Math & physics reasoning on OlympiadBench Phys
14.41AccuracyINFUSER
Evaluation Results
| Method | Links | |
|---|---|---|
| INFUSERModel Backbone=Qwen3-8B-Base2026.06 | 14.41 | |
| R-ZeroModel Backbone=Qwen3-8B-Base2026.06 | 13.98 | |
| AZRModel Backbone=Qwen3-8B-Base2026.06 | 13.14 | |
| BaseModel Backbone=Qwen3-8B-Base2026.06 | 12.29 | |
| R-ZeroModel Backbone=Qwen3-4B-Base2026.06 | 11.44 | |
| INFUSERModel Backbone=Qwen3-4B-Base2026.06 | 10.31 | |
| BaseModel Backbone=Qwen3-4B-Base2026.06 | 10.17 | |
| AZRModel Backbone=Qwen3-4B-Base2026.06 | 10.17 |