Mathematical Reasoning on Olympiad (Avg@32, Pass@32)
49.7Average@32SCOPE (Ours)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| SCOPE (Ours)Teacher Model=SkyWork-OR1-Math-7B, Student Model=DeepSeek-R1-Distill-Qwen-1.5B2026.04 | 49.7 | 70.9 | |
| w/ OPDTeacher Model=SkyWork-OR1-Math-7B, Student Model=DeepSeek-R1-Distill-Qwen-1.5B2026.04 | 44.9 | 69.3 | |
| R1-Distill-Qwen-1.5BTeacher Model=SkyWork-OR1-Math-7B, Student Model=DeepSeek-R1-Distill-Qwen-1.5B2026.04 | 44.2 | 67.5 | |
| w/ GRPOTeacher Model=SkyWork-OR1-Math-7B, Student Model=DeepSeek-R1-Distill-Qwen-1.5B2026.04 | 40.5 | 67.7 | |
| w/ KDTeacher Model=SkyWork-OR1-Math-7B, Student Model=DeepSeek-R1-Distill-Qwen-1.5B2026.04 | 39.7 | 66.9 | |
| w/ GRPOTeacher Model=Qwen3-8B-Instruct, Student Model=Qwen3-1.7B-Base2026.04 | 25.9 | 50.9 | |
| w/ OPDTeacher Model=Qwen3-8B-Instruct, Student Model=Qwen3-1.7B-Base2026.04 | 25.3 | 53.7 | |
| SCOPE (Ours)Teacher Model=Qwen3-8B-Instruct, Student Model=Qwen3-1.7B-Base2026.04 | 24.6 | 54.6 | |
| Qwen3-1.7B-BaseTeacher Model=Qwen3-8B-Instruct, Student Model=Qwen3-1.7B-Base2026.04 | 16.3 | 48 | |
| w/ KDTeacher Model=Qwen3-8B-Instruct, Student Model=Qwen3-1.7B-Base2026.04 | 13.3 | 41.4 |