Models
requested versus observed model, with the trust the observed value deserves
| requested | observed | runs | source | verification |
|---|---|---|---|---|
| auto_codex | auto_codex | 66 | agent_reported | unverified x 66 |
| qwen/Qwen3.8-27B | qwen/Qwen3.8-27B | 33 | agent_reported | unverified x 33 |
| qwen/Qwen3.8-Flash-Next | qwen/Qwen3.8-Flash-Next | 33 | agent_reported | unverified x 33 |