Mathematical Reasoning on MATH500 (Accuracy and Length)
88.4AccuracyTACLer
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| TACLerMode=Thinking2026.01 | 88.4 | 3,010 | |
| DeepScaleRMode=Thinking2026.01 | 87.8 | 3,030 | |
| FastCuRLMode=Thinking2026.01 | 87.8 | 3,894 | |
| STILL-3Mode=Thinking2026.01 | 83.6 | 3,797 | |
| R1-QwenMode=Thinking2026.01 | 81.2 | 4,856 |