Reasoning on Average (MATH, OlympicBench, MinervaMath, AMC2023, GPQA) (test)
59.83Pass@1W2SR-P
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| W2SR-PStudent=Qwen2.5 32B, Teacher=Qwen2.5-14B (R)2025.05 | 59.83 | 175.04 | |
| GRPOStudent=Qwen2.5 32B2025.05 | 59.57 | — | |
| W2SRStudent=Qwen2.5 32B, Teacher=Qwen2.5-14B (R)2025.05 | 58.73 | 134.66 | |
| W2SR-PStudent=Qwen2.5 32B, Teacher=Qwen2.5-7B (R)2025.05 | 58.42 | 94.13 | |
| W2SRStudent=Qwen2.5 32B, Teacher=Qwen2.5-7B (R)2025.05 | 56.43 | 81.08 | |
| GRPOStudent=Qwen2.5 14B2025.05 | 54.71 | — | |
| Qwen2.5-14B (R)Role=Standalone Teacher2025.05 | 54.71 | — | |
| W2SR-PStudent=Qwen2.5 14B, Teacher=Qwen2.5-7B (R)2025.05 | 53.18 | 66.33 | |
| W2SR-PStudent=Qwen2.5 32B, Teacher=Qwen2.5-1.5B (R)2025.05 | 50.93 | 68.61 | |
| W2SR-PStudent=Qwen2.5 Math-7B, Teacher=Qwen2.5-1.5B (R)2025.05 | 50.76 | 141.14 | |
| Qwen2.5-7B (R)Role=Standalone Teacher2025.05 | 50.47 | — | |
| W2SRStudent=Qwen2.5 14B, Teacher=Qwen2.5-7B (R)2025.05 | 49.88 | 52.62 | |
| LIMOStudent=Qwen2.5 32B2025.05 | 49.57 | — | |
| GRPOStudent=Qwen2.5 Math-7B2025.05 | 49.32 | — | |
| W2SR-PStudent=Qwen2.5 14B, Teacher=Qwen2.5-1.5B (R)2025.05 | 47.06 | 71.45 | |
| LIMOStudent=Qwen2.5 14B2025.05 | 45.89 | — | |
| W2SRStudent=Qwen2.5 Math-7B, Teacher=Qwen2.5-1.5B (R)2025.05 | 45.57 | 108.41 | |
| W2SRStudent=Qwen2.5 32B, Teacher=Qwen2.5-1.5B (R)2025.05 | 45.41 | 49.79 | |
| W2SRStudent=Qwen2.5 14B, Teacher=Qwen2.5-1.5B (R)2025.05 | 42.62 | 46.94 | |
| W2SR-PStudent=Qwen2.5 Math-7B, Teacher=Qwen2.5-0.5B (R)2025.05 | 41.65 | 63.25 | |
| LIMOStudent=Qwen2.5 Math-7B2025.05 | 40 | — | |
| VanillaStudent=Qwen2.5 32B2025.05 | 38.92 | — | |
| VanillaStudent=Qwen2.5 14B2025.05 | 35.91 | — | |
| W2SRStudent=Qwen2.5 Math-7B, Teacher=Qwen2.5-0.5B (R)2025.05 | 31.79 | 48.46 | |
| Qwen2.5-1.5B (R)Role=Standalone Teacher2025.05 | 30.86 | — | |
| VanillaStudent=Qwen2.5 Math-7B2025.05 | 30.71 | — | |
| Qwen2.5-0.5B (R)Role=Standalone Teacher2025.05 | 21.25 | — |