[KongchangAI]
Benchmark

MedGUIDE

A benchmark for evaluating clinical guideline reasoning ability, whose core metrics include weighted accuracy and guideline adherence.

Timeline (last 90 days)

Oct 30

On the MedGUIDE benchmark, GUIDER with a Qwen3-14B backbone achieves 82.5 weighted accuracy and 73.5 guideline compliance.

Unverified50%
Oct 30

On MedGUIDE, GUIDER exceeds the strongest clinical reinforcement learning baseline by 3.3 and 3.1 points in weighted accuracy and guideline compliance, respectively.

Unverified50%

All Facts (2)

Source Articles