Medical Visual Question Answering on MedX-Bench
29.8AccuracyViToS
Evaluation results
| Method | Links | |
|---|---|---|
| ViToSBackbone=Lingshu-32B2026.06 | 29.8 | |
| PDropBackbone=Lingshu-32B2026.06 | 28.45 | |
| BaselineBackbone=Lingshu-32B2026.06 | 28.4 | |
| FastVBackbone=Lingshu-32B2026.06 | 28.21 | |
| ViToSBackbone=Lingshu-7B2026.06 | 25.8 | |
| ViToSBackbone=HuatuoGPT-Vision-7B2026.06 | 23.9 | |
| PDropBackbone=Lingshu-7B2026.06 | 23.51 | |
| BaselineBackbone=Lingshu-7B2026.06 | 23.5 | |
| VisionZipBackbone=Lingshu-7B2026.06 | 23.47 | |
| VScanBackbone=Lingshu-7B2026.06 | 23.09 | |
| VisionZipBackbone=Lingshu-32B2026.06 | 22.45 | |
| FastVBackbone=Lingshu-7B2026.06 | 22.39 | |
| BaselineBackbone=HuatuoGPT-Vision-7B2026.06 | 22.35 | |
| VScanBackbone=Lingshu-32B2026.06 | 22.2 | |
| PDropBackbone=HuatuoGPT-Vision-7B2026.06 | 21.95 | |
| FastVBackbone=HuatuoGPT-Vision-7B2026.06 | 21.06 | |
| VisionZipBackbone=HuatuoGPT-Vision-7B2026.06 | 17.5 | |
| VScanBackbone=HuatuoGPT-Vision-7B2026.06 | 17.3 |