ "missing_advisor_contract": [
713
+ "report.md"
714
+ ],
715
+ "missing_required": [],
716
+ "run_dir": "runs/ctt_learned_dominance_val_to_test",
717
+ "status": "pass"
718
+ },
719
+ {
720
+ "markdown_report_policy": "warning_only_consolidated_readme",
721
+ "missing_advisor_contract": [
722
+ "report.md"
723
+ ],
724
+ "missing_required": [],
725
+ "run_dir": "runs/ctt_residual_smoke_proxy",
726
+ "status": "pass"
727
+ },
728
+ {
729
+ "markdown_report_policy": "warning_only_consolidated_readme",
730
+ "missing_advisor_contract": [
731
+ "config.yaml",
732
+ "report.md"
733
+ ],
734
+ "missing_required": [],
735
+ "run_dir": "runs/ctt_test_rollout_comparison",
736
+ "status": "pass"
737
+ },
738
+ {
739
+ "markdown_report_policy": "warning_only_consolidated_readme",
740
+ "missing_advisor_contract": [
741
+ "report.md"
742
+ ],
743
+ "missing_required": [],
744
+ "run_dir": "runs/ctt_val_proxy_comparison",
745
+ "status": "pass"
746
+ },
747
+ {
748
+ "markdown_report_policy": "warning_only_consolidated_readme",
749
+ "missing_advisor_contract": [
750
+ "config.yaml",
751
+ "report.md"
752
+ ],
753
+ "missing_required": [],
754
+ "run_dir": "runs/ctt_val_rollout_comparison",
755
+ "status": "pass"
756
+ },
757
+ {
758
+ "markdown_report_policy": "warning_only_consolidated_readme",
759
+ "missing_advisor_contract": [
760
+ "config.yaml",
761
+ "train.log",
762
+ "eval.log",
763
+ "metrics_by_task.json",
764
+ "metrics_by_seed.json",
765
+ "report.md"
766
+ ],
767
+ "missing_required": [],
768
+ "run_dir": "runs/data_accounting",
769
+ "status": "pass"
770
+ },
771
+ {
772
+ "markdown_report_policy": "warning_only_consolidated_readme",
773
+ "missing_advisor_contract": [
774
+ "train.log",
775
+ "eval.log",
776
+ "metrics_by_task.json",
777
+ "metrics_by_seed.json",
778
+ "report.md"
779
+ ],
780
+ "missing_required": [],
781
+ "run_dir": "runs/paper_ctt_audit",
782
+ "status": "pass"
783
+ }
784
+ ],
785
+ "schema_version": 1,
786
+ "summary": {
787
+ "num_failures": 0,
788
+ "num_forbidden_matches": 0,
789
+ "num_paper_inputs": 37,
790
+ "num_run_dirs_in_paper": 32,
791
+ "num_warnings": 59,
792
+ "status": "pass"
793
+ }
794
+ }
workspace/runs/paper_ctt_audit/command.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ python scripts/audit_ctt_paper_artifacts.py --repo-root . --paper latex/main.tex --out-dir runs/paper_ctt_audit
workspace/runs/paper_ctt_audit/config.yaml ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ paper: latex/main.tex
2
+ markdown_policy: consolidated_readme_only
3
+ strict_report_md: false
workspace/runs/paper_ctt_audit/data_hash.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ 9c456e02ceffa581945d3d32e68a4c6b4a1f52d27d07a762d38ae3edbc5b536b
workspace/runs/paper_ctt_audit/git_hash.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ 14e81292e2004925f81e0bb4810f6f144bfd87b7
workspace/runs/paper_ctt_audit/metrics.json ADDED
@@ -0,0 +1,794 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "audit_policy": {
3
+ "markdown_note": "Advisor checklist asks for report.md files, but the current workspace policy keeps README.md as the only Markdown document. Missing report.md is recorded as a warning, not regenerated here.",
4
+ "markdown_policy": "consolidated_readme_only"
5
+ },
6
+ "forbidden_patterns": [],
7
+ "paper": "latex/main.tex",
8
+ "paper_inputs": {
9
+ "num_inputs": 37,
10
+ "num_missing": 0,
11
+ "rows": [
12
+ {
13
+ "line": 216,
14
+ "raw": "../runs/data_accounting/table",
15
+ "resolved": "runs/data_accounting/table.tex",
16
+ "run_dir": "runs/data_accounting",
17
+ "status": "pass"
18
+ },
19
+ {
20
+ "line": 363,
21
+ "raw": "../paper/sections/theory",
22
+ "resolved": "paper/sections/theory.tex",
23
+ "status": "pass"
24
+ },
25
+ {
26
+ "line": 372,
27
+ "raw": "tables/main_results",
