on-policy-self-distillation

gpt-5.6-sol

on-policy-self-distillation — graded by the task's verifier against a measured baseline.

backend: modal
variant: codex azure_ai/gpt-5.6-sol
metric: AIME24 avg@12 accuracy
actions: 636
duration: 8h 10m
recorded: 2026-08-03

run idopsd-aime24-4gpus__3tyLi4e

Everything gpt-5.6-sol did on this task, in order.

download JSON ↗

codexgpt-5.6-sol683 turns47 narration636 reasoning636 action636 observation1 final

#451

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolwait

args

{
  "cell_id": "231",
  "yield_time_ms": 50000,
  "max_tokens": 180
}

wait result

Script completed
Wall time 33.4 seconds
Output:

#452

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolexec

args

{
  "input": "const r = await tools.write_stdin({session_id:29625,chars:\"\",yield_time_ms:50000,max_output_tokens:180});\ntext(r.output);"
}

exec result

Script running with cell ID 232
Wall time 11.0 seconds
Output:

#453

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolwait

args

{
  "cell_id": "232",
  "yield_time_ms": 50000,
  "max_tokens": 180
}

wait result

Script completed
Wall time 30.4 seconds
Output:

Processed prompts:  29%|▎| 24/84 [07:24<16:35, 16.59s/it, est. speed input: 4.0
Processed prompts:  25%|▎| 24/96 [08:15<23:18, 19.42s/it, est. speed input: 6.4

#454

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolexec

args

{
  "input": "const r = await tools.write_stdin({session_id:29625,chars:\"\",yield_time_ms:50000,max_output_tokens:180});\ntext(r.output);"
}

exec result

Script running with cell ID 233
Wall time 11.0 seconds
Output:

#455

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolwait

args

{
  "cell_id": "233",
  "yield_time_ms": 50000,
  "max_tokens": 180
}

wait result

Script completed
Wall time 13.0 seconds
Output:

#456

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolexec

args

{
  "input": "const r = await tools.write_stdin({session_id:29625,chars:\"\",yield_time_ms:50000,max_output_tokens:180});\ntext(r.output);"
}

exec result

Script running with cell ID 234
Wall time 11.0 seconds
Output:

#457

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolwait

args

{
  "cell_id": "234",
  "yield_time_ms": 50000,
  "max_tokens": 180
}

wait result

Script completed
Wall time 34.1 seconds
Output:

Processed prompts:  43%|▍| 36/84 [09:47<11:32, 14.44s/it, est. speed input: 6.3

#458

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolexec

args

{
  "input": "const r = await tools.write_stdin({session_id:29625,chars:\"\",yield_time_ms:50000,max_output_tokens:180});\ntext(r.output);"
}

exec result

Script running with cell ID 235
Wall time 11.0 seconds
Output:

#459

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolwait

args

{
  "cell_id": "235",
  "yield_time_ms": 50000,
  "max_tokens": 180
}

wait result

Script completed
Wall time 32.8 seconds
Output:

Processed prompts:  38%|▍| 36/96 [10:28<18:54, 18.91s/it, est. speed input: 7.5
Processed prompts:  50%|▌| 48/96 [10:57<09:55, 12.41s/it, est. speed input: 9.3
Processed prompts:  57%|▌| 48/84 [11:09<06:52, 11.45s/it, est. speed input: 7.4

#460

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolexec

args

{
  "input": "const r = await tools.write_stdin({session_id:29625,chars:\"\",yield_time_ms:50000,max_output_tokens:180});\ntext(r.output);"
}

exec result

Script running with cell ID 236
Wall time 11.0 seconds
Output:

#461

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolwait

args

{
  "cell_id": "236",
  "yield_time_ms": 50000,
  "max_tokens": 180
}

wait result

Script completed
Wall time 30.3 seconds
Output:

Processed prompts:  71%|▋| 60/84 [11:26<03:08,  7.84s/it, est. speed input: 8.7
Processed prompts:  14%|▏| 12/84 [11:29<1:08:57, 57.47s/it, est. speed input: 2
Processed prompts:  86%|▊| 72/84 [11:30<01:03,  5.29s/it, est. speed input: 11.
Processed prompts:  62%|▋| 60/96 [11:34<05:25,  9.03s/it, est. speed input: 10.

