Medical Symptom Tracking on MMXU
46.8Worsen RateCitrus-V
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Citrus-VSize=8B2026.03 | 46.8 | 71.3 | 53.2 | 57.1 | |
| GPT-4oSize=–2026.03 | 48 | 67.5 | 55.9 | 57.1 | |
| InternVL2.5Size=8B2026.03 | 48.6 | 60.7 | 40.2 | 49.8 | |
| Claude-3.5Size=–2026.03 | 49.4 | 51.8 | 49.3 | 50.2 | |
| Qwen2.5-VLSize=7B2026.03 | 49.9 | 54.5 | 51.3 | 51.9 | |
| LingshuSize=7B2026.03 | 59.7 | 73.4 | 52.9 | 62 | |
| MEDIC-ADSize=7B2026.03 | 66.3 | 71.4 | 58.8 | 65.5 |