Text-only Question Answering on PMC-MI-Bench
13BLEU@4M3LLM-8B
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| M3LLM-8BParameters=8B2025.11 | 13 | 38.5 | 73.4 | 86.4 | |
| HealthGPT-14BParameters=14B2025.11 | 11.8 | 39 | 69 | 84.2 | |
| LLaVA-Med-7BParameters=7B2025.11 | 11.5 | 36.3 | 69.6 | 83.5 | |
| HuatuoGPT-Vision-7BParameters=7B2025.11 | 11.3 | 37.9 | 69.1 | 83.9 | |
| Lingshu-7BParameters=7B2025.11 | 11 | 38.8 | 68.3 | 85.3 | |
| QWen2.5-VL-7BParameters=7B2025.11 | 10.9 | 37.7 | 70.3 | 84.4 | |
| LLaVA-NeXT-7BParameters=7B2025.11 | 9.8 | 35.2 | 68.3 | 83.2 | |
| LLaVA-7BParameters=7B2025.11 | 9.3 | 32.4 | 66.1 | 81.9 | |
| InternVL3-8BParameters=8B2025.11 | 9.1 | 35.4 | 68.3 | 83.3 | |
| MedGemma-27BParameters=27B2025.11 | 7.1 | 31.6 | 65.3 | 82.3 |