Mathematical Reasoning on Olympiad (Avg@8, Pass@8)
17.22Avg@8PowerOPD
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| PowerOPDalpha=1, Teacher=Qwen3-4B, Student=Qwen3-0.6B-Base2026.06 | 17.22 | 35.85 | 2,701 | |
| PowerOPDalpha=500, Teacher=Qwen3-4B, Student=Qwen3-0.6B-Base2026.06 | 17.22 | 35.85 | 2,185 | |
| PowerOPDalpha=50, Teacher=Qwen3-4B, Student=Qwen3-0.6B-Base2026.06 | 17.2 | 36.59 | 2,571 | |
| PowerOPDalpha=10, Teacher=Qwen3-4B, Student=Qwen3-0.6B-Base2026.06 | 17.11 | 35.26 | 2,664 | |
| PowerOPDalpha=100, Teacher=Qwen3-4B, Student=Qwen3-0.6B-Base2026.06 | 17.04 | 36.15 | 2,490 | |
| PowerOPDalpha=0.5, Teacher=Qwen3-4B, Student=Qwen3-0.6B-Base2026.06 | 16.96 | 36.15 | 2,704 | |
| PowerOPDalpha=5, Teacher=Qwen3-4B, Student=Qwen3-0.6B-Base2026.06 | 16.87 | 35.85 | 2,677 | |
| PowerOPDalpha=0.1, Teacher=Qwen3-4B, Student=Qwen3-0.6B-Base2026.06 | 16.48 | 34.67 | 2,776 | |
| VanillaParam=+ Clip, Teacher=Qwen3-4B, Student=Qwen3-0.6B-Base2026.06 | 15.67 | 33.78 | 2,809 | |
| VanillaParam=+ Z-Score, Teacher=Qwen3-4B, Student=Qwen3-0.6B-Base2026.06 | 12.61 | 31.41 | 2,750 | |
| VanillaTeacher=Qwen3-4B, Student=Qwen3-0.6B-Base2026.06 | 12.5 | 30.07 | 4,087 | |
| Full-vocabTeacher=Qwen3-4B, Student=Qwen3-0.6B-Base2026.06 | 5.62 | 25 | 4,094 | |
| VanillaParam=+ Tanh, Teacher=Qwen3-4B, Student=Qwen3-0.6B-Base2026.06 | 4.62 | 15 | 3,872 |