28
+ "resolved": "latex/tables/main_results.tex",
29
+ "status": "pass"
30
+ },
31
+ {
32
+ "line": 373,
33
+ "raw": "tables/car_decomposition",
34
+ "resolved": "latex/tables/car_decomposition.tex",
35
+ "status": "pass"
36
+ },
37
+ {
38
+ "line": 381,
39
+ "raw": "tables/source_score_sweep",
40
+ "resolved": "latex/tables/source_score_sweep.tex",
41
+ "status": "pass"
42
+ },
43
+ {
44
+ "line": 382,
45
+ "raw": "tables/selector_calibration",
46
+ "resolved": "latex/tables/selector_calibration.tex",
47
+ "status": "pass"
48
+ },
49
+ {
50
+ "line": 411,
51
+ "raw": "../runs/ctt_residual_smoke_proxy/table",
52
+ "resolved": "runs/ctt_residual_smoke_proxy/table.tex",
53
+ "run_dir": "runs/ctt_residual_smoke_proxy",
54
+ "status": "pass"
55
+ },
56
+ {
57
+ "line": 439,
58
+ "raw": "../runs/ctt_val_proxy_comparison/table",
59
+ "resolved": "runs/ctt_val_proxy_comparison/table.tex",
60
+ "run_dir": "runs/ctt_val_proxy_comparison",
61
+ "status": "pass"
62
+ },
63
+ {
64
+ "line": 491,
65
+ "raw": "../runs/action_bound_audit_rgb_refs/table",
66
+ "resolved": "runs/action_bound_audit_rgb_refs/table.tex",
67
+ "run_dir": "runs/action_bound_audit_rgb_refs",
68
+ "status": "pass"
69
+ },
70
+ {
71
+ "line": 536,
72
+ "raw": "../runs/ctt_base_context_obs_val_noclip_rollout_comparison/table",
73
+ "resolved": "runs/ctt_base_context_obs_val_noclip_rollout_comparison/table.tex",
74
+ "run_dir": "runs/ctt_base_context_obs_val_noclip_rollout_comparison",
75
+ "status": "pass"
76
+ },
77
+ {
78
+ "line": 547,
79
+ "raw": "../runs/ctt_base_context_obs_test_noclip_rollout_comparison/table",
80
+ "resolved": "runs/ctt_base_context_obs_test_noclip_rollout_comparison/table.tex",
81
+ "run_dir": "runs/ctt_base_context_obs_test_noclip_rollout_comparison",
82
+ "status": "pass"
83
+ },
84
+ {
85
+ "line": 568,
86
+ "raw": "../runs/ctt_base_context_obs_val_scaled0215_rollout_comparison/table",
87
+ "resolved": "runs/ctt_base_context_obs_val_scaled0215_rollout_comparison/table.tex",
88
+ "run_dir": "runs/ctt_base_context_obs_val_scaled0215_rollout_comparison",
89
+ "status": "pass"
90
+ },
91
+ {
92
+ "line": 579,
93
+ "raw": "../runs/ctt_base_context_obs_test_scaled0215_rollout_comparison/table",
94
+ "resolved": "runs/ctt_base_context_obs_test_scaled0215_rollout_comparison/table.tex",
95
+ "run_dir": "runs/ctt_base_context_obs_test_scaled0215_rollout_comparison",
96
+ "status": "pass"
97
+ },
98
+ {
99
+ "line": 598,
100
+ "raw": "../runs/ctt_base_context_obs_val_perdim_trainmax_rollout_comparison/table",
101
+ "resolved": "runs/ctt_base_context_obs_val_perdim_trainmax_rollout_comparison/table.tex",
102
+ "run_dir": "runs/ctt_base_context_obs_val_perdim_trainmax_rollout_comparison",
103
+ "status": "pass"
104
+ },
105
+ {
106
+ "line": 609,
107
+ "raw": "../runs/ctt_base_context_obs_test_perdim_trainmax_rollout_comparison/table",
108
+ "resolved": "runs/ctt_base_context_obs_test_perdim_trainmax_rollout_comparison/table.tex",
109
+ "run_dir": "runs/ctt_base_context_obs_test_perdim_trainmax_rollout_comparison",
110
+ "status": "pass"
111
+ },
112
+ {
113
+ "line": 631,
114
+ "raw": "../runs/ctt_base_context_obs_val_envclip_rollout_comparison/table",
115
+ "resolved": "runs/ctt_base_context_obs_val_envclip_rollout_comparison/table.tex",
116
+ "run_dir": "runs/ctt_base_context_obs_val_envclip_rollout_comparison",
117
+ "status": "pass"
118
+ },
119
+ {
120
+ "line": 642,
121
+ "raw": "../runs/ctt_base_context_obs_test_envclip_rollout_comparison/table",
122
+ "resolved": "runs/ctt_base_context_obs_test_envclip_rollout_comparison/table.tex",
123
+ "run_dir": "runs/ctt_base_context_obs_test_envclip_rollout_comparison",
124
+ "status": "pass"
125
+ },
126
+ {
127
+ "line": 663,
128
+ "raw": "../runs/ctt_base_context_obs_val_tanh_rollout_comparison/table",
129
