Mathematical Problem Solving on AIME24
54.1Pass@1Base Model
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Base ModelModel Scale=Qwen3-4B, Thinking Mode=Thinking2026.04 | 54.1 | 75.5 | |
| MIXED-CUTSModel Scale=Qwen3-4B, Thinking Mode=Non-thinking2026.04 | 46 | 73.5 | |
| GRPOModel Scale=Qwen3-4B, Thinking Mode=Non-thinking2026.04 | 32.5 | 63.8 | |
| MIXED-CUTSModel Scale=Qwen3-1.7B, Thinking Mode=Non-thinking2026.04 | 32.3 | 62 | |
| GRPOModel Scale=Qwen3-1.7B, Thinking Mode=Non-thinking2026.04 | 29.5 | 60.7 | |
| Base ModelModel Scale=Qwen3-1.7B, Thinking Mode=Thinking2026.04 | 28.9 | 62.8 | |
| Base ModelModel Scale=Qwen3-4B, Thinking Mode=Non-thinking2026.04 | 24.2 | 54 | |
| Base ModelModel Scale=Qwen3-1.7B, Thinking Mode=Non-thinking2026.04 | 12.9 | 36.5 |