AI BENCHMARK PROFILE

MedDDC-Eval

Health & Life SciencesKnowledge & Reasoning

MedDDC-Eval evaluates multi-turn medical consultation agents by decoupling diagnosis from the diagnostic reader, using a frozen shared reader to score policies. It reports diagnostic support, coverage, and efficiency across Record and Dialogue splits.

Released: 2026-07-21
Readiness: Paper only
Primary field: Health & Life Sciences

Why it matters

Coupled evaluation confounds history elicitation with terminal diagnosis. This testbed enables fair comparison and evaluation-informed policy optimization, providing a more reliable measure of diagnostic support.

Motivation

Evaluating multi-turn medical consultation agents requires judging the diagnostic support provided by the histories they elicit through interaction.

Primary resources

Paper ↗Hugging Face ↗

Benchmark Radar records only publicly supported details and links back to primary sources for verification.