Safety Evaluation on O2M Harmful Clinical Queries
13.43ASRH-R Demon
Evaluation Results
| Method | Links | |
|---|---|---|
| H-R DemonModel=Llava-Med-v1, Demonstration budget (n)=42025.06 | 13.43 | |
| H-R DemonModel=Llava-Med-v1, Demonstration budget (n)=22025.06 | 17.88 | |
| H-R DemonModel=Llava-Med-v1, Demonstration budget (n)=82025.06 | 21.31 | |
| H-R DemonModel=Llava-Med-v1, Demonstration budget (n)=162025.06 | 32.63 | |
| H-R DemonModel=Llava-Med-v1.5, Demonstration budget (n)=42025.06 | 42.61 | |
| H-R DemonModel=Llava-Med-v1.5, Demonstration budget (n)=22025.06 | 44.75 | |
| H-R DemonModel=Llava-Med-v1.5, Demonstration budget (n)=82025.06 | 45.1 | |
| H-R DemonModel=Llava-Med-v1.5, Demonstration budget (n)=162025.06 | 47.32 | |
| B-A DemonModel=Llava-Med-v1.5, Demonstration budget (n)=22025.06 | 56.82 | |
| B-A DemonModel=Llava-Med-v1.5, Demonstration budget (n)=162025.06 | 56.92 | |
| B-A DemonModel=Llava-Med-v1.5, Demonstration budget (n)=42025.06 | 57.53 | |
| B-A DemonModel=Llava-Med-v1.5, Demonstration budget (n)=82025.06 | 57.68 | |
| B-A DemonModel=Llava-Med-v1, Demonstration budget (n)=162025.06 | 60.76 | |
| Baseline (No Demon)Model=Llava-Med-v1.5, Demonstration budget (n)=02025.06 | 61.52 | |
| B-A DemonModel=Llava-Med-v1, Demonstration budget (n)=82025.06 | 65.05 | |
| B-A DemonModel=Llava-Med-v1, Demonstration budget (n)=42025.06 | 68.79 | |
| Baseline (No Demon)Model=Llava-Med-v1, Demonstration budget (n)=02025.06 | 72.58 | |
| B-A DemonModel=Llava-Med-v1, Demonstration budget (n)=22025.06 | 74.9 |