+ "resolved": "runs/ctt_base_context_obs_val_tanh_rollout_comparison/table.tex",
130
+ "run_dir": "runs/ctt_base_context_obs_val_tanh_rollout_comparison",
131
+ "status": "pass"
132
+ },
133
+ {
134
+ "line": 673,
135
+ "raw": "../runs/ctt_base_context_obs_test_tanh_rollout_comparison/table",
136
+ "resolved": "runs/ctt_base_context_obs_test_tanh_rollout_comparison/table.tex",
137
+ "run_dir": "runs/ctt_base_context_obs_test_tanh_rollout_comparison",
138
+ "status": "pass"
139
+ },
140
+ {
141
+ "line": 694,
142
+ "raw": "../runs/ctt_val_rollout_comparison/table",
143
+ "resolved": "runs/ctt_val_rollout_comparison/table.tex",
144
+ "run_dir": "runs/ctt_val_rollout_comparison",
145
+ "status": "pass"
146
+ },
147
+ {
148
+ "line": 715,
149
+ "raw": "../runs/ctt_base_context_obs_val_rollout_comparison/table",
150
+ "resolved": "runs/ctt_base_context_obs_val_rollout_comparison/table.tex",
151
+ "run_dir": "runs/ctt_base_context_obs_val_rollout_comparison",
152
+ "status": "pass"
153
+ },
154
+ {
155
+ "line": 740,
156
+ "raw": "../runs/ctt_base_context_obj_val_rollout_comparison/table",
157
+ "resolved": "runs/ctt_base_context_obj_val_rollout_comparison/table.tex",
158
+ "run_dir": "runs/ctt_base_context_obj_val_rollout_comparison",
159
+ "status": "pass"
160
+ },
161
+ {
162
+ "line": 758,
163
+ "raw": "../runs/ctt_test_rollout_comparison/table",
164
+ "resolved": "runs/ctt_test_rollout_comparison/table.tex",
165
+ "run_dir": "runs/ctt_test_rollout_comparison",
166
+ "status": "pass"
167
+ },
168
+ {
169
+ "line": 780,
170
+ "raw": "../runs/ctt_base_context_obs_test_rollout_comparison/table",
171
+ "resolved": "runs/ctt_base_context_obs_test_rollout_comparison/table.tex",
172
+ "run_dir": "runs/ctt_base_context_obs_test_rollout_comparison",
173
+ "status": "pass"
174
+ },
175
+ {
176
+ "line": 798,
177
+ "raw": "../runs/ctt_dominance_val_to_test/table",
178
+ "resolved": "runs/ctt_dominance_val_to_test/table.tex",
179
+ "run_dir": "runs/ctt_dominance_val_to_test",
180
+ "status": "pass"
181
+ },
182
+ {
183
+ "line": 821,
184
+ "raw": "../runs/ctt_learned_dominance_val_to_test/table",
185
+ "resolved": "runs/ctt_learned_dominance_val_to_test/table.tex",
186
+ "run_dir": "runs/ctt_learned_dominance_val_to_test",
187
+ "status": "pass"
188
+ },
189
+ {
190
+ "line": 853,
191
+ "raw": "../runs/ctt_base_context_obs_learned_dominance_context_val_to_test/table",
192
+ "resolved": "runs/ctt_base_context_obs_learned_dominance_context_val_to_test/table.tex",
193
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_context_val_to_test",
194
+ "status": "pass"
195
+ },
196
+ {
197
+ "line": 873,
198
+ "raw": "../runs/ctt_base_context_obs_learned_dominance_context_tangent_success_tanh_val_to_test/table",
199
+ "resolved": "runs/ctt_base_context_obs_learned_dominance_context_tangent_success_tanh_val_to_test/table.tex",
200
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_context_tangent_success_tanh_val_to_test",
201
+ "status": "pass"
202
+ },
203
+ {
204
+ "line": 892,
205
+ "raw": "../runs/ctt_base_context_obs_learned_dominance_context_success_tanh_train_to_test/table",
206
+ "resolved": "runs/ctt_base_context_obs_learned_dominance_context_success_tanh_train_to_test/table.tex",
207
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_context_success_tanh_train_to_test",
208
+ "status": "pass"
209
+ },
210
+ {
211
+ "line": 914,
212
+ "raw": "../runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/table",
213
+ "resolved": "runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test/table.tex",
214
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test",
215
+ "status": "pass"
216
+ },
217
+ {
218
+ "line": 939,
219
+ "raw": "../runs/ctt_base_context_obs_learned_dominance_source_envclip_train_to_test/table",
220
