Medical Reasoning on Humanity's Last Exam (HLE) Medical
20.8AccuracyMedReason
Evaluation Results
| Method | Links | |
|---|---|---|
| MedReasonBackbone=Qwen2.5-7B-Instruct2026.02 | 20.8 | |
| MedVerseBackbone=LLaMA-3.1-8B-Instruct2026.02 | 20.6 | |
| MedReasonBackbone=LLaMA-3.1-8B-Instruct2026.02 | 20.2 | |
| MedVerseBackbone=Qwen2.5-7B-Instruct2026.02 | 19.6 | |
| Original (Qwen2.5-7B-Instruct)Backbone=Qwen2.5-7B-Instruct2026.02 | 18.4 | |
| Huatuo-o1Backbone=LLaMA-3.1-8B-Instruct2026.02 | 14.6 | |
| Original (LLaMA-3.1-8B-Instruct)Backbone=LLaMA-3.1-8B-Instruct2026.02 | 13.6 |