Adversarial detection on Expanded benchmark Misapplied theorem
13Caught CountTheoria
Evaluation results
| Method | Links | ||
|---|---|---|---|
| TheoriaModel=GPT-5.5, Web search=Enabled, Evaluation protocol=Structured (step-level typed judges with BEFORE→AFTER diffs)2026.07 | 13 | 92.9 | |
| Holistic judgeModel=GPT-5.5, Web search=Enabled, Evaluation protocol=Holistic (single-call global assessment)2026.07 | 13 | 92.9 |