Reproduce “Auditable Clinical Support with Decision Agents” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Tiancheng Bao, Riga Wu, Jan Wienhold, Sebastian Billig, Patrick Winnersbach, Hsuan-Chih Chen, Akis Giannoukos, Jiayang Li and 8 more

Decision2Care makes clinical work-ups auditable by asking a model to assign probabilities to predefined options and using a fixed controller to select actions, while logging the questions, clinical state, scores, and rules. Across 546 retrospective emergency admissions, mean diagnostic accuracy was 86.5%, versus 86.1% for one MIRA-v2 configuration. This does not establish prospective clinical benefit or safety.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.