Mathematical Reasoning on HMMT-26 (mean@8, pass@8)
43.94Mean@8DeepSeek-R1-Distill-Qwen-32B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DeepSeek-R1-Distill-Qwen-32B2026.05 | 43.94 | 63.64 | |
| SkyWork-OR1-Math-7B2026.05 | 36.74 | 54.55 | |
| GRPOStudent Model=DeepSeek-R1-Distill-Qwen-7B, Teacher Model=DeepSeek-R1-Distill-Qwen-32B2026.05 | 36.74 | 54.55 | |
| LGRStudent Model=DeepSeek-R1-Distill-Qwen-7B, Teacher Model=DeepSeek-R1-Distill-Qwen-32B2026.05 | 34.85 | 51.52 | |
| OPD (9k)Student Model=DeepSeek-R1-Distill-Qwen-7B, Teacher Model=DeepSeek-R1-Distill-Qwen-32B2026.05 | 32.95 | 45.45 | |
| REOPOLDStudent Model=DeepSeek-R1-Distill-Qwen-7B, Teacher Model=DeepSeek-R1-Distill-Qwen-32B2026.05 | 32.95 | 46.88 | |
| JSDStudent Model=DeepSeek-R1-Distill-Qwen-7B, Teacher Model=DeepSeek-R1-Distill-Qwen-32B2026.05 | 31.82 | 42.42 | |
| LGR (topk)Student Model=DeepSeek-R1-Distill-Qwen-7B, Teacher Model=DeepSeek-R1-Distill-Qwen-32B2026.05 | 30.68 | 36.36 | |
| OPD (topk)Student Model=DeepSeek-R1-Distill-Qwen-7B, Teacher Model=DeepSeek-R1-Distill-Qwen-32B2026.05 | 30.3 | 42.42 | |
| Base ModelStudent Model=DeepSeek-R1-Distill-Qwen-7B2026.05 | 28.03 | 39.39 | |
| LGR (topk)Student Model=DeepSeek-R1-Distill-Qwen-1.5B, Teacher Model=SkyWork-OR1-Math-7B2026.05 | 24.62 | 34.38 | |
| LGRStudent Model=DeepSeek-R1-Distill-Qwen-1.5B, Teacher Model=SkyWork-OR1-Math-7B2026.05 | 23.86 | 37.5 | |
| OPD (9k)Student Model=DeepSeek-R1-Distill-Qwen-1.5B, Teacher Model=SkyWork-OR1-Math-7B2026.05 | 22.73 | 27.27 | |
| REOPOLDStudent Model=DeepSeek-R1-Distill-Qwen-1.5B, Teacher Model=SkyWork-OR1-Math-7B2026.05 | 22.35 | 27.27 | |
| OPD (topk)Student Model=DeepSeek-R1-Distill-Qwen-1.5B, Teacher Model=SkyWork-OR1-Math-7B2026.05 | 21.59 | 33.33 | |
| JSDStudent Model=DeepSeek-R1-Distill-Qwen-1.5B, Teacher Model=SkyWork-OR1-Math-7B2026.05 | 21.21 | 33.33 | |
| GRPOStudent Model=DeepSeek-R1-Distill-Qwen-1.5B, Teacher Model=SkyWork-OR1-Math-7B2026.05 | 20.46 | 27.27 | |
| Base ModelStudent Model=DeepSeek-R1-Distill-Qwen-1.5B2026.05 | 12.5 | 27.27 |