Medical Visual Question Answering on MMMU Med (Accuracy)
78AccuracyViToS
Evaluation results
| Method | Links | |
|---|---|---|
| ViToSBackbone=Lingshu-7B2026.06 | 78 | |
| ViToS2026.06 | 78 | |
| ViTAR2026.06 | 72 | |
| PDropBackbone=Lingshu-7B2026.06 | 66.67 | |
| ViToSBackbone=Lingshu-32B2026.06 | 66.67 | |
| VScanBackbone=Lingshu-7B2026.06 | 64.83 | |
| BaselineBackbone=Lingshu-7B2026.06 | 63.33 | |
| VisionZipBackbone=Lingshu-7B2026.06 | 62.76 | |
| BaselineBackbone=Lingshu-32B2026.06 | 62 | |
| VScanBackbone=Lingshu-32B2026.06 | 60 | |
| FastVBackbone=Lingshu-32B2026.06 | 59.31 | |
| MedCCO2026.06 | 59.3 | |
| VisionZipBackbone=Lingshu-32B2026.06 | 58.67 | |
| PDropBackbone=Lingshu-32B2026.06 | 58.67 | |
| FastVBackbone=Lingshu-7B2026.06 | 56.55 | |
| MEDIC-ADSize=7B2026.03 | 54.2 | |
| LingshuSize=7B2026.03 | 54 | |
| ViToSBackbone=HuatuoGPT-Vision-7B2026.06 | 50 | |
| Citrus-VSize=8B2026.03 | 46.4 | |
| VScanBackbone=HuatuoGPT-Vision-7B2026.06 | 46 | |
| MedVLThinker2026.06 | 45.5 | |
| BaselineBackbone=HuatuoGPT-Vision-7B2026.06 | 45.33 | |
| FastVBackbone=HuatuoGPT-Vision-7B2026.06 | 44.83 | |
| VisionZipBackbone=HuatuoGPT-Vision-7B2026.06 | 44 | |
| PDropBackbone=HuatuoGPT-Vision-7B2026.06 | 44 | |
| MedVLM-R12026.06 | 35.5 | |
| Med-R12026.06 | 32.7 | |
| LLaVA-MedSize=7B2026.03 | 29.3 | |
| LingshuSize=7B, Training protocol=full finetuning2026.03 | 24.8 |