Blinded A/B preference on Clinician Validation Spine cohort (R1)
94.2Win RateCLR-voyance-8B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CLR-voyance-8Bbaseline_comparison=HuatuoGPT-o1-7B2026.05 | 94.2 | 88 | |
| CLR-voyance-8Bbaseline_comparison=Qwen3-8B (base)2026.05 | 89.5 | 82 | |
| CLR-voyance-8Bbaseline_comparison=GPT-52026.05 | 85.6 | 78 |