Question Answering on Medbullets-4
71.4AccuracyMeerkat
Evaluation Results
| Method | Links | |
|---|---|---|
| MeerkatModel size=70B2024.03 | 71.4 | |
| GPT-4Shots=5-shot2024.03 | 68.8 | |
| MeerkatModel size=8B2024.03 | 62.3 | |
| OriginalBackbone Model=Lingshu-7B2025.12 | 62.01 | |
| PSIBackbone Model=Lingshu-7B2025.12 | 60.06 | |
| MeerkatModel size=7B2024.03 | 58.7 | |
| ModelMergeBackbone Model=Lingshu-7B2025.12 | 54.22 | |
| BioMistralModel size=7B2024.03 | 52.3 | |
| MediTronModel size=7B2024.03 | 51.1 | |
| GPT-3.5Shots=5-shot2024.03 | 51 | |
| RESTABackbone Model=Lingshu-7B2025.12 | 49.35 | |
| ModelMergeBackbone Model=HuatuoGPT-Vision-7B2025.12 | 46.1 | |
| OriginalBackbone Model=HuatuoGPT-Vision-7B2025.12 | 43.51 | |
| PSIBackbone Model=HuatuoGPT-Vision-7B2025.12 | 41.88 | |
| RESTABackbone Model=HuatuoGPT-Vision-7B2025.12 | 32.47 |