+ "resolved": "runs/ctt_base_context_obs_learned_dominance_source_envclip_train_to_test/table.tex",
221
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_source_envclip_train_to_test",
222
+ "status": "pass"
223
+ },
224
+ {
225
+ "line": 975,
226
+ "raw": "../runs/ctt_base_context_obs_test_envclip_k16_rollout_comparison/table",
227
+ "resolved": "runs/ctt_base_context_obs_test_envclip_k16_rollout_comparison/table.tex",
228
+ "run_dir": "runs/ctt_base_context_obs_test_envclip_k16_rollout_comparison",
229
+ "status": "pass"
230
+ },
231
+ {
232
+ "line": 996,
233
+ "raw": "../runs/ctt_base_context_obs_dominance_envclip_k16_train_to_test/table",
234
+ "resolved": "runs/ctt_base_context_obs_dominance_envclip_k16_train_to_test/table.tex",
235
+ "run_dir": "runs/ctt_base_context_obs_dominance_envclip_k16_train_to_test",
236
+ "status": "pass"
237
+ },
238
+ {
239
+ "line": 1020,
240
+ "raw": "../runs/ctt_base_context_obs_learned_dominance_chartcompat_obs_utility_task_envclip_k16_train_to_test/table",
241
+ "resolved": "runs/ctt_base_context_obs_learned_dominance_chartcompat_obs_utility_task_envclip_k16_train_to_test/table.tex",
242
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_chartcompat_obs_utility_task_envclip_k16_train_to_test",
243
+ "status": "pass"
244
+ },
245
+ {
246
+ "line": 1077,
247
+ "raw": "../runs/ctt_base_context_obs_learned_dominance_train_to_test/table",
248
+ "resolved": "runs/ctt_base_context_obs_learned_dominance_train_to_test/table.tex",
249
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_train_to_test",
250
+ "status": "pass"
251
+ },
252
+ {
253
+ "line": 1101,
254
+ "raw": "../runs/ctt_base_context_obs_nonlinear_dominance_basic_positive_train_to_test/table.tex",
255
+ "resolved": "runs/ctt_base_context_obs_nonlinear_dominance_basic_positive_train_to_test/table.tex",
256
+ "run_dir": "runs/ctt_base_context_obs_nonlinear_dominance_basic_positive_train_to_test",
257
+ "status": "pass"
258
+ },
259
+ {
260
+ "line": 1204,
261
+ "raw": "../runs/paper_ctt_audit/table",
262
+ "resolved": "runs/paper_ctt_audit/table.tex",
263
+ "run_dir": "runs/paper_ctt_audit",
264
+ "status": "pass"
265
+ }
266
+ ],
267
+ "run_dirs": [
268
+ "runs/action_bound_audit_rgb_refs",
269
+ "runs/ctt_base_context_obj_val_rollout_comparison",
270
+ "runs/ctt_base_context_obs_dominance_envclip_k16_train_to_test",
271
+ "runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test",
272
+ "runs/ctt_base_context_obs_learned_dominance_chartcompat_obs_utility_task_envclip_k16_train_to_test",
273
+ "runs/ctt_base_context_obs_learned_dominance_context_success_tanh_train_to_test",
274
+ "runs/ctt_base_context_obs_learned_dominance_context_tangent_success_tanh_val_to_test",
275
+ "runs/ctt_base_context_obs_learned_dominance_context_val_to_test",
276
+ "runs/ctt_base_context_obs_learned_dominance_source_envclip_train_to_test",
277
+ "runs/ctt_base_context_obs_learned_dominance_train_to_test",
278
+ "runs/ctt_base_context_obs_nonlinear_dominance_basic_positive_train_to_test",
279
+ "runs/ctt_base_context_obs_test_envclip_k16_rollout_comparison",
280
+ "runs/ctt_base_context_obs_test_envclip_rollout_comparison",
281
+ "runs/ctt_base_context_obs_test_noclip_rollout_comparison",
282
+ "runs/ctt_base_context_obs_test_perdim_trainmax_rollout_comparison",
283
+ "runs/ctt_base_context_obs_test_rollout_comparison",
284
+ "runs/ctt_base_context_obs_test_scaled0215_rollout_comparison",
285
+ "runs/ctt_base_context_obs_test_tanh_rollout_comparison",
286
+ "runs/ctt_base_context_obs_val_envclip_rollout_comparison",
287
+ "runs/ctt_base_context_obs_val_noclip_rollout_comparison",
288
+ "runs/ctt_base_context_obs_val_perdim_trainmax_rollout_comparison",
289
+ "runs/ctt_base_context_obs_val_rollout_comparison",
290
+ "runs/ctt_base_context_obs_val_scaled0215_rollout_comparison",
291
