Closed-format Question Answering on MSK (Musculoskeletal)
88.8AccuracyHR1.5
Evaluation Results
| Method | Links | |
|---|---|---|
| HR1.52026.07 | 88.8 | |
| HR1.5+2026.07 | 82.2 | |
| HR12026.07 | 80 | |
| GPT-5.42026.07 | 76.6 | |
| Gemini-3-Flash-Preview2026.07 | 68.3 | |
| MedGemma-4B2026.07 | 68.1 | |
| Claude Opus 4.72026.07 | 58.7 | |
| LLaMa-4-Scout2026.07 | 55 | |
| MedGemma-27B2026.07 | 52.8 |