Question Answering on MedMCQA
6.43FDR (%)SGR
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| SGRModel=Llama-3.1-70B-Instruct2025.10 | 6.43 | 44.34 | 33.67 | — | |
| SGRModel=Qwen3-32B2025.10 | 8.08 | 2.94 | 2.09 | — | |
| SGRModel=Llama-3.1-8B-Instruct2025.10 | 8.54 | 2.03 | 1.27 | — | |
| FDR searchModel=Llama-3.1-8B-Instruct2025.10 | 9.39 | 30.41 | 20.27 | — | |
| FDR searchModel=Qwen3-32B2025.10 | 9.55 | 48.6 | 36.05 | — | |
| Conformal LabelingModel=Llama-3.1-8B-Instruct2025.10 | 9.8 | 31.72 | 21.2 | — | |
| Conformal LabelingModel=Qwen3-32B2025.10 | 9.81 | 49.85 | 37.05 | — | |
| Conformal LabelingModel=Llama-3.1-70B-Instruct2025.10 | 9.86 | 69.37 | 54.83 | — | |
| FDR searchModel=Llama-3.1-70B-Instruct2025.10 | 10.58 | 71.15 | 56.67 | — | |
| AI onlyModel=Llama-3.1-8B-Instruct2025.10 | — | — | — | 40.35 | |
| AI onlyModel=Qwen3-32B2025.10 | — | — | — | 33.44 | |
| AI onlyModel=Llama-3.1-70B-Instruct2025.10 | — | — | — | 28.9 |