Mathematical Reasoning on AMC23 (Maj@16, Avg@16)
95Maj@16GRPO
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GRPOModel=Qwen3-4B-Base2026.06 | 95 | 87.2 | |
| PAECModel=Qwen3-4B-Base2026.06 | 92.5 | 87.8 | |
| BaseModel=Qwen3-4B-Base2026.06 | 90 | 82.9 | |
| Global Entropy RegModel=Qwen3-4B-Base2026.06 | 90 | 82.9 | |
| PAECModel=Llama3.1-8B2026.06 | 15 | 10.4 | |
| Global Entropy RegModel=Llama3.1-8B2026.06 | 10 | 8.4 | |
| GRPOModel=Llama3.1-8B2026.06 | 7.5 | 7.2 | |
| BaseModel=Llama3.1-8B2026.06 | 5 | 3 |