[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-DEEP-PolyU--Awesome-GraphRAG":3,"tool-DEEP-PolyU--Awesome-GraphRAG":64},[4,17,27,35,43,56],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":16},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,3,"2026-04-05T11:01:52",[13,14,15],"开发框架","图像","Agent","ready",{"id":18,"name":19,"github_repo":20,"description_zh":21,"stars":22,"difficulty_score":23,"last_commit_at":24,"category_tags":25,"status":16},1381,"everything-claude-code","affaan-m\u002Feverything-claude-code","everything-claude-code 是一套专为 AI 编程助手（如 Claude Code、Codex、Cursor 等）打造的高性能优化系统。它不仅仅是一组配置文件，而是一个经过长期实战打磨的完整框架，旨在解决 AI 代理在实际开发中面临的效率低下、记忆丢失、安全隐患及缺乏持续学习能力等核心痛点。\n\n通过引入技能模块化、直觉增强、记忆持久化机制以及内置的安全扫描功能，everything-claude-code 能显著提升 AI 在复杂任务中的表现，帮助开发者构建更稳定、更智能的生产级 AI 代理。其独特的“研究优先”开发理念和针对 Token 消耗的优化策略，使得模型响应更快、成本更低，同时有效防御潜在的攻击向量。\n\n这套工具特别适合软件开发者、AI 研究人员以及希望深度定制 AI 工作流的技术团队使用。无论您是在构建大型代码库，还是需要 AI 协助进行安全审计与自动化测试，everything-claude-code 都能提供强大的底层支持。作为一个曾荣获 Anthropic 黑客大奖的开源项目，它融合了多语言支持与丰富的实战钩子（hooks），让 AI 真正成长为懂上",140436,2,"2026-04-05T23:32:43",[13,15,26],"语言模型",{"id":28,"name":29,"github_repo":30,"description_zh":31,"stars":32,"difficulty_score":23,"last_commit_at":33,"category_tags":34,"status":16},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",107662,"2026-04-03T11:11:01",[13,14,15],{"id":36,"name":37,"github_repo":38,"description_zh":39,"stars":40,"difficulty_score":23,"last_commit_at":41,"category_tags":42,"status":16},3704,"NextChat","ChatGPTNextWeb\u002FNextChat","NextChat 是一款轻量且极速的 AI 助手，旨在为用户提供流畅、跨平台的大模型交互体验。它完美解决了用户在多设备间切换时难以保持对话连续性，以及面对众多 AI 模型不知如何统一管理的痛点。无论是日常办公、学习辅助还是创意激发，NextChat 都能让用户随时随地通过网页、iOS、Android、Windows、MacOS 或 Linux 端无缝接入智能服务。\n\n这款工具非常适合普通用户、学生、职场人士以及需要私有化部署的企业团队使用。对于开发者而言，它也提供了便捷的自托管方案，支持一键部署到 Vercel 或 Zeabur 等平台。\n\nNextChat 的核心亮点在于其广泛的模型兼容性，原生支持 Claude、DeepSeek、GPT-4 及 Gemini Pro 等主流大模型，让用户在一个界面即可自由切换不同 AI 能力。此外，它还率先支持 MCP（Model Context Protocol）协议，增强了上下文处理能力。针对企业用户，NextChat 提供专业版解决方案，具备品牌定制、细粒度权限控制、内部知识库整合及安全审计等功能，满足公司对数据隐私和个性化管理的高标准要求。",87618,"2026-04-05T07:20:52",[13,26],{"id":44,"name":45,"github_repo":46,"description_zh":47,"stars":48,"difficulty_score":23,"last_commit_at":49,"category_tags":50,"status":16},2268,"ML-For-Beginners","microsoft\u002FML-For-Beginners","ML-For-Beginners 是由微软推出的一套系统化机器学习入门课程，旨在帮助零基础用户轻松掌握经典机器学习知识。这套课程将学习路径规划为 12 周，包含 26 节精炼课程和 52 道配套测验，内容涵盖从基础概念到实际应用的完整流程，有效解决了初学者面对庞大知识体系时无从下手、缺乏结构化指导的痛点。\n\n无论是希望转型的开发者、需要补充算法背景的研究人员，还是对人工智能充满好奇的普通爱好者，都能从中受益。课程不仅提供了清晰的理论讲解，还强调动手实践，让用户在循序渐进中建立扎实的技能基础。其独特的亮点在于强大的多语言支持，通过自动化机制提供了包括简体中文在内的 50 多种语言版本，极大地降低了全球不同背景用户的学习门槛。此外，项目采用开源协作模式，社区活跃且内容持续更新，确保学习者能获取前沿且准确的技术资讯。如果你正寻找一条清晰、友好且专业的机器学习入门之路，ML-For-Beginners 将是理想的起点。",84991,"2026-04-05T10:45:23",[14,51,52,53,15,54,26,13,55],"数据工具","视频","插件","其他","音频",{"id":57,"name":58,"github_repo":59,"description_zh":60,"stars":61,"difficulty_score":10,"last_commit_at":62,"category_tags":63,"status":16},3128,"ragflow","infiniflow\u002Fragflow","RAGFlow 是一款领先的开源检索增强生成（RAG）引擎，旨在为大语言模型构建更精准、可靠的上下文层。它巧妙地将前沿的 RAG 技术与智能体（Agent）能力相结合，不仅支持从各类文档中高效提取知识，还能让模型基于这些知识进行逻辑推理和任务执行。\n\n在大模型应用中，幻觉问题和知识滞后是常见痛点。RAGFlow 通过深度解析复杂文档结构（如表格、图表及混合排版），显著提升了信息检索的准确度，从而有效减少模型“胡编乱造”的现象，确保回答既有据可依又具备时效性。其内置的智能体机制更进一步，使系统不仅能回答问题，还能自主规划步骤解决复杂问题。\n\n这款工具特别适合开发者、企业技术团队以及 AI 研究人员使用。无论是希望快速搭建私有知识库问答系统，还是致力于探索大模型在垂直领域落地的创新者，都能从中受益。RAGFlow 提供了可视化的工作流编排界面和灵活的 API 接口，既降低了非算法背景用户的上手门槛，也满足了专业开发者对系统深度定制的需求。作为基于 Apache 2.0 协议开源的项目，它正成为连接通用大模型与行业专有知识之间的重要桥梁。",77062,"2026-04-04T04:44:48",[15,14,13,26,54],{"id":65,"github_repo":66,"name":67,"description_en":68,"description_zh":69,"ai_summary_zh":69,"readme_en":70,"readme_zh":71,"quickstart_zh":72,"use_case_zh":73,"hero_image_url":74,"owner_login":75,"owner_name":76,"owner_avatar_url":77,"owner_bio":78,"owner_company":79,"owner_location":79,"owner_email":79,"owner_twitter":79,"owner_website":80,"owner_url":81,"languages":79,"stars":82,"forks":83,"last_commit_at":84,"license":85,"difficulty_score":86,"env_os":78,"env_gpu":87,"env_ram":87,"env_deps":88,"category_tags":90,"github_topics":91,"view_count":10,"oss_zip_url":79,"oss_zip_packed_at":79,"status":16,"created_at":99,"updated_at":100,"faqs":101,"releases":120},1212,"DEEP-PolyU\u002FAwesome-GraphRAG","Awesome-GraphRAG","Awesome-GraphRAG: A curated list of resources (surveys, papers, benchmarks, and opensource projects) on graph-based retrieval-augmented generation. ","Awesome-GraphRAG 是一个精心整理的资源库，专注于基于图的检索增强生成（GraphRAG）技术。它汇集了综述、论文、基准测试和开源项目，帮助用户快速掌握GraphRAG的最新进展。传统RAG在处理实体关系和多跳推理时效率较低，而GraphRAG通过图结构表示知识（如实体关系和领域层次），支持更精准的检索和推理，解决了领域特定大语言模型中的知识碎片化问题。这个资源库特别适合AI研究人员、开发者和工程师使用，能高效整合GraphRAG相关工具和方法。亮点包括其分类体系基于权威综述论文，以及整合了已被ICLR收录的GraphRAG Benchmark和LinearRAG等实用工具。持续更新，是探索高效知识检索系统的必备指南。","# Awesome-GraphRAG (GraphRAG Survey)\n\n\u003Cdiv align=\"center\">\n    \u003Ca href=\"https:\u002F\u002Fawesome.re\">\u003Cimg src=\"https:\u002F\u002Fawesome.re\u002Fbadge.svg\"\u002F>\u003C\u002Fa>\n    \u003Ca href=\"http:\u002F\u002Fmakeapullrequest.com\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FPRs-welcome-green.svg\"\u002F>\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Farxiv.org\u002Fabs\u002F2501.13958\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FPaper-Arxiv-red?logo=arxiv&style=flat-square\" alt=\"arXiv:2506.08938\">\u003C\u002Fa>\n    \u003Ca href=\"http:\u002F\u002Fmakeapullrequest.com\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flast-commit\u002FDEEP-PolyU\u002FAwesome-GraphRAG?color=blue\"\u002F>\u003C\u002Fa>\n    \u003Ca href=\"http:\u002F\u002Fmakeapullrequest.com\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FDEEP-PolyU\u002FAwesome-GraphRAG\"\u002F>\u003C\u002Fa>\n\u003C\u002Fdiv>\n\n\nThis repository contains a curated list of resources on graph-based retrieval-augmented generation (GraphRAG), which are classified according to \"[**A Survey of Graph Retrieval-Augmented Generation for Customized Large Language Models**](https:\u002F\u002Farxiv.org\u002Fabs\u002F2501.13958)\". Continuously updating, stay tuned!\n\n\n**📃 Please [cite our paper](#-citation)** if you find our survey or repository helpful!\n\n\n# 🎉 News\n- **[2026-01-26]** Our **[GraphRAG Benchmark](https:\u002F\u002Fgithub.com\u002FGraphRAG-Bench\u002FGraphRAG-Benchmark)** is accepted by ICLR’26. \n- **[2026-01-26]** Our **[LinearRAG](https:\u002F\u002Fgithub.com\u002FDEEP-PolyU\u002FLinearRAG)** is accepted by ICLR’26.\n- **[2025-11-08]** Our **[LogicRAG](https:\u002F\u002Fgithub.com\u002FchensyCN\u002FLogicRAG.git)** is accepted by AAAI'26.\n- **[2025-10-27]** We release **[LinearRAG](https:\u002F\u002Fgithub.com\u002FDEEP-PolyU\u002FLinearRAG)**, a relation-free graph construction method for efficient GraphRAG.\n- **[2025-06-06]** We release the **[GraphRAG Benchmark](https:\u002F\u002Fgithub.com\u002FGraphRAG-Bench\u002FGraphRAG-Benchmark.git)** for evaluating GraphRAG models.\n- **[2025-05-14]** We release the [GraphRAG Benchmark dataset](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002FGraphRAG-Bench\u002FGraphRAG-Bench).\n- **[2025-01-21]** We release the [GraphRAG survey](https:\u002F\u002Fgithub.com\u002FDEEP-PolyU\u002FAwesome-GraphRAG).\n\n---\n\n\u003Cdiv>\n\u003Ch3 align=\"left\">\n       \u003Cp align=\"center\">\u003Cimg width=\"100%\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FDEEP-PolyU_Awesome-GraphRAG_readme_22be195ae39e.png\" \u002F>\u003C\u002Fp>\n    \u003Cp align=\"center\">\u003Cem>Overview of traditional RAG and two typical GraphRAG workflows. \u003C\u002Fem>\u003C\u002Fp>\n\u003C\u002Fdiv>\n\n- **Non-graph RAG** organizes the\ncorpus into chunks, ranks them by similarity, and retrieves the most relevant text for generating responses.\n- **Knowledge-based\nGraphRAG** extracts detailed knowledge graphs from the corpus using entity recognition and relation extraction, offering\nfine-grained, domain-specific information.\n- **Index-based GraphRAG** summarizes the corpus into high-level topic nodes, which\nare linked to form an index graph, while the fact linking maps topics to text.\n\n---\n\n# RAG vs. GraphRAG\nGraphRAG is a new paradigm of RAG that revolutionizes domain-specific LLM applications, by addressing traditional RAG limitations through three key innovations: **(i) graph-structured knowledge representation** that explicitly captures\nentity relationships and domain hierarchies, **(ii) graph-aware retrieval mechanisms** that enable multi-hop reasoning and context-preserving knowledge acquisition, and **(iii) structure-guided\nknowledge search algorithms** that ensure efficient retrieval across large-scale corpora.\n    \n\n\u003Ch3 align=\"center\">\n   \u003Cp align=\"center\">\u003Cimg width=\"100%\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FDEEP-PolyU_Awesome-GraphRAG_readme_b0d1526bdca9.png\" \u002F>\u003C\u002Fp>\n    \u003Cp align=\"center\">\u003Cem>Comparison between traditional RAG and GraphRAG.\u003C\u002Fem>\u003C\u002Fp>\n\n\n\n# 📫 Contact Us\nWe welcome researchers to share related work to enrich this list or provide insightful comments on our survey. Feel free to reach out to the corresponding co-first authors: [Qinggang Zhang](https:\u002F\u002Fqing145.github.io\u002F), [Shengyuan Chen](https:\u002F\u002Fchensycn.github.io\u002F).\n\n\n## Table of Content\n- [🍀 Citation](#-citation)\n- [📫 Contact Us](#-contact-us)\n- [📈 Trend of GraphRAG Research](#-trend-of-graphrag-research)\n- [📜 Research Papers](#-research-papers)\n    - [Knowledge Organization](#knowledge-organization)\n        - [Graph for Knowledge Indexing](#graphs-for-knowledge-indexing)\n        - [Graph as Knowledge Carrier](#graphs-as-knowledge-carrier)\n            - [Knowledge Graph Construction from Corpus](#knowledge-graph-construction-from-corpus)\n            - [GraphRAG with Existing KGs](#graphrag-with-existing-kgs)\n        - [Hybrid GraphRAG](#hybrid-graphrag)\n    - [Knowledge Retrieval](#knowledge-retrieval)\n        - [Semantics Similarity-based Retriever](#semantics-similarity-based-retriever)\n        - [Logical Reasoning-based Retriever](#logical-reasoning-based-retriever)\n        - [LLM-based Retriever](#llm-based-retriever)\n        - [GNN-based Retriever](#gnn-based-retriever)\n        - [Multi-round Retriever](#multi-round-retriever)\n        - [Post-retrieval](#post-retrieval)\n        - [Hybrid Retriever](#hybrid-retriever)\n    - [Knowledge Integration](#knowledge-integration)\n        - [Fine-tuning](#fine-tuning)\n            - [Fine-tuning with Node-level Knowledge](#fine-tuning-with-node-level-knowledge)\n            - [Fine-tuning with Path-level Knowledge](#fine-tuning-with-path-level-knowledge)\n            - [Fine-tuning with Subgraph-level Knowledge](#fine-tuning-with-subgraph-level-knowledge)\n        - [In-context Learning](#in-context-learning)\n            - [Graph-enhanced Chain-of-Thought](#graph-enhanced-chain-of-thought)\n            - [Collaborative Knowledge Graph Refinement](#collaborative-knowledge-graph-refinement)\n- [📚 Related Survey Papers](#-related-survey-papers)\n- [🏆 Benchmarks](#-benchmarks)\n- [💻 Open-source Projects](#-open-source-projects)\n\n\n# 📈 Trend of GraphRAG Research\n\n\u003Ch3 align=\"center\">\n   \u003Cp align=\"center\">\u003Cimg width=\"100%\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FDEEP-PolyU_Awesome-GraphRAG_readme_4610c437cc87.png\" \u002F>\u003C\u002Fp>\n    \u003Cp align=\"center\">\u003Cem>The development trends in the field of GraphRAG with representative works.\u003C\u002Fem>\u003C\u002Fp>\n\n# 📜 Research Papers\n## Knowledge Organization\n\n### Graphs for Knowledge Indexing\n- (arXiv 2025) **LinearRAG: Linear Graph Retrieval Augmented Generation on Large-scale Corpora**  [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2510.10114)\n- (EMNLP 2025) **Don’t Forget the Base Retriever! A Low-Resource Graph-based Retriever for Multi-hop Question Answering** [[Paper]](https:\u002F\u002Faclanthology.org\u002F2025.emnlp-industry.174\u002F)\n- (arXiv 2025) **Query-Centric Graph Retrieval Augmented Generation** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2509.22009)\n- (arXiv 2025) **Multi-Agent GraphRAG: A Text-to-Cypher Framework for Labeled Property Graphs** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2510.09156)\n- (arXiv 2025) **Grounded by Experience: Generative Healthcare Prediction Augmented with Hierarchical Agentic Retrieval** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2511.13293)\n- (ICML 2025) **HippoRAG2: From RAG to Memory: Non-Parametric Continual Learning for Large Language Models** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2502.14802)\n- (arXiv 2025) **PersonaAgent with GraphRAG: Community-Aware Knowledge Graphs for Personalized LLM** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2511.17467)\n- (arXiv 2025) **E^2GraphRAG: Streamlining Graph-based RAG for High Efficiency and Effectiveness** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2505.24226)\n- (arXiv 2025) **DIGIMON: A unified and modular graph-based RAG framework** [[Paper]](https:\u002F\u002Fgithub.com\u002FJayLZhou\u002FGraphRAG.git)\n- (arXiv 2025) **ArchRAG: Attributed Community-based Hierarchical Retrieval-Augmented Generation** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2502.09891)\n- (arXiv 2025) **KET-RAG: A Cost-Efficient Multi-Granular Indexing Framework for Graph-RAG** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2502.09304)\n- (arXiv 2025) **PIKE-RAG: sPecIalized KnowledgE and Rationale Augmented Generation** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2501.11551)\n- (EMNLP 2025 Findings) **Retrieval-Augmented Generation with Hierarchical Knowledge** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2503.10150)\n- (arXiv 2024) **Graph Neural Network Enhanced Retrieval for Question Answering of LLMs** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2406.06572)\n- (arXiv 2024) **KAG: Boosting LLMs in Professional Domains via Knowledge Augmented Generation** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2409.13731)\n- (arXiv 2024) **OG-RAG: Ontology-Grounded Retrieval-Augmented Generation For Large Language Models** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2412.15235)\n- (arXiv 2024) **GRAG: Graph Retrieval-Augmented Generation** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2405.16506)\n- (arXiv 2024) **Empowering Large Language Models to Set up a Knowledge Retrieval Indexer via Self-Learning** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2405.16933)\n- (ICLR 2024) **RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrieval** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2401.18059)\n- (AAAI 2024) **Knowledge graph prompting for multi-document question answering** [[Paper]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1609\u002Faaai.v38i17.29889)\n- (arXiv 2024) **GraphCoder: Enhancing Repository-Level Code Completion via Code Context Graph-based Retrieval and Language Model** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2406.07003)\n- (NeurIPS 2023) **Avis: Autonomous visual information seeking with large language model agent** [[Paper]](https:\u002F\u002Fopenreview.net\u002Fforum?id=7EMphtUgCI&noteId=yGw4rbGozi)\n- (CoRL 2023) **Sayplan: Grounding large language models using 3d scene graphs for scalable robot task planning** [[Paper]](https:\u002F\u002Fproceedings.mlr.press\u002Fv229\u002Frana23a\u002Frana23a.pdf)\n- (arXiv 2020) **Answering complex open-domain questions with multi-hop dense retrieval** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2009.12756)\n- (arXiv 2019) **Knowledge guided text retrieval and reading for open domain question answering** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F1911.03868)\n\n### Graphs as Knowledge Carrier\n#### Knowledge Graph Construction from Corpus\n- (AAAI 2026) **You Don’t Need Pre-built Graphs for RAG: Retrieval Augmented Generation with Adaptive Reasoning Structures** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2508.06105)\n- (arXiv 2025) **AutoGraph-R1: End-to-End Reinforcement Learning for Knowledge Graph Construction** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2510.15339)\n- (arXiv 2025) **AGRAG: Advanced Graph-based Retrieval-Augmented Generation for LLMs** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2511.05549)\n- (EMNLP 2025) **MaGiX: A Multi-Granular Adaptive Graph Intelligence Framework for Enhancing Cross-Lingual RAG** [[Paper]](https:\u002F\u002Faclanthology.org\u002Fanthology-files\u002Fanthology-files\u002Fpdf\u002Ffindings\u002F2025.findings-emnlp.279.pdf)\n- (CIKM 2025) **Context-Aware Fine-Grained Graph RAG for Query-Focused Summarization** [[Paper]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3746252.3760935)\n- (CIKM 2025) **DocPolicyKG: A Lightweight LLM-Based Framework for Knowledge Graph Construction from Chinese Policy Documents** [[Paper]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fabs\u002F10.1145\u002F3746252.3760904)\n- (arXiv 2025) **SUBQRAG: SUB-QUESTION DRIVEN DYNAMIC GRAPH RAG** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2510.07718)\n- (arXiv 2025) **Ontology Learning and Knowledge Graph Construction: A Comparison of Approaches and Their Impact on RAG Performance** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2511.05991)\n- (NeurIPS 2025) **GFM-RAG: Graph Foundation Model for Retrieval Augmented Generation** [Paper](https:\u002F\u002Farxiv.org\u002Fabs\u002F2502.01113)\n- (arXiv 2025) **G-reasoner: Foundation Models for Unified Reasoning over Graph-structured Knowledge** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2509.24276)\n- (CVPR 2025) **Medical Graph RAG: Towards Safe Medical Large Language Model via Graph Retrieval-Augmented Generation** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2408.04187)\n- (arXiv 2025) **Youtu-GraphRAG: Vertically Unified Agents for Graph Retrieval-Augmented Complex Reasoning** [[Paper]](https:\u002F\u002Fwww.arxiv.org\u002Fabs\u002F2508.19855)\n- (arXiv 2025) **Retrieval-Augmented Generation with Hierarchical Knowledge** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2503.10150)\n- (arXiv 2025) **MedRAG: Enhancing Retrieval-augmented Generation with Knowledge Graph-Elicited Reasoning for Healthcare Copilot** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2502.04413)\n- (arXiv 2025) **PathRAG: Pruning Graph-based Retrieval Augmented Generation with Relational Paths** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2502.14902)\n- (EDBT 2025) **DBCopilot: Natural Language Querying over Massive Databases via Schema Routing** [[Paper]](https:\u002F\u002Fopenproceedings.org\u002F2025\u002Fconf\u002Fedbt\u002Fpaper-209.pdf)\n- (arXiv 2024) **From local to global: A graph rag approach to query-focused summarization** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2404.16130)\n- (EMNLP 2024) **Structure Guided Prompt: Instructing Large Language Model in Multi-Step Reasoning by Exploring Graph Structure of the Text** [[Paper]](https:\u002F\u002Faclanthology.org\u002F2024.emnlp-main.528.pdf)\n- (EMNLP 2024 Findings) **GraphReader: Building Graph-based Agent to Enhance Long-Context Abilities of Large Language Models** [[Paper]](https:\u002F\u002Faclanthology.org\u002F2024.findings-emnlp.746\u002F)\n- (SIGIR 2024) **Retrieval-augmented generation with knowledge graphs for customer service question answering** [[Paper]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fabs\u002F10.1145\u002F3626772.3661370)\n- (arXiv 2024) **DynaGRAG | Exploring the Topology of Information for Advancing Language Understanding and Generation in Graph Retrieval-Augmented Generation** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2412.18644)\n- (arXiv 2024) **FastRAG: Retrieval Augmented Generation for Semi-structured Data** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2411.13773)\n- (TechRxiv 2024) **LuminiRAG: Vision-Enhanced Graph RAG for Complex Multi-Modal Document Understanding** [[Paper]](https:\u002F\u002Fwww.techrxiv.org\u002Fusers\u002F867713\u002Farticles\u002F1248304-luminirag-vision-enhanced-graph-rag-for-complex-multi-modal-document-understanding)\n- (BigData 2023) **AutoKG: Efficient automated knowledge graph generation for language models** [[Paper]](https:\u002F\u002Fieeexplore.ieee.org\u002Fabstract\u002Fdocument\u002F10386454)\n- (ACL 2019) **Using Local Knowledge Graph Construction to Scale Seq2Seq Models to Multi-Document Inputs** [[Paper]](https:\u002F\u002Faclanthology.org\u002FD19-1428.pdf)\n- (SIGIR 2019) **Answering complex questions by joining multi-document evidence with quasi knowledge graphs** [[Paper]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3331184.3331252)\n\n#### GraphRAG with Existing KGs\n- (arXiv 2025) **GraphSearch: An Agentic Deep Searching Workflow for Graph Retrieval-Augmented Generation** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2509.22009)\n- (arXiv 2025) **Detecting Hallucinations in Graph Retrieval-Augmented Generation via Attention Patterns and Semantic Alignment** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2512.09148)\n- （arXiv 2025） **Inference Scaled GraphRAG: Improving Multi Hop Question Answering on Knowledge Graphs** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2506.19967)\n- (AAAI 2025) **LightPROF: A Lightweight Reasoning Framework for Large Language Model on Knowledge Graph** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2504.03137)\n- (ICLR 2025) **Simple is Effective: The Roles of Graphs and Large Language Models in Knowledge-Graph-Based Retrieval-Augmented Generation** [[Paper]](https:\u002F\u002Fopenreview.net\u002Fforum?id=JvkuZZ04O7)\n- (arXiv 2025) **Empowering GraphRAG with Knowledge Filtering and Integration** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2503.13804)\n- (arXiv 2024)**StructRAG: Boosting Knowledge Intensive Reasoning of LLMs via Inference-time Hybrid Information Structurization** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2410.08815)\n- (ICLR 2024) **Reasoning on Graphs: Faithful and Interpretable Large Language Model Reasoning** [[Paper]](https:\u002F\u002Fopenreview.net\u002Fforum?id=ZGNWW7xZ6Q)\n- (AAAI 2024) **Mitigating large language model hallucinations via autonomous knowledge graph-based retrofitting** [[Paper]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1609\u002Faaai.v38i16.29770)\n- (ICLR 2024) **Think-on-Graph: Deep and Responsible Reasoning of Large Language Model on Knowledge Graph** [[Paper]](https:\u002F\u002Fopenreview.net\u002Fforum?id=nnVO1PvbTv)\n- (Bioinformatics 2024) **Biomedical knowledge graph-enhanced prompt generation for large language models** [[Paper]](https:\u002F\u002Facademic.oup.com\u002Fbioinformatics\u002Farticle\u002F40\u002F9\u002Fbtae560\u002F7759620)\n- (NeurIPS 2024) **KnowGPT: Knowledge Graph based PrompTing for Large Language Models** [[Paper]](https:\u002F\u002Fopenreview.net\u002Fforum?id=PacBluO5m7&referrer=%5Bthe%20profile%20of%20Daochen%20Zha%5D(%2Fprofile%3Fid%3D~Daochen_Zha1))\n- (ACL 2024 Findings) **Knowledge Graph-Enhanced Large Language Models via Path Selection** [[Paper]](https:\u002F\u002Faclanthology.org\u002F2024.findings-acl.376\u002F)\n- (IEEE VIS 2024) **KNOWNET: Guided Health Information Seeking from LLMs via Knowledge Graph Integration** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2407.13598)\n- (CoLM 2024) **ProLLM: Protein Chain-of-Thoughts Enhanced LLM for Protein-Protein Interaction Prediction** [[Paper]](https:\u002F\u002Fopenreview.net\u002Fforum?id=2nTzomzjjb#discussion)\n- (arXiv 2024) **LEGO-GraphRAG: Modularizing Graph-based Retrieval-Augmented Generation for Design Space Exploration** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2411.05844)\n- (arXiv 2024) **Think-on-Graph 2.0: Deep and Faithful Large Language Model Reasoning with Knowledge-guided Retrieval Augmented Generation** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2407.10805)\n\n### Hybrid GraphRAG\n- (NAACL 2025) **Knowledge Graph-Guided Retrieval Augmented Generation** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2502.06864)\n- (ACL 2024 Findings) **HybGRAG: Hybrid Retrieval-Augmented Generation on Textual and Relational Knowledge Bases**[[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2412.16311)\n- (arXiv 2024) **Graph of Records: Boosting Retrieval Augmented Generation for Long-context Summarization with Graphs** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2410.11001)\n- (arXiv 2024) **Medical graph rag: Towards safe medical large language model via graph retrieval-augmented generation** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2408.04187)\n- (arXiv 2024) **Codexgraph: Bridging large language models and code repositories via code graph databases** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2408.03910)\n\n## Knowledge Retrieval\n\n### Semantics Similarity-based