Mathematical Reasoning on AMC 2023 (Pass@8 accuracy)
93.7Pass@8 AccuracyDiScO
Evaluation Results
| Method | Links | |
|---|---|---|
| DiScOParameter Scale=7B, Evaluation Protocol=Pass@8 (n=32)2026.06 | 93.7 | |
| ReasonFluxParameter Scale=7B, Evaluation Protocol=Pass@8 (n=32)2026.06 | 93.1 | |
| Qwen2.5-MathParameter Scale=7B, Evaluation Protocol=Pass@8 (n=32)2026.06 | 90.8 | |
| R1-Distill-QwenParameter Scale=7B, Evaluation Protocol=Pass@8 (n=32)2026.06 | 89.9 | |
| Qwen2.5-Math-InstParameter Scale=7B, Evaluation Protocol=Pass@8 (n=32)2026.06 | 84 |