DiReCT: Diagnostic Reasoning for Clinical Notes via Large Language Models (NeurIPS 2024 D&B Track)
-
Updated
Mar 6, 2025 - Python
DiReCT: Diagnostic Reasoning for Clinical Notes via Large Language Models (NeurIPS 2024 D&B Track)
Claude Code skill for multi-persona Bayesian diagnostic reasoning using NEJM Clinical Problem-Solving format with likelihood ratios, evidence-based medicine, and medical AI education
🔥 让AI像专家一样诊断 | 3种创新方法 | 跨领域通用 | 精美可视化
A multi-agent LLM framework for diagnostic, multi-objective calibration of complex simulation models. Pluggable model and LLM layers. Demonstrated on distributed hydrology, designed to generalize.
Code and data to reproduce the main analyses and figures in “Hickam’s Dictum: An Analysis of Multiple Diagnoses” (Journal of General Internal Medicine, published October 28, 2024; DOI: 10.1007/s11606-024-09120-y).
Clinical reasoning likelihood-ratio workflows for LLM diagnostic-reasoning evaluation
Computational illustrations of diagnostic underdetermination, fiber reduction, diagnostic gain, and adaptive clinical inquiry.
DEIC research prototype for hidden-state inference, source reliability, diagnostic querying, and adaptive recovery.
Code and frozen research artifacts for “Large language models generate diagnostic likelihood ratios with low mean bias but wide dispersion” (Scientific Reports, 2026).
To associate your repository with the diagnostic-reasoning topic, visit your repo's landing page and select "manage topics."