Mathematical Reasoning on AIME 2026 (Exact Match)
50.8Exact Match AccuracyOPD
Evaluation Results
| Method | Links | |
|---|---|---|
| OPDTeacher-Student Setting=DeepSeek-R1 -> DeepSeek-R1-Distill-Qwen-7B2026.06 | 50.8 | |
| CDMTeacher-Student Setting=DeepSeek-R1 -> DeepSeek-R1-Distill-Qwen-7B2026.06 | 48 | |
| ALMTeacher-Student Setting=DeepSeek-R1 -> DeepSeek-R1-Distill-Qwen-7B2026.06 | 47.3 | |
| SFTTeacher-Student Setting=DeepSeek-R1 -> DeepSeek-R1-Distill-Qwen-7B2026.06 | 46.7 | |
| OPDTeacher-Student Setting=Qwen3-8B -> Llama3.1-8B-SFT2026.06 | 41.7 | |
| CDMTeacher-Student Setting=Qwen3-8B -> Llama3.1-8B-SFT2026.06 | 37.3 | |
| ALMTeacher-Student Setting=Qwen3-8B -> Llama3.1-8B-SFT2026.06 | 37.1 | |
| SFTTeacher-Student Setting=Qwen3-8B -> Llama3.1-8B-SFT2026.06 | 36.5 |