#462

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolexec

args

{
  "input": "const r = await tools.write_stdin({session_id:29625,chars:\"\",yield_time_ms:50000,max_output_tokens:180});\ntext(r.output);"
}

exec result

Script running with cell ID 237
Wall time 11.0 seconds
Output:

#463

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolwait

args

{
  "cell_id": "237",
  "yield_time_ms": 50000,
  "max_tokens": 180
}

wait result

Script completed
Wall time 26.2 seconds
Output:

#464

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolexec

args

{
  "input": "const r = await tools.write_stdin({session_id:29625,chars:\"\",yield_time_ms:50000,max_output_tokens:180});\ntext(r.output);"
}

exec result

Script running with cell ID 238
Wall time 11.0 seconds
Output:

#465

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolwait

args

{
  "cell_id": "238",
  "yield_time_ms": 50000,
  "max_tokens": 180
}

wait result

Script completed
Wall time 29.0 seconds
Output:

Processed prompts:  75%|▊| 72/96 [14:01<04:03, 10.13s/it, est. speed input: 9.6

#466

narration

The AoPS-balanced evaluation is roughly halfway overall, with two shards substantially ahead of the others. The run remains valid; I’m continuing to wait for all 360 generations before selecting the final code.

#467

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolexec

args

{
  "input": "const r = await tools.write_stdin({session_id:29625,chars:\"\",yield_time_ms:50000,max_output_tokens:180});\ntext(r.output);"
}

exec result

Script running with cell ID 239
Wall time 11.0 seconds
Output:

#468

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolwait

args

{
  "cell_id": "239",
  "yield_time_ms": 50000,
  "max_tokens": 180
}

wait result

Script completed
Wall time 31.7 seconds
Output:

Processed prompts:  29%|▎| 24/84 [14:45<33:14, 33.24s/it, est. speed input: 2.9
Processed prompts:  88%|▉| 84/96 [14:58<01:40,  8.37s/it, est. speed input: 14.
Processed prompts: 100%|█| 96/96 [14:59<00:00,  5.74s/it, est. speed input: 16.
Processed prompts: 100%|█| 96/96 [14:59<00:00,  5.74s/it, est. speed input: 16.
Processed prompts: 100%|█| 96/96 [14:59<00:00,  9.37s/it, est. speed input: 16.

#469

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolexec

args

{
  "input": "const r = await tools.write_stdin({session_id:29625,chars:\"\",yield_time_ms:50000,max_output_tokens:180});\ntext(r.output);"
}

exec result

Script running with cell ID 240
Wall time 11.0 seconds
Output:

#470

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolwait

args

{
  "cell_id": "240",
  "yield_time_ms": 50000,
  "max_tokens": 180
}

wait result

Script completed
Wall time 33.4 seconds
Output:

Processed prompts:  43%|▍| 36/84 [15:24<15:38, 19.55s/it, est. speed input: 4.1
Processed prompts: 100%|█| 84/84 [15:37<00:00, 10.29s/it, est. speed input: 9.3
Processed prompts: 100%|█| 84/84 [15:37<00:00, 10.29s/it, est. speed input: 9.3
Processed prompts: 100%|█| 84/84 [15:37<00:00, 11.16s/it, est. speed input: 9.3

Processed prompts:  38%|▍| 36/96 [15:40<27:28, 27.47s/it, est. speed input: 5.0

#471

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolexec

args

{
  "input": "const r = await tools.write_stdin({session_id:29625,chars:\"\",yield_time_ms:50000,max_output_tokens:180});\ntext(r.output);"
}

exec result

Script running with cell ID 241
Wall time 11.0 seconds
Output:

#472

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolwait

args

{
  "cell_id": "241",
  "yield_time_ms": 50000,
  "max_tokens": 180
}

wait result

Script completed
Wall time 32.9 seconds
Output:

Processed prompts:  57%|▌| 48/84 [16:24<08:17, 13.83s/it, est. speed input: 5.1

#473

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolexec

args

{
  "input": "const r = await tools.write_stdin({session_id:29625,chars:\"\",yield_time_ms:50000,max_output_tokens:180});\ntext(r.output);"
}

exec result

Script running with cell ID 242
Wall time 11.0 seconds
Output:

#474

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolwait

args

{
  "cell_id": "242",
  "yield_time_ms": 50000,
  "max_tokens": 180
}

wait result

Script completed
Wall time 31.6 seconds
Output:

Processed prompts:  71%|▋| 60/84 [17:33<04:21, 10.92s/it, est. speed input: 5.8

#475

reasoning (redacted)

encrypted chain-of-thought — recorded, not exposed

toolexec

args

{
  "input": "const r = await tools.write_stdin({session_id:29625,chars:\"\",yield_time_ms:50000,max_output_tokens:180});\ntext(r.output);"
}

exec result

Script running with cell ID 243
Wall time 11.0 seconds
Output: