This repository serves as a comprehensive, up-to-date curated list of research papers and developments in Retrieval-Augmented Generation (RAG) for Large Language Models (LLMs). It aims to provide researchers and practitioners with a centralized, organized resource to track the rapidly evolving RAG landscape, facilitating informed decision-making and accelerating innovation in the field.

How It Works

This project functions as a living bibliography, meticulously cataloging recent advancements in RAG. It categorizes papers by conference (e.g., NIPS, EMNLP, ACL, ICML, ICLR), by RAG sub-areas (e.g., Methods & Pipeline, Evaluation Datasets, Applications, Chunk & Database), and by specific findings or techniques. The repository actively encourages community contributions via pull requests to maintain its currency and breadth.

Quick Start & Requirements

This repository is a curated list of research papers and does not involve software installation or execution. It serves as a knowledge base rather than a runnable tool.

Highlighted Details

• Extensive coverage of RAG papers from 2024 and 2025, categorized by major AI conferences.
• Detailed sections on RAG methodologies, evaluation datasets, applications (e.g., medical, recommendation), and specific techniques like graph-enhanced RAG, multimodal RAG, and temporal memory.
• Regular updates highlight the latest research, including code releases and accepted papers from top-tier conferences.
• Includes a "Findings" section summarizing novel techniques and approaches within RAG research.

Maintenance & Community

The project is actively maintained and encourages community contributions through pull requests. Interested parties can contact the maintainers via email at jiangyijcx@163.com for collaboration or updates.

Licensing & Compatibility

The provided README content does not specify a software license. As a curated list of research papers, it does not impose direct licensing restrictions on users, but the underlying research papers themselves will have their own licenses.

Limitations & Caveats

This repository is a dynamic collection of research papers and is explicitly stated to be "still being improved and updated." It does not provide a unified RAG framework or implementation, but rather serves as a pointer to the latest academic work in the domain.