Absence Detection on CheXpert
67.06Base AccuracyLLaVA-Med
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| LLaVA-MedChain-of-Thought (CoT)=false2026.05 | 67.06 | 75.35 | 8.29 | 42 | |
| MedGemmaChain-of-Thought (CoT)=true2026.05 | 32.54 | 88.17 | 55.63 | 282 | |
| MedGemmaModel size=27B, Chain-of-Thought (CoT)=false2026.05 | 31.16 | 97.24 | 66.08 | 335 | |
| LLaVA-1.5Chain-of-Thought (CoT)=false2026.05 | 27.42 | 96.06 | 68.64 | 348 | |
| Qwen2.5-VLChain-of-Thought (CoT)=true2026.05 | 27.22 | 99.8 | 72.58 | 368 | |
| MedGemmaChain-of-Thought (CoT)=false2026.05 | 24.06 | 93.29 | 69.23 | 351 | |
| LLaVA-MedChain-of-Thought (CoT)=true2026.05 | 11.83 | 99.8 | 87.97 | 446 | |
| CheXagentChain-of-Thought (CoT)=false2026.05 | 6.51 | 43.2 | 36.69 | 186 | |
| GPT-4oChain-of-Thought (CoT)=false2026.05 | 2.17 | 94.87 | 92.7 | 470 | |
| Qwen2.5-VLChain-of-Thought (CoT)=false2026.05 | 0.39 | 99.8 | 99.41 | 504 |