Healthcare QA on HealthBench 500
41.16Rubric ScorePOP
Evaluation Results
| Method | Links | |
|---|---|---|
| POPBackbone=Qwen-2.5-7B-Inst2026.04 | 41.16 | |
| Qwen-2.5-7B-InstBackbone=Qwen-2.5-7B-Inst2026.04 | 40.83 | |
| POPBackbone=Qwen-2.5-7B2026.04 | 29.97 | |
| Qwen-2.5-7BBackbone=Qwen-2.5-7B2026.04 | 26.36 | |
| Train on DBackbone=Qwen-2.5-7B-Inst2026.04 | 26.14 | |
| Train on DBackbone=Qwen-2.5-7B2026.04 | 23.89 |