Mathematical Reasoning on Average (AMC23, Minerva, Olympiad, Math500, AIME24, AIME25)
38.9Acc@16Generate
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GenerateModel=Qwen2.5-7B-Instruct2026.02 | 38.9 | 4,458 | |
| Self-VerifyModel=Qwen2.5-7B-Instruct2026.02 | 38.4 | 1,152 | |
| Self-VerifyModel=Qwen2.5-3B-Instruct2026.02 | 29.5 | 1,936 | |
| GenerateModel=Qwen2.5-3B-Instruct2026.02 | 28.5 | 3,273 | |
| Self-VerifyModel=Qwen2.5-1.5B-Instruct2026.02 | 21.7 | 1,227 | |
| GenerateModel=Qwen2.5-1.5B-Instruct2026.02 | 20.2 | 1,604 |