Medical Question Answering on MedXpertQA-U
12.39Pass@1Qwen3-1.7B
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen3-1.7BTrain Size=100%2026.05 | 12.39 | |
| SHIFTTrain Size=0.1%2026.05 | 12.22 | |
| SHIFTTrain Size=0.2%2026.05 | 12.22 | |
| ClusterTrain Size=0.2%2026.05 | 11.54 | |
| CoT SimilarityTrain Size=0.1%2026.05 | 10.19 | |
| CoT SimilarityTrain Size=0.2%2026.05 | 9.68 | |
| SC-EntropyTrain Size=0.2%2026.05 | 9.17 | |
| Random (avg)Train Size=0.2%2026.05 | 9.1 | |
| SC-EntropyTrain Size=0.1%2026.05 | 8.83 | |
| A-PPLTrain Size=0.2%2026.05 | 8.83 | |
| CoreSetTrain Size=0.2%2026.05 | 8.66 | |
| Random (avg)Train Size=0.1%2026.05 | 8.56 | |
| Q-PPLTrain Size=0.1%2026.05 | 8.49 | |
| CoreSetTrain Size=0.1%2026.05 | 7.64 | |
| Q-PPLTrain Size=0.2%2026.05 | 7.64 | |
| ClusterTrain Size=0.1%2026.05 | 7.3 | |
| A-PPLTrain Size=0.1%2026.05 | 7.3 | |
| Qwen3-1.7BTrain Size=N/A2026.05 | 4.92 |