19 open-source projects similar to nju-rl/ma-rag, ranked by shared indexed features. Tags may describe platforms or build tools rather than the same primary purpose. Check each project’s use case, license, and deployment requirements before treating it as a replacement.
2026-06-03 We add demo inference entry. Try your custom question with bash rundemoinference.sh. - 2026-03-26 We release code and paper: Unified-MAS: Universally Generating Domain-Specific Nodes for Empowering Automatic Multi-Agent Systems.
This is the code repository for paper: KERAP: A Knowledge-Enhanced Reasoning Approach for Accurate Zero-shot Diagnosis Prediction Using Multi-agent LLMs and EHR, which aims to enhance LLM's zero-shot diagnosis prediction task utilizing KG through multi-agent framework and multi-stage reasoning…
Automated Drug Combination Extraction (DCE) from large-scale biomedical literature is important for precision medicine and pharmacological research. However, existing extraction methodologies predominantly focus on binary interactions. When tasked with complex, variable-length n-ary…
UCAgents: Unidirectional Convergence for Visual Evidence Anchored Multi-Agent Medical Decision-Making A novel multi-agent framework for medical decision-making, leveraging unidirectional convergence and visual evidence anchoring to enhance accuracy and interpretability in medical Visual Question…
News - Introduction - Key Features - Methodology - Experiments - Setup - Experiment Scripts - Citation
A Self-Evolving Multi-Agent Framework for Medical Multi-Disciplinary Team (MDT) Consultations.
An agentic RL framework to enhance retreival-augmented reasoning in Diagnostic Policy
AMG-RAG (Agentic Medical Graph-RAG) is a comprehensive framework that automates the construction and continuous updating of Medical Knowledge Graphs (MKGs), integrates reasoning, and retrieves current external evidence for medical Question Answering (QA). Our approach addresses the challenge of…
Nithin Sivakumaran | Justin Chih-Yao Chen | David Wan | Yue Zhang | Jaehong Yoon | Elias Stengel-Eskin | Mohit Bansal
ConfAgents is our proposed adaptive multi-agent framework. This project aims to enhance the efficiency and robustness of complex problem-solving by intelligently allocating collaborative resources through a principled, two-stage process.
Meissa is a multi-modal medical agent, built on trajectory-based agentic behavior distillation framework.
Observational studies can yield clinically actionable evidence at scale, but executing them on real-world databases is open-ended and requires coherent decisions across cohort construction, analysis, and reporting. Prior evaluations of LLM agents emphasize isolated steps or single answers,…
repo for Tree-of-Reasoning: Towards Complex Medical Diagnosis via Multi-Agent Reasoning with Evidence Tree
This repository provides a reference implementation of Med-Inquire and EvoClinician, including: