Uncertainty Quantification on MedQA (test)
0.635AUROCSAR
Evaluation Results
| Method | Links | |
|---|---|---|
| SARBackbone=WizardLM-13b2023.07 | 0.635 | |
| SEBackbone=WizardLM-13b2023.07 | 0.62 | |
| SARBackbone=LLaMA-2-13b-chat2023.07 | 0.616 | |
| SEBackbone=LLaMA-2-13b-chat2023.07 | 0.609 | |
| LN-PEBackbone=WizardLM-13b2023.07 | 0.609 | |
| SEBackbone=Vicuna-13b2023.07 | 0.599 | |
| SARBackbone=Vicuna-13b2023.07 | 0.598 | |
| LN-PEBackbone=Vicuna-13b2023.07 | 0.572 | |
| LN-PEBackbone=LLaMA-2-13b-chat2023.07 | 0.562 |