Retriever\n- (AAAI 2024) **StructuGraphRAG: Structured Document-Informed Knowledge Graphs for Retrieval-Augmented Generation** [[Paper]](https:\u002F\u002Fojs.aaai.org\u002Findex.php\u002FAAAI-SS\u002Farticle\u002Fview\u002F31798\u002F33965)\n- (arXiv 2024) **G-Retriever: Retrieval-Augmented Generation for Textual Graph Understanding and Question Answering** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2402.07630)\n- (arXiv 2024) **CancerKG.ORG A Web-scale, Interactive, Verifiable Knowledge Graph-LLM Hybrid for Assisting with Optimal Cancer Treatment and Care** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2501.00223)\n- (arXiv 2024) **Empowering Large Language Models to Set up a Knowledge Retrieval Indexer via Self-Learning** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2405.16933)\n- (arXiv 2024) **GraphCoder: Enhancing Repository-Level Code Completion via Code Context Graph-based Retrieval and Language Model** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2406.07003)\n- (arXiv 2024) **Medical Graph RAG: Towards Safe Medical Large Language Model via Graph Retrieval-Augmented Generation** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2408.04187)\n- (arXiv 2024) **How to Make LLMs Strong Node Classifiers?** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2410.02296)\n\n### Logical Reasoning-based Retriever\n- (AAAI 2026) **You Don’t Need Pre-built Graphs for RAG: Retrieval Augmented Generation with Adaptive Reasoning Structures** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2508.06105)\n- (NeurIPS 2024) **KnowGPT: Knowledge Graph based PrompTing for Large Language Models** [[Paper]](https:\u002F\u002Fopenreview.net\u002Fforum?id=PacBluO5m7&referrer=%5Bthe%20profile%20of%20Daochen%20Zha%5D(%2Fprofile%3Fid%3D~Daochen_Zha1))\n- (ACL 2024 Findings) **Knowledge Graph-Enhanced Large Language Models via Path Selection** [[Paper]](https:\u002F\u002Faclanthology.org\u002F2024.findings-acl.376\u002F)\n- (ICLR 2024) **Think-on-Graph: Deep and Responsible Reasoning of Large Language Model on Knowledge Graph** [[Paper]](https:\u002F\u002Fopenreview.net\u002Fforum?id=nnVO1PvbTv)\n- (CIKM 2024) **RD-P: A Trustworthy Retrieval-Augmented Prompter with Knowledge Graphs for LLMs** [[Paper]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3627673.3679659)\n- (arXiv 2024) **RuleRAG: Rule-Guided Retrieval-Augmented Generation with Language Models for Question Answering** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2410.22353)\n- (LHB 2024) **Intelligent question answering for water conservancy project inspection driven by knowledge graph and large language model collaboration** [[Paper]](https:\u002F\u002Fwww.tandfonline.com\u002Fdoi\u002Ffull\u002F10.1080\u002F27678490.2024.2397337)\n- (arXiv 2024) **RiTeK: A Dataset for Large Language Models Complex Reasoning over Textual Knowledge Graphs** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2410.13987)\n\n### LLM-based Retriever\n- (AAAI 2024) **Knowledge graph prompting for multi-document question answering** [[Paper]](https:\u002F\u002Fwww.overleaf.com\u002Fproject\u002F667419080bc7191bc75f5880)\n- (EMNLP 2024) **Structure Guided Prompt: Instructing Large Language Model in Multi-Step Reasoning by Exploring Graph Structure of the Text** [[Paper]](https:\u002F\u002Faclanthology.org\u002F2024.emnlp-main.528.pdf)\n- (ACML 2024) **Enhancing Textbook Question Answering with Knowledge Graph-Augmented Large Language Models** [[Paper]](https:\u002F\u002Fopenreview.net\u002Fforum?id=ATiIqCCqR2)\n- (ICLR 2024) **Think-on-Graph: Deep and Responsible Reasoning of Large Language Model on Knowledge Graph** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2307.07697)\n- (arXiv 2024) **LightRAG: Simple and Fast Retrieval-Augmented Generation** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2410.05779)\n- (arXiv 2024) **MEG: Medical Knowledge-Augmented Large Language Models for Question Answering** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2411.03883)\n- (arXiv 2024) **From local to global: A graph rag approach to query-focused summarization** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2404.16130)\n\n### GNN-based Retriever\n- (arXiv 2025) **CG-RAG: Research Question Answering by Citation Graph Retrieval-Augmented LLMs** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2501.15067)\n- (arXiv 2024) **Advanced RAG Models with Graph Structures: Optimizing Complex Knowledge Reasoning and Text Generation** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2411.03572)\n- (arXiv 2024) **Language Models are Graph Learners** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2410.02296)\n- (arXiv 2024) **Graph Neural Network Enhanced Retrieval for Question Answering of LLMs** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2406.06572)\n- (arXiv 2024) **Knowledge Graph-Augmented Language Models for Knowledge-Grounded Dialogue Generation** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2305.18846)\n\n\n### Multi-round