Differential Diagnosis on Eurorad (test)
74.9Diagnostic Accuracy (Baseline)Qwen3-VL-30B
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| Qwen3-VL-30BDomain=General2026.05 | 74.9 | 79.6 | 4.7 | 58.6 | 80.5 | 7 | 2.2 | |
| InternVL3.5-30BDomain=General2026.05 | 74 | 77.5 | 3.5 | 68.3 | 75.5 | 8.8 | 5.2 | |
| Lingshu-32BDomain=Medical2026.05 | 72.2 | 74.8 | 2.6 | 54.1 | 74.2 | 6.5 | 3.8 | |
| Qwen3-VL-8BDomain=General2026.05 | 69.6 | 74.5 | 4.9 | 77.5 | 74.4 | 9.7 | 4.8 | |
| Lingshu-7BDomain=Medical2026.05 | 66.1 | 67.8 | 1.7 | 58.2 | 68.2 | 3.5 | 1.7 | |
| InternVL3.5-8BDomain=General2026.05 | 64 | 71 | 6.9 | 81.9 | 73.7 | 12.7 | 5.8 | |
| Qwen3-VL-2BDomain=General2026.05 | 54.6 | 61.6 | 7 | 49.6 | 60.7 | 8.4 | 1.3 |