Differential Diagnosis on NOVA OOD
20.36Diagnostic AccuracyQwen2.5-VL-72B
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Qwen2.5-VL-72BModel Category=General Domain2026.06 | 20.36 | 42.33 | 61.09 | |
| Lingshu-32BModel Category=Medical Domain2026.06 | 14.68 | 28.7 | 48.84 | |
| BrReMark w/o synthesisModel Category=Thinking with Images2026.06 | 11.53 | 25.19 | 43.71 | |
| InternVL3.5-8BModel Category=General Domain2026.06 | 11.2 | 26.04 | 40.85 | |
| Lingshu-7BModel Category=Medical Domain2026.06 | 11.09 | 21.03 | 41.76 | |
| BrReMarkModel Category=Thinking with Images2026.06 | 10.43 | 26.32 | 43.07 | |
| MedVLM-R1Model Category=Medical Domain2026.06 | 10.14 | 15.09 | 23.02 | |
| BrReMark w/o rfmtModel Category=Thinking with Images2026.06 | 8.9 | 24.11 | 35.65 | |
| BrReMark w/o rlocModel Category=Thinking with Images2026.06 | 8.83 | 24.29 | 34.05 | |
| BrReMark w/o rllmModel Category=Thinking with Images2026.06 | 7.95 | 24.78 | 33.39 | |
| GRITModel Category=Thinking with Images2026.06 | 6.8 | 12.52 | 32.23 | |
| LLaVA-v1.5-13BModel Category=General Domain2026.06 | 6.72 | 10.68 | 48.3 | |
| HuatuoGPT-V-7BModel Category=Medical Domain2026.06 | 6.68 | 16.17 | 36.26 | |
| Qwen2.5-VL-7BModel Category=General Domain2026.06 | 6.28 | 29.35 | 30.23 |