Mathematical Reasoning on Math 500 (pass@4, pass@8, pass@16)
98.44Pass@4 Success RateQwen2.5-7B EP-GRPO
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Qwen2.5-7B EP-GRPOModel=Qwen2.5-7B, Method=EP-GRPO2026.05 | 98.44 | 99.98 | 100 | |
| Qwen2.5-7B GRPOModel=Qwen2.5-7B, Method=GRPO2026.05 | 97.78 | 99.95 | 100 | |
| Qwen2.5-3B EP-GRPOModel=Qwen2.5-3B, Method=EP-GRPO2026.05 | 96.58 | 99.89 | 100 | |
| Qwen2.5-3B GRPO (More Rollouts)Model=Qwen2.5-3B, Method=GRPO, Rollouts=More2026.05 | 96 | 99.85 | 100 | |
| Qwen2.5-3B GRPO (Higher Temp)Model=Qwen2.5-3B, Method=GRPO, Temperature=Higher2026.05 | 94.48 | 99.7 | 100 | |
| Qwen2.5-3B GRPOModel=Qwen2.5-3B, Method=GRPO2026.05 | 94.26 | 99.68 | 100 | |
| Qwen2.5-3B BaseModel=Qwen2.5-3B, Method=Base2026.05 | 78.16 | 95.28 | 99.79 | |
| Qwen3-235B-A22B-InstructModel=Qwen3-235B-A22B-Instruct, Category=Commercial Models2026.05 | 76.06 | 78.07 | 80 | |
| Qwen2.5-7B BaseModel=Qwen2.5-7B, Method=Base2026.05 | 74.78 | 93.7 | 99.62 | |
| DeepSeek-R1-671B-0528Model=DeepSeek-R1-671B-0528, Category=Commercial Models2026.05 | 66.63 | 69.33 | 70 |