Math on Math
60.8Math ScoreGlobal Surgery
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Global SurgeryBackbone=Qwen-2.5-7B, Training Strategy=MIXED Training2026.02 | 60.8 | 43 | |
| Chat-Math-IFBackbone=Qwen-2.5-7B, Training Strategy=SEQUENTIAL Training2026.02 | 59.9 | 40.3 | |
| Normal MixingBackbone=Qwen-2.5-7B, Training Strategy=MIXED Training2026.02 | 59.3 | 44.6 | |
| Modular Gradient SurgeryBackbone=Qwen-2.5-7B, Training Strategy=MIXED Training2026.02 | 58.8 | 45 | |
| Modular Gradient SurgeryBackbone=Llama-3.1-8B, Training Strategy=MIXED Training2026.02 | 36 | 32.6 | |
| Global SurgeryBackbone=Llama-3.1-8B, Training Strategy=MIXED Training2026.02 | 35.8 | 30.3 | |
| Normal MixingBackbone=Llama-3.1-8B, Training Strategy=MIXED Training2026.02 | 35 | 27.3 | |
| Math-Chat-IFBackbone=Qwen-2.5-7B, Training Strategy=SEQUENTIAL Training2026.02 | 0.9 | 20.2 |