Misconception Classification on Misconception Classification
95.99MAP@3Qwen-3-4B (student)
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen-3-4B (student)Type=Our Models, Time (h)=0.008, Hardware=8 × H202026.05 | 95.99 | |
| Qwen-2.5-72B (teacher)Type=Our Models, Time (h)=0.186, Hardware=8 × H202026.05 | 94.97 | |
| GPT-5Type=API Model, Time (h)=1.50, Hardware=Cloud API2026.05 | 81.37 | |
| Claude-4-SonnetType=API Model, Time (h)=1.80, Hardware=Cloud API2026.05 | 76.65 | |
| GPT-OSS-120BType=Self-deployment Model, Time (h)=1.10, Hardware=32 × H202026.05 | 76.61 | |
| Qwen-2.5-72BType=Self-deployment Model, Time (h)=1.30, Hardware=32 × H202026.05 | 72.85 |