Chest X-ray classification on MIMIC-CXR MAIRA (test)
51.6Macro F1Med-PaLM M
Evaluation Results
| Method | Links | |
|---|---|---|
| Med-PaLM MModel size category=SOTA VLM, Evaluation protocol=zero-shot2025.07 | 51.6 | |
| MedGemma 1.5Model Scale=4B, Temperature=0.02026.04 | 41.5 | |
| MedGemma 4BModel size category=Small Models, Evaluation protocol=zero-shot2025.07 | 40.5 | |
| MedGemma 1Model Scale=4B, Temperature=0.02026.04 | 40.5 | |
| MedGemma 1Model Scale=27B, Temperature=0.02026.04 | 40.2 | |
| Gemini 3 ProTemperature=default2026.04 | 39.4 | |
| Gemini 3 FlashTemperature=default2026.04 | 37.8 | |
| Gemini v2.5 FlashModel size category=Large Models, Evaluation protocol=zero-shot2025.07 | 32.2 | |
| Gemini v2.5 ProModel size category=Large Models, Evaluation protocol=zero-shot2025.07 | 31.9 | |
| Qwen3 VLModel Scale=4B, Temperature=default2026.04 | 31.1 | |
| Gemma 3 4BModel size category=Small Models, Evaluation protocol=zero-shot2025.07 | 26.7 | |
| Gemma 3 27BModel size category=Small Models, Evaluation protocol=zero-shot2025.07 | 25 |