Instruction Following on IF
39.2IF ScoreChat-Math-IF
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Chat-Math-IFBackbone=Qwen-2.5-7B, Training Strategy=SEQUENTIAL Training2026.02 | 39.2 | 40.3 | |
| Modular Gradient SurgeryBackbone=Qwen-2.5-7B, Training Strategy=MIXED Training2026.02 | 33.7 | 45 | |
| Normal MixingBackbone=Qwen-2.5-7B, Training Strategy=MIXED Training2026.02 | 33.2 | 44.6 | |
| Global SurgeryBackbone=Llama-3.1-8B, Training Strategy=MIXED Training2026.02 | 31.1 | 30.3 | |
| Modular Gradient SurgeryBackbone=Llama-3.1-8B, Training Strategy=MIXED Training2026.02 | 30 | 32.6 | |
| Global SurgeryBackbone=Qwen-2.5-7B, Training Strategy=MIXED Training2026.02 | 29.9 | 43 | |
| Math-Chat-IFBackbone=Qwen-2.5-7B, Training Strategy=SEQUENTIAL Training2026.02 | 26.2 | 20.2 | |
| Normal MixingBackbone=Llama-3.1-8B, Training Strategy=MIXED Training2026.02 | 25.4 | 27.3 |