+ "runs/ctt_base_context_obs_val_tanh_rollout_comparison",
292
+ "runs/ctt_dominance_val_to_test",
293
+ "runs/ctt_learned_dominance_val_to_test",
294
+ "runs/ctt_residual_smoke_proxy",
295
+ "runs/ctt_test_rollout_comparison",
296
+ "runs/ctt_val_proxy_comparison",
297
+ "runs/ctt_val_rollout_comparison",
298
+ "runs/data_accounting",
299
+ "runs/paper_ctt_audit"
300
+ ]
301
+ },
302
+ "required_paths": [
303
+ {
304
+ "detail": "Canonical measured/proxy metrics.",
305
+ "name": "metrics_module",
306
+ "path": "cil/metrics.py",
307
+ "status": "pass"
308
+ },
309
+ {
310
+ "detail": "JSON/TeX metric export and proxy guards.",
311
+ "name": "metrics_eval",
312
+ "path": "scripts/eval_metrics.py",
313
+ "status": "pass"
314
+ },
315
+ {
316
+ "detail": "Regression tests for metric separation.",
317
+ "name": "metrics_tests",
318
+ "path": "tests/test_metrics.py",
319
+ "status": "pass"
320
+ },
321
+ {
322
+ "detail": "Chart database export.",
323
+ "name": "chart_export",
324
+ "path": "scripts/export_cil_charts.py",
325
+ "status": "pass"
326
+ },
327
+ {
328
+ "detail": "Chart leakage audit.",
329
+ "name": "chart_audit",
330
+ "path": "scripts/audit_cil_charts.py",
331
+ "status": "pass"
332
+ },
333
+ {
334
+ "detail": "Scripted data accounting.",
335
+ "name": "data_accounting",
336
+ "path": "runs/data_accounting/table.json",
337
+ "status": "pass"
338
+ },
339
+ {
340
+ "detail": "Data accounting table.",
341
+ "name": "data_accounting_table",
342
+ "path": "runs/data_accounting/table.tex",
343
+ "status": "pass"
344
+ },
345
+ {
346
+ "detail": "Leakage audit artifact.",
347
+ "name": "leakage_audit",
348
+ "path": "runs/leakage_audit/report.json",
349
+ "status": "pass"
350
+ },
351
+ {
352
+ "detail": "Causal Tangent Transport module.",
353
+ "name": "ctt_model",
354
+ "path": "cil/models/ctt.py",
355
+ "status": "pass"
356
+ },
357
+ {
358
+ "detail": "Tangent-code encoder/decoder helpers.",
359
+ "name": "tangent_encoder",
360
+ "path": "cil/models/tangent_encoder.py",
361
+ "status": "pass"
362
+ },
363
+ {
364
+ "detail": "Chart encoder.",
365
+ "name": "chart_encoder",
366
+ "path": "cil/models/chart_encoder.py",
367
+ "status": "pass"
368
+ },
369
+ {
370
+ "detail": "Utility energy scorer.",
371
+ "name": "utility_energy",
372
+ "path": "cil/models/utility_energy.py",
373
+ "status": "pass"
374
+ },
375
+ {
376
+ "detail": "CTT training script.",
377
+ "name": "ctt_train",
378
+ "path": "scripts/train_ctt.py",
379
+ "status": "pass"
380
+ },
381
+ {
382
+ "detail": "Proxy support evaluation.",
383
+ "name": "ctt_proxy_eval",
384
+ "path": "scripts/eval_ctt_proxy.py",
385
+ "status": "pass"
386
+ },
387
+ {
388
+ "detail": "Measured rollout evaluation.",
389
+ "name": "ctt_rollout_eval",
390
+ "path": "scripts/eval_ctt_rollout.py",
391
+ "status": "pass"
392
+ },
393
+ {
394
+ "detail": "Utility energy training.",
395
+ "name": "utility_train",
396
+ "path": "scripts/train_utility_energy.py",
397
+ "status": "pass"
398
+ },
399
+ {
400
+ "detail": "Calibrated dominance rule.",
401
+ "name": "dominance_calibration",
402
+ "path": "scripts/calibrate_dominance.py",
403
+ "status": "pass"
404
+ },
405
+ {
406
+ "detail": "Theory section included by paper.",
407
+ "name": "theory_tex",
408
+ "path": "paper/sections/theory.tex",
409
+ "status": "pass"
410
+ },
411
+ {
412
+ "detail": "Compiled paper PDF.",
413
+ "name": "paper_pdf",
414
+ "path": "latex/main.pdf",
415
+ "status": "pass"
416
+ },
417
+ {
418
+ "detail": "CTT loss weights and variants are config-driven.",
419
+ "files": [
420
+ "configs/ctt/gated_residual_full.yaml",
421
+ "configs/ctt/gated_residual_smoke.yaml",
422
+ "configs/ctt/residual_full.yaml",
423
+ "configs/ctt/residual_smoke.yaml"
424
+ ],
425
+ "name": "ctt_configs",
426
+ "path": "configs/ctt/*.yaml",
