Disease Recognition on CDDMBench
91.5AccuracyQwen-VL-Chat-AG (7B)
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen-VL-Chat-AG (7B)Method=SFT (Unfrozen encoder)2026.01 | 91.5 | |
| Qwen2.5-VL-3B-InstructMethod=Reasoning-Enhanced GRPO2026.01 | 72.5 | |
| Qwen2.5-VL-3B-InstructMethod=GRPO2026.01 | 69.43 | |
| Qwen-VL-Chat-AG* (7B)Method=SFT (Frozen encoder)2026.01 | 66.1 | |
| Qwen2.5-VL-3B-InstructMethod=SFT2026.01 | 58.84 | |
| Gpt-5-NanoMethod=+Judge2026.01 | 33.7 | |
| Gpt-5-NanoMethod=Expl. Caption2026.01 | 31.6 | |
| Gpt-5-NanoMethod=+Few-shot2026.01 | 29.8 | |
| Qwen-VL-Chat (7B)Method=+Judge2026.01 | 25.39 | |
| Qwen-VL-Chat (7B)Method=+Few-shot2026.01 | 24.49 | |
| Qwen-VL-Chat (7B)Method=Expl. Caption2026.01 | 12.1 | |
| Gpt-5-NanoMethod=Zero-shot2026.01 | 11 | |
| Qwen2.5-VL-3B-InstructMethod=Few-shot2026.01 | 6.96 | |
| Qwen-VL-Chat (7B)Method=Zero-shot2026.01 | 5 | |
| Qwen2.5-VL-3B-InstructMethod=Zero-shot2026.01 | 4.84 |