Visual Question Answering on PMC-MI-Bench (test)
15BLEU@4M3LLM-8B
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| M3LLM-8BParameters=8B2025.11 | 15 | 37.8 | 70.1 | 78.2 | |
| Lingshu-7BParameters=7B2025.11 | 9.8 | 30.4 | 66.8 | 73.9 | |
| HealthGPT-14BParameters=14B2025.11 | 9.3 | 30.8 | 66.3 | 73.7 | |
| HuatuoGPT-Vision-7BParameters=7B2025.11 | 9.1 | 31.4 | 66.9 | 74.7 | |
| QWen2.5-VL-7BParameters=7B2025.11 | 8.5 | 29.5 | 64.7 | 73.9 | |
| LLaVA-Med-7BParameters=7B2025.11 | 5.8 | 23.7 | 58.8 | 63 | |
| LLaVA-NeXT-7BParameters=7B2025.11 | 4.5 | 26.5 | 59.3 | 68.7 | |
| LLaVA-7BParameters=7B2025.11 | 3.9 | 27.1 | 58.3 | 68.2 | |
| InternVL3-8BParameters=8B2025.11 | 3.8 | 22.5 | 55.7 | 71.4 | |
| MedGemma-27BParameters=27B2025.11 | 3.4 | 26.7 | 62.5 | 74.2 |