Retriever\n- (arXiv 2024) **Graph Chain-of-Thought: Augmenting Large Language Models by Reasoning on Graphs** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2404.07103)\n- (arXiv 2024) **Generative Subgraph Retrieval for Knowledge Graph-Grounded Dialog Generation** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2410.09350)\n- (arXiv 2024) **Graph of Records: Boosting Retrieval Augmented Generation for Long-context Summarization with Graphs** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2410.11001)\n### Post-retrieval \n- (ACL 2024) **Boosting Language Models Reasoning with Chain-of-Knowledge Prompting** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2306.06427)\n- (ACL 2024 Findings) **Call Me When Necessary: LLMs can Efficiently and Faithfully Reason over Structured Environments** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2403.08593)\n- (arXiv 2024) **Graph-constrained Reasoning: Faithful Reasoning on Knowledge Graphs with Large Language Models** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2410.13080)\n- (arXiv 2024) **Mitigating Large Language Model Hallucinations via Autonomous Knowledge Graph-based Retrofitting** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2311.13314)\n\n### Hybrid Retriever\n- (arXiv 2024) **Think-on-Graph 2.0: Deep and Faithful Large Language Model Reasoning with Knowledge-guided Retrieval Augmented Generation** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2407.10805)\n- (arXiv 2024) **StructRAG: Boosting Knowledge Intensive Reasoning of LLMs via Inference-time Hybrid Information Structurization** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2410.08815)\n\n## Knowledge Integration\n### Fine-tuning\n#### Fine-tuning with Node-level Knowledge\n- (arXiv 2025) **Large Language Models based Graph Convolution for Text-Attributed Networks?** [[Paper]](https:\u002F\u002Fopenreview.net\u002Fforum?id=x5FfUvsLIE)\n- (SIGIR 2024) **Graphgpt: Graph instruction tuning for large language models** [[Paper]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3626772.3657775)\n#### Fine-tuning with Path-level Knowledge\n- (AAAI 2024) **Exploring large language model for graph data understanding in online job recommendations** [[Paper]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1609\u002Faaai.v38i8.28769)\n- (arXiv 2024) **MuseGraph: Graph-oriented Instruction Tuning of Large Language Models for Generic Graph Mining** [[Paper]](https:\u002F\u002FarXiv.org\u002Fpdf\u002F2403.04780)\n- (WWW 2023) **Structure pretraining and prompt tuning for knowledge graph transfer** [[Paper]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3543507.3583301)\n- (ICLR 2023) **Reasoning on graphs: Faithful and interpretable large language model reasonin**g [[Paper]](https:\u002F\u002Fopenreview.net\u002Fforum?id=ZGNWW7xZ6Q)\n\n#### Fine-tuning with Subgraph-level Knowledge\n- (ICML 2024) **Llaga: Large language and graph assistant** [[Paper]](https:\u002F\u002Fopenreview.net\u002Fpdf?id=B48Pzc4oKi)\n- (KDD 2024) **Graphwiz: An instruction-following language model for graph problems** [[Paper]](https:\u002F\u002Fgraph-wiz.github.io\u002F)\n- (AAAI 2024) **Graph neural prompting with large language models** [[Paper]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1609\u002Faaai.v38i17.29875)\n- (ACL 2024 Findings) **Rho:Reducing hallucination in open-domain dialogues with knowledge\ngrounding** [[Paper]](https:\u002F\u002Faclanthology.org\u002F2023.findings-acl.275\u002F)\n- (EACL 2024 Findings) **Language is All a Graph Needs** [[Paper]](https:\u002F\u002Faclanthology.org\u002F2024.findings-eacl.132.pdf)\n\n### In-context Learning\n#### Graph-enhanced Chain-of-Thought\n- (KBS 2025) **Different paths to the same destination: Diversifying LLMs generation for multi-hop open-domain question answering** [[Paper]](https:\u002F\u002Fwww.sciencedirect.com\u002Fscience\u002Farticle\u002Fabs\u002Fpii\u002FS0950705124014230)\n- (ICLR 2024) **Reasoning on Graphs: Faithful and Interpretable Large Language Model Reasoning** [[Paper]](https:\u002F\u002Fopenreview.net\u002Fforum?id=ZGNWW7xZ6Q)\n- (ICLR 2024) **Think-on-Graph: Deep and Responsible Reasoning of Large Language Model on Knowledge Graph** [[Paper]](https:\u002F\u002Fopenreview.net\u002Fforum?id=nnVO1PvbTv)\n- (arXiv 2024) **Think-on-Graph 2.0: Deep and Faithful Large Language Model Reasoning with Knowledge-guided Retrieval Augmented Generation** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2407.10805)\n- (arXiv 2024) **Graph Chain-of-Thought: Augmenting Large Language Models by Reasoning on Graphs** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2404.07103)\n- (ICLR 2024) **Chain-of-Knowledge: Grounding Large Language Models via Dynamic Knowledge Adapting over Heterogeneous Sources** [[Paper]](https:\u002F\u002Fopenreview.net\u002Fforum?id=cPgh4gWZlz)\n- (ACL 2024 Findings) **Visual In-Context Learning for Large Vision-Language Models** [[Paper]](https:\u002F\u002Fwww.semanticscholar.org\u002FPaper\u002FVisual-In-Context-Learning-for-Large-Models-Zhou-Li\u002Fb00d1028291ae64e9d7485a34ec5f1b7b5a37909)\n- (NeurIPS 2023) **What makes good examples for visual in-context learning?