Model monitor
MiniMax M2.7 Reference
minimax/minimax-m2.7 · via openrouter
As of 2026-08-23: reference route · 418 complete panels attest measurement across 1 task families.
Reference route: this model is the co-probed confound control. It never opens a public incident — its record attests measurement, not verdicts.
Certified claim, per family: H0: E[panel_mean_t | prior completed panels] ≥ frozen_null_floor. Fixed panel composition does not assume or establish temporal independence. Recent panels below link to their hash-committed evidence.
Task family
Code reasoning
- Frozen null floor
- 0.700
- Anytime level α
- 0.050
- Completed panels
- 418
- CS lower
- 0.581
- CS upper
- 0.642
Limited history: the chart is capped; current counts and confidence bounds are not.
Panel record — 418 completed panels
| Cycle | Completed | Member scores | Mean | Evidence |
|---|---|---|---|---|
| 460 | 2026-08-23 13:42 UTC | 0.00 rate limited0.00 rate limited1.001.001.00 | 0.600 | 90f9b4e51693… |
| 459 | 2026-08-23 12:40 UTC | 1.000.00 rate limited0.00 rate limited1.001.00 | 0.600 | 6f63c10d8e5f… |
| 458 | 2026-08-23 11:40 UTC | 1.001.001.001.001.00 | 1.000 | a2ad9e95b30e… |
| 457 | 2026-08-23 10:39 UTC | 1.001.001.001.001.00 | 1.000 | 97c79b7933a8… |
| 456 | 2026-08-23 09:34 UTC | 1.001.000.00 rate limited0.00 rate limited1.00 | 0.600 | bb80734b271d… |
| 455 | 2026-08-23 08:34 UTC | 1.001.001.001.001.00 | 1.000 | 92ddd53be4ec… |
| 454 | 2026-08-23 07:30 UTC | 0.00 rate limited1.000.00 rate limited0.00 rate limited0.00 rate limited | 0.200 | 8ff2cba7a8f0… |
| 453 | 2026-08-23 06:30 UTC | 1.001.001.001.001.00 | 1.000 | 6eb31bf630fe… |
| 452 | 2026-08-23 05:28 UTC | 1.000.00 rate limited0.00 rate limited0.00 truncated0.00 rate limited | 0.200 | f796d726dc92… |
| 451 | 2026-08-23 04:26 UTC | 1.000.00 rate limited0.00 rate limited1.001.00 | 0.600 | 59def933d70b… |
| 450 | 2026-08-23 03:21 UTC | 1.001.001.001.001.00 | 1.000 | b7339ac0b2f9… |
| 449 | 2026-08-23 02:21 UTC | 0.00 rate limited0.00 rate limited1.001.001.00 | 0.600 | b88363bb900f… |
| 448 | 2026-08-23 01:20 UTC | 0.00 rate limited1.001.001.001.00 | 0.800 | 64fcaa564030… |
| 447 | 2026-08-23 00:20 UTC | 1.000.00 rate limited1.001.000.00 rate limited | 0.600 | 8b33d1081df1… |
| 446 | 2026-08-22 23:18 UTC | 1.001.001.001.001.00 | 1.000 | 5c4b31fb0966… |
| 445 | 2026-08-22 22:13 UTC | 1.001.001.001.001.00 | 1.000 | 41e55b950e5f… |
| 444 | 2026-08-22 21:08 UTC | 1.001.000.00 rate limited1.001.00 | 0.800 | ac4ce6b5dae2… |
| 443 | 2026-08-22 20:05 UTC | 1.000.00 provider error1.000.00 provider error1.00 | 0.600 | 84903dfb9ecf… |
| 442 | 2026-08-22 19:03 UTC | 1.001.000.00 rate limited1.001.00 | 0.800 | 05a4a9ca35e4… |
| 441 | 2026-08-22 17:58 UTC | 1.000.00 rate limited1.001.000.00 rate limited | 0.600 | 322e49ee5cf3… |
| 440 | 2026-08-22 16:59 UTC | 1.001.001.001.001.00 | 1.000 | 9fc002782e49… |
| 439 | 2026-08-22 15:53 UTC | 1.001.001.001.001.00 | 1.000 | 92c2f0286b27… |
| 438 | 2026-08-22 14:53 UTC | 1.001.001.001.001.00 | 1.000 | 333546d56261… |
| 437 | 2026-08-22 13:52 UTC | 1.001.000.00 rate limited1.001.00 | 0.800 | e44915047f5f… |
| 436 | 2026-08-22 12:51 UTC | 1.000.00 rate limited1.000.00 rate limited1.00 | 0.600 | 3b20e10ed74e… |
| 435 | 2026-08-22 11:50 UTC | 1.001.001.001.001.00 | 1.000 | 1e987988c776… |
| 434 | 2026-08-22 10:45 UTC | 1.000.00 rate limited0.00 rate limited1.001.00 | 0.600 | 78c62438e669… |
| 433 | 2026-08-22 09:45 UTC | 1.001.000.00 truncated1.001.00 | 0.800 | 8d458e099d19… |
| 432 | 2026-08-22 08:45 UTC | 1.001.001.001.001.00 | 1.000 | 22c0b4e87df1… |
| 431 | 2026-08-22 07:40 UTC | 0.00 rate limited1.001.001.000.00 truncated | 0.600 | 8fe4fe93b4b0… |
Newest first; older panels are retained and replayable even where this table truncates. A failed member scores 0 and stays on the record. Each evidence link opens the panel’s full hash-committed record.
Epoch & design record
- Current epoch
17d9ded1-4381-4d83-9292-4e120bf1d079- Opened
- 2026-08-05 12:27 UTC
- Previous epoch
87f9507e-e714-42c6-9d16-71fb0235e292- Transition
- checker_refresh
- Design hash
bc3668bd52d9d48803b744c2c90f4ed816d7e394525b32b4219f6bab5c9c6df2- Panel hash
827de544c7641d140ac37e321bdb2578a41e291915f519183abbc5bd68153e81- Pool hash
cbfba66601ff51f43c83017d6de9e49988afcee72235f5dca89d705f4f8bb4cc- Panel size
- 5
- Terminal failure
- zero — a failed member scores 0 and stays on the record
- Pool version
code_reasoning-v4- Checker version
code_reasoning-v2- Reference mean μ₀
- 0.700
- Minimum effect δ
- 0.000
- Evidence log e
- 28.65 · max 33.54 · alarm at 3.00 (= ln 1/α)
- Raw scores kept
- 1242
- Provider spend, this epoch
- $3.3857987
The persisted monitor state is the authority a certificate replays byte-for-byte; the chart band above is a faithful replay of the same recorded panel means. Spend sums completed probe runs and includes reserved estimates where a run’s true cost was unrecoverable.