427
+ "status": "pass"
428
+ }
429
+ ],
430
+ "required_phrases": [
431
+ {
432
+ "detail": "Paper title/introduction names CTT as the central method.",
433
+ "name": "title_ctt",
434
+ "needle": "Causal Tangent Transport",
435
+ "status": "pass"
436
+ },
437
+ {
438
+ "detail": "Paper defines train-positive source-to-target transport.",
439
+ "name": "transport_operator",
440
+ "needle": "T_{\\phi}(z_s,z_t,\\xi_s^+)",
441
+ "status": "pass"
442
+ },
443
+ {
444
+ "detail": "Paper names measured rollout positive-tangent recall separately.",
445
+ "name": "outcome_ptr",
446
+ "needle": "OutcomePTR",
447
+ "status": "pass"
448
+ },
449
+ {
450
+ "detail": "Paper names distance-only Proxy Positive Tangent Coverage.",
451
+ "name": "pptc",
452
+ "needle": "PPTC",
453
+ "status": "pass"
454
+ },
455
+ {
456
+ "detail": "Paper exposes the support part of CAR.",
457
+ "name": "support_gap",
458
+ "needle": "SupportGap",
459
+ "status": "pass"
460
+ },
461
+ {
462
+ "detail": "Paper exposes the selector part of CAR.",
463
+ "name": "selector_gap",
464
+ "needle": "SelectorGap",
465
+ "status": "pass"
466
+ }
467
+ ],
468
+ "run_artifacts": [
469
+ {
470
+ "markdown_report_policy": "warning_only_consolidated_readme",
471
+ "missing_advisor_contract": [
472
+ "metrics_by_task.json",
473
+ "metrics_by_seed.json",
474
+ "report.md"
475
+ ],
476
+ "missing_required": [],
477
+ "run_dir": "runs/action_bound_audit_rgb_refs",
478
+ "status": "pass"
479
+ },
480
+ {
481
+ "markdown_report_policy": "warning_only_consolidated_readme",
482
+ "missing_advisor_contract": [
483
+ "config.yaml",
484
+ "report.md"
485
+ ],
486
+ "missing_required": [],
487
+ "run_dir": "runs/ctt_base_context_obj_val_rollout_comparison",
488
+ "status": "pass"
489
+ },
490
+ {
491
+ "markdown_report_policy": "warning_only_consolidated_readme",
492
+ "missing_advisor_contract": [
493
+ "report.md"
494
+ ],
495
+ "missing_required": [],
496
+ "run_dir": "runs/ctt_base_context_obs_dominance_envclip_k16_train_to_test",
497
+ "status": "pass"
498
+ },
499
+ {
500
+ "markdown_report_policy": "warning_only_consolidated_readme",
501
+ "missing_advisor_contract": [
502
+ "report.md"
503
+ ],
504
+ "missing_required": [],
505
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_basic_envclip_train_to_test",
506
+ "status": "pass"
507
+ },
508
+ {
509
+ "markdown_report_policy": "warning_only_consolidated_readme",
510
+ "missing_advisor_contract": [
511
+ "report.md"
512
+ ],
513
+ "missing_required": [],
514
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_chartcompat_obs_utility_task_envclip_k16_train_to_test",
515
+ "status": "pass"
516
+ },
517
+ {
518
+ "markdown_report_policy": "warning_only_consolidated_readme",
519
+ "missing_advisor_contract": [
520
+ "report.md"
521
+ ],
522
+ "missing_required": [],
523
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_context_success_tanh_train_to_test",
524
+ "status": "pass"
525
+ },
526
+ {
527
+ "markdown_report_policy": "warning_only_consolidated_readme",
528
+ "missing_advisor_contract": [
529
+ "report.md"
530
+ ],
531
+ "missing_required": [],
532
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_context_tangent_success_tanh_val_to_test",
533
+ "status": "pass"
534
+ },
535
+ {
536
+ "markdown_report_policy": "warning_only_consolidated_readme",
537
+ "missing_advisor_contract": [
538
+ "report.md"
539
+ ],
540
+ "missing_required": [],
541
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_context_val_to_test",
542
+ "status": "pass"
543
+ },
544
+ {
545
+ "markdown_report_policy": "warning_only_consolidated_readme",
546
+ "missing_advisor_contract": [
547
+ "report.md"
548
+ ],
549
+ "missing_required": [],
550
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_source_envclip_train_to_test",
551
+ "status": "pass"
552
+ },
553
+ {
554
+ "markdown_report_policy": "warning_only_consolidated_readme",
555
+ "missing_advisor_contract": [
556
+ "report.md"
557
+ ],
558
+ "missing_required": [],
559
+ "run_dir": "runs/ctt_base_context_obs_learned_dominance_train_to_test",
560
+ "status": "pass"
561
+ },
562
+ {
563
+ "markdown_report_policy": "warning_only_consolidated_readme",
564
+ "missing_advisor_contract": [
565
+ "report.md"
566
+ ],
567
+ "missing_required": [],
568
+ "run_dir": "runs/ctt_base_context_obs_nonlinear_dominance_basic_positive_train_to_test",
569
+ "status": "pass"
570
+ },
571
+ {
572
+ "markdown_report_policy": "warning_only_consolidated_readme",
573
+ "missing_advisor_contract": [
574
+ "config.yaml",
575
+ "report.md"
576
+ ],
577
+ "missing_required": [],
578
+ "run_dir": "runs/ctt_base_context_obs_test_envclip_k16_rollout_comparison",
579
+ "status": "pass"
580
+ },
581
+ {
582
+ "markdown_report_policy": "warning_only_consolidated_readme",
583
+ "missing_advisor_contract": [
584
+ "config.yaml",
585
+ "report.md"
586
+ ],
587
+ "missing_required": [],
588
+ "run_dir": "runs/ctt_base_context_obs_test_envclip_rollout_comparison",
589
+ "status": "pass"
590
+ },
591
+ {
592
+ "markdown_report_policy": "warning_only_consolidated_readme",
593
+ "missing_advisor_contract": [
594
+ "config.yaml",
595
+ "report.md"
596
+ ],
597
+ "missing_required": [],
598
+ "run_dir": "runs/ctt_base_context_obs_test_noclip_rollout_comparison",
599
+ "status": "pass"
600
+ },
601
+ {
602
+ "markdown_report_policy": "warning_only_consolidated_readme",
603
+ "missing_advisor_contract": [
604
+ "config.yaml",
605
+ "report.md"
606
+ ],
607
+ "missing_required": [],
608
+ "run_dir": "runs/ctt_base_context_obs_test_perdim_trainmax_rollout_comparison",
609
+ "status": "pass"
610
+ },
611
+ {
612
+ "markdown_report_policy": "warning_only_consolidated_readme",
613
+ "missing_advisor_contract": [
614
+ "config.yaml",
615
+ "report.md"
616
+ ],
617
+ "missing_required": [],
618
+ "run_dir": "runs/ctt_base_context_obs_test_rollout_comparison",
619
+ "status": "pass"
620
+ },
621
+ {
622
+ "markdown_report_policy": "warning_only_consolidated_readme",
623
+ "missing_advisor_contract": [
624
+ "config.yaml",
625
+ "report.md"
626
+ ],
627
+ "missing_required": [],
628
+ "run_dir": "runs/ctt_base_context_obs_test_scaled0215_rollout_comparison",
629
+ "status": "pass"
630
+ },
631
+ {
632
+ "markdown_report_policy": "warning_only_consolidated_readme",
633
+ "missing_advisor_contract": [
634
+ "config.yaml",
635
+ "report.md"
636
+ ],
637
+ "missing_required": [],
638
+ "run_dir": "runs/ctt_base_context_obs_test_tanh_rollout_comparison",
639
+ "status": "pass"
640
+ },
641
+ {
642
+ "markdown_report_policy": "warning_only_consolidated_readme",
643
+ "missing_advisor_contract": [
644
+ "config.yaml",
645
+ "report.md"
646
+ ],
647
+ "missing_required": [],
648
+ "run_dir": "runs/ctt_base_context_obs_val_envclip_rollout_comparison",
649
+ "status": "pass"
650
+ },
651
+ {
652
+ "markdown_report_policy": "warning_only_consolidated_readme",
653
+ "missing_advisor_contract": [
654
+ "config.yaml",
655
+ "report.md"
656
+ ],
657
+ "missing_required": [],
658
+ "run_dir": "runs/ctt_base_context_obs_val_noclip_rollout_comparison",
659
+ "status": "pass"
660
+ },
661
+ {
662
+ "markdown_report_policy": "warning_only_consolidated_readme",
663
+ "missing_advisor_contract": [
664
+ "config.yaml",
665
+ "report.md"
666
+ ],
667
+ "missing_required": [],
668
+ "run_dir": "runs/ctt_base_context_obs_val_perdim_trainmax_rollout_comparison",
669
+ "status": "pass"
670
+ },
671
+ {
672
+ "markdown_report_policy": "warning_only_consolidated_readme",
673
+ "missing_advisor_contract": [
674
+ "config.yaml",
675
+ "report.md"
676
+ ],
677
+ "missing_required": [],
678
+ "run_dir": "runs/ctt_base_context_obs_val_rollout_comparison",
679
+ "status": "pass"
680
+ },
681
+ {
682
+ "markdown_report_policy": "warning_only_consolidated_readme",
683
+ "missing_advisor_contract": [
684
+ "config.yaml",
685
+ "report.md"
686
+ ],
687
+ "missing_required": [],
688
+ "run_dir": "runs/ctt_base_context_obs_val_scaled0215_rollout_comparison",
689
+ "status": "pass"
690
+ },
691
+ {
692
+ "markdown_report_policy": "warning_only_consolidated_readme",
693
+ "missing_advisor_contract": [
694
+ "config.yaml",
695
+ "report.md"
696
+ ],
697
+ "missing_required": [],
698
+ "run_dir": "runs/ctt_base_context_obs_val_tanh_rollout_comparison",
699
+ "status": "pass"
700
+ },
701
+ {
702
+ "markdown_report_policy": "warning_only_consolidated_readme",
703
+ "missing_advisor_contract": [
704
+ "report.md"
705
+ ],
706
+ "missing_required": [],
707
+ "run_dir": "runs/ctt_dominance_val_to_test",
708
+ "status": "pass"
709
+ },
710
+ {
711
+ "markdown_report_policy": "warning_only_consolidated_readme",
712
+ "missing_advisor_contract": [
713
+ "report.md"
714
+ ],
715
+ "missing_required": [],
716
+ "run_dir": "runs/ctt_learned_dominance_val_to_test",
717
+ "status": "pass"
718
+ },
719
+ {
720
+ "markdown_report_policy": "warning_only_consolidated_readme",
721
+ "missing_advisor_contract": [
722
+ "report.md"
723
+ ],
724
+ "missing_required": [],
725
+ "run_dir": "runs/ctt_residual_smoke_proxy",
726
+ "status": "pass"
727
+ },
728
+ {
729
+ "markdown_report_policy": "warning_only_consolidated_readme",
730
+ "missing_advisor_contract": [
731
+ "config.yaml",
732
+ "report.md"
733
+ ],
734
+ "missing_required": [],
735
+ "run_dir": "runs/ctt_test_rollout_comparison",
736
+ "status": "pass"
737
+ },
738
+ {
739
+ "markdown_report_policy": "warning_only_consolidated_readme",
740
+ "missing_advisor_contract": [
741
+ "report.md"
742
+ ],
743
+ "missing_required": [],
744
+ "run_dir": "runs/ctt_val_proxy_comparison",
745
+ "status": "pass"
746
+ },
747
+ {
748
+ "markdown_report_policy": "warning_only_consolidated_readme",
749
+ "missing_advisor_contract": [
750
+ "config.yaml",
751
+ "report.md"
752
+ ],
753
+ "missing_required": [],
754
+ "run_dir": "runs/ctt_val_rollout_comparison",
755
+ "status": "pass"
756
+ },
757
+ {
758
+ "markdown_report_policy": "warning_only_consolidated_readme",
759
+ "missing_advisor_contract": [
760
+ "config.yaml",
761
+ "train.log",
762
+ "eval.log",
763
+ "metrics_by_task.json",
764
+ "metrics_by_seed.json",
765
+ "report.md"
766
+ ],
767
+ "missing_required": [],
768
+ "run_dir": "runs/data_accounting",
769
+ "status": "pass"
770
+ },
771
+ {
772
+ "markdown_report_policy": "warning_only_consolidated_readme",
773
+ "missing_advisor_contract": [
774
+ "train.log",
775
+ "eval.log",
776
+ "metrics_by_task.json",
777
+ "metrics_by_seed.json",
778
+ "report.md"
779
+ ],
780
+ "missing_required": [],
781
+ "run_dir": "runs/paper_ctt_audit",
782
+ "status": "pass"
783
+ }
784
+ ],
785
+ "schema_version": 1,
786
+ "summary": {
787
+ "num_failures": 0,
788
+ "num_forbidden_matches": 0,
789
+ "num_paper_inputs": 37,
790
+ "num_run_dirs_in_paper": 32,
791
+ "num_warnings": 59,
792
+ "status": "pass"
793
+ }
794
+ }
workspace/runs/paper_ctt_audit/split_hash.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ aaa857b92cbe5d455009f3ca3380e8b124908aa6c4c99fc1da408f738ed2dcf6
workspace/runs/paper_ctt_audit/table.tex ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ % Auto-generated by scripts/audit_ctt_paper_artifacts.py
2
+ \begin{tabular}{lrr}
3
+ \toprule
4
+ Audit item & Count & Status \\
5
+ \midrule
6
+ Forbidden language matches & 0 & pass \\
7
+ Paper inputs & 37 & pass \\
8
+ Run dirs in paper & 32 & pass \\
9
+ Implementation paths & 20 & pass \\
10
+ Advisor-contract warnings & 59 & warn \\
11
+ \bottomrule
12
+ \end{tabular}