Multilingual Reasoning on Polymath Low
96.5Accuracy (en)Base
Evaluation Results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| BaseModel=Qwen3-4B2025.10 | 96.5 | 88 | 93.9 | 89.6 | 85.3 | 90.7 | 85.1 | 83.2 | 29.3 | 69.9 | 81.1 | 0 | |
| Selective translationModel=Qwen3-4B2025.10 | 96.3 | 88.3 | 94.4 | 90.4 | 86.1 | 91.5 | 88.3 | 86.7 | 81.3 | 77.1 | 88 | 19.3 | |
| Full translationModel=Qwen3-4B2025.10 | 96 | 88.3 | 93.3 | 90.9 | 87.5 | 92.5 | 89.6 | 90.4 | 85.3 | 80.5 | 89.4 | 100 |