Mathematical Reasoning on SVAMP (Accuracy, Micro avg.)
90.33Accuracy (SVAMP)TraceLift
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| TraceLiftModel=Qwen2.5-7B2026.05 | 90.33 | 70.19 | |
| Exec-onlyModel=Qwen2.5-14B2026.05 | 88.67 | 71.93 | |
| TraceLiftModel=Qwen2.5-14B2026.05 | 88.67 | 72.25 | |
| BaseModel=Qwen2.5-14B2026.05 | 86 | 71.2 | |
| Exec-onlyModel=Qwen2.5-7B2026.05 | 67.67 | 65.62 | |
| BaseModel=Qwen2.5-7B2026.05 | 58.67 | 46.51 | |
| TraceLiftModel=Llama3.1-8B2026.05 | 47.67 | 22.86 | |
| Exec-onlyModel=Llama3.1-8B2026.05 | 46.33 | 22.19 | |
| BaseModel=Llama3.1-8B2026.05 | 34.33 | 12.74 |