Error Recognition on Math500
57.1Top-1 AccuracyCORRECT
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| CORRECTSynthesized by=GPT-4o-mini2025.09 | 57.1 | 87.8 | 95.9 | |
| CORRECTSynthesized by=GPT-5-Nano2025.09 | 57.1 | 87.8 | 95.9 | |
| BaselineSynthesized by=GPT-5-Nano2025.09 | 41.8 | 57.1 | 64.3 | |
| LLM-as-a-JudgeSynthesized by=GPT-4o-mini2025.09 | 10.2 | 23.4 | 35.6 |