Medical Visual Question Answering on Medical-Diff-VQA (test)
49.6BLEU-4Qwen3.5 4B (SFT) + SVE
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| Qwen3.5 4B (SFT) + SVEBase Model=Qwen3.5, Parameters=4B, Training Protocol=SFT, Visual Encoder=Stateful Visual Encoder2026.06 | 49.6 | 40.9 | 66.3 | 178.9 | 32.2 | 12.45 | 89.21 | |
| Qwen3.5 4B (SFT)Base Model=Qwen3.5, Parameters=4B, Training Protocol=SFT2026.06 | 47.9 | 40.6 | 62.7 | 145.1 | 31.55 | 11.95 | 86.83 |