Mathematical Reasoning on AIME 25 (Pass@1, Pass@4, Pass@8)
40Pass@8MOTAB
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| MOTABStudent LLM=Qwen3-1.7B, Teacher Model=Qwen3-32B2026.05 | 40 | 32.05 | 14.58 | |
| ImitKDStudent LLM=Qwen3-1.7B, Teacher Model=Qwen3-32B2026.05 | 36.67 | 28.33 | 12.5 | |
| SKDStudent LLM=Qwen3-1.7B, Teacher Model=Qwen3-32B, beta=0.82026.05 | 36.67 | 28.24 | 13.33 | |
| RFTStudent LLM=Qwen3-1.7B, Teacher Model=Qwen3-32B2026.05 | 33.33 | 24.67 | 12.08 | |
| SKDStudent LLM=Qwen3-1.7B, Teacher Model=Qwen3-32B, beta=1.02026.05 | 33.33 | 23.9 | 9.17 | |
| SFTStudent LLM=Qwen3-1.7B, Teacher Model=Qwen3-32B2026.05 | 26.67 | 22.29 | 11.25 | |
| SKDStudent LLM=Qwen3-1.7B, Teacher Model=Qwen3-32B, beta=0.62026.05 | 26.67 | 22.33 | 9.59 | |
| ImitKDStudent LLM=Qwen2.5-7B-Instruct, Teacher Model=Qwen3-32B2026.05 | 26.67 | 16.57 | 5.83 | |
| SKDStudent LLM=Qwen2.5-7B-Instruct, Teacher Model=Qwen3-32B, beta=0.82026.05 | 26.67 | 18.81 | 7.5 | |
| MOTABStudent LLM=Qwen2.5-7B-Instruct, Teacher Model=Qwen3-32B2026.05 | 26.67 | 19.52 | 7.92 | |
| SKDStudent LLM=Qwen2.5-7B-Instruct, Teacher Model=Qwen3-32B, beta=0.62026.05 | 23.33 | 17.86 | 7.92 | |
| SKDStudent LLM=Qwen2.5-7B-Instruct, Teacher Model=Qwen3-32B, beta=1.02026.05 | 23.33 | 17.14 | 6.67 | |
| RFTStudent LLM=Qwen2.5-7B-Instruct, Teacher Model=Qwen3-32B2026.05 | 20 | 14 | 5 | |
| MOTABStudent LLM=Llama3.2-3B-Instruct, Teacher Model=Qwen3-32B2026.05 | 16.67 | — | — | |
| SFTStudent LLM=Qwen2.5-7B-Instruct, Teacher Model=Qwen3-32B2026.05 | 16.67 | 10 | 4.58 | |
| SKDStudent LLM=Llama3.2-3B-Instruct, Teacher Model=Qwen3-32B, beta=0.82026.05 | 13.33 | 9.05 | 2.92 | |
| SFTStudent LLM=Llama3.2-3B-Instruct, Teacher Model=Qwen3-32B2026.05 | 6.67 | 3.33 | 0.83 | |
| RFTStudent LLM=Llama3.2-3B-Instruct, Teacher Model=Qwen3-32B2026.05 | 6.67 | 3.33 | 0.83 | |
| ImitKDStudent LLM=Llama3.2-3B-Instruct, Teacher Model=Qwen3-32B2026.05 | 6.67 | 4.76 | 1.67 | |
| SKDStudent LLM=Llama3.2-3B-Instruct, Teacher Model=Qwen3-32B, beta=0.62026.05 | 6.67 | 4.95 | 2.08 | |
| SKDStudent LLM=Llama3.2-3B-Instruct, Teacher Model=Qwen3-32B, beta=1.02026.05 | 6.67 | 5.24 | 1.67 |