** [[Paper]](https:\u002F\u002Fproceedings.neurips.cc\u002Fpaper_files\u002Fpaper\u002F2023\u002Fhash\u002F398ae57ed4fda79d0781c65c926d667b-Abstract-Conference.html)\n- (ACL 2023) **Plan-and-Solve Prompting: Improving Zero-Shot Chain-of-Thought Reasoning by Large Language Models** [[Paper]](https:\u002F\u002Faclanthology.org\u002F2023.acl-long.147.pdf)\n- (AAAI 2024) **When Do Program-of-Thought Works for Reasoning?** [[Paper]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1609\u002Faaai.v38i16.29721)\n- (ICLR 2022) **An Explanation of In-context Learning as Implicit Bayesian Inference** [[Paper]](https:\u002F\u002Fopenreview.net\u002Fforum?id=RdJVFCHjUMI)\n- (EMNLP 2023) **KnowledGPT: Enhancing Large Language Models with Retrieval and Storage Access on Knowledge Bases** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2308.11761)\n\n#### Collaborative Knowledge Graph Refinement\n- (AAAI 2024) **Mitigating large language model hallucinations via autonomous knowledge graph-based retrofitting** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2311.13314)\n- (ACL 2024 Findings) **Knowledge Graph-Enhanced Large Language Models via Path Selection** [[Paper]](https:\u002F\u002Faclanthology.org\u002F2024.findings-acl.376\u002F)\n- (NeurIPS 2024) **Plan-on-Graph: Self-Correcting Adaptive Planning of Large Language Model on Knowledge Graphs** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2410.23875)\n- (arXiv 2024) **Explore then Determine: A GNN-LLM Synergy Framework for Reasoning over Knowledge Graph** [[Paper]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2406.01145)\n- (ACL 2024) **CogMG: Collaborative Augmentation Between Large Language Model and Knowledge Graph** [[Paper]](https:\u002F\u002Faclanthology.org\u002F2024.acl-demos.35\u002F)\n\n\n\n# 📚 Related Survey Papers\n- (arXiv 2025) **Retrieval-Augmented Generation with Graphs (GraphRAG)** [[Paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2501.00309)\n- (arXiv 2024) **Graph Retrieval-Augmented Generation: A Survey** [[Paper]](https:\u002F\u002FarXiv.org\u002Fpdf\u002F2408.08921)\n- (AIxSET 2024) **Graph Retrieval-Augmented Generation for Large Language Models: A Survey** [[Paper]](https:\u002F\u002Fpapers.ssrn.com\u002Fsol3\u002FDelivery.cfm?abstractid=4895062)\n\nTo explore the applications of LLMs on graph tasks, we recommend the following repositories:\n- [Awesome-LLMs-in-Graph-tasks](https:\u002F\u002Fgithub.com\u002FyhLeeee\u002FAwesome-LLMs-in-Graph-tasks) by [Yuhan Li](https:\u002F\u002Fyhleeee.github.io\u002F) from HKUST(GZ).\n- [Awesome-Graph-LLM](https:\u002F\u002Fgithub.com\u002FXiaoxinHe\u002FAwesome-Graph-LLM) by [Xiaoxin He](https:\u002F\u002Fxiaoxinhe.github.io\u002F) from NUS.\n- [Awesome-Graph-Prompt](https:\u002F\u002Fgithub.com\u002FWxxShirley\u002FAwesome-Graph-Prompt), created by [Xixi Wu](https:\u002F\u002Fwxxshirley.github.io\u002F) from CUHK.\n\n\n# 🏆 Benchmarks\n| Dataset | Task | Paper | Repo |\n| --- | --- | --- | --- |\n| GraphRAG-Bench | GraphRAG evaluation | [[arXiv 2025]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2506.05690) | [[Github]](https:\u002F\u002Fgithub.com\u002FGraphRAG-Bench\u002FGraphRAG-Benchmark) |\n| DIGIMON | Large-scale graphRAG | [[arXiv 2025]](https:\u002F\u002Fwww.arxiv.org\u002Fpdf\u002F2503.04338) | [[Github]](https:\u002F\u002Fgithub.com\u002FJayLZhou\u002FGraphRAG) |\n| PolyG | GraphRAG evaluation | [[arXiv 2025]](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2504.02112) | [[Github]](https:\u002F\u002Fgithub.com\u002FLiu-rj\u002FPolyG) |\n| SimpleQuestion | Simple Question Answering | [[arXiv 2015]](https:\u002F\u002FarXiv.org\u002Fabs\u002F1506.02075) | [[Github]](https:\u002F\u002Fgithub.com\u002FJerryzhao-z\u002Fsimple-question-answering-with-memory-networks) |\n| WebQ | Simple Question Answering | [[EMNLP 2013]](https:\u002F\u002Fnlp.stanford.edu\u002Fpubs\u002FsemparseEMNLP13.pdf) | [[CodaLab]](https:\u002F\u002Fworksheets.codalab.org\u002Fworksheets\u002F0xba659fe363cb46e7a505c5b6a774dc8a) |\n|Multihop-RAG | Multi-hop Reasoning | [[COLING 2024]](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2401.15391) |    [[Github]](https:\u002F\u002Fgithub.com\u002Fyixuantt\u002FMultiHop-RAG\u002F) |\n| CWQ | Multi-hop Reasoning | [[NAACL 2018]](https:\u002F\u002Faclanthology.org\u002FN18-1059\u002F) | [[TAU-NLP]](https:\u002F\u002Fwww.tau-nlp.org\u002Fcompwebq) |\n| MetaQA | Multi-hop Reasoning | [[AAAI 2018]](https:\u002F\u002FarXiv.org\u002Fabs\u002F1709.04071) | [[Github]](https:\u002F\u002Fgithub.com\u002Fyuyuz\u002FMetaQA) |\n| MetaQA-3 | Multi-hop Reasoning | [[AAAI 2018]](https:\u002F\u002FarXiv.org\u002Fabs\u002F1709.04071) | [[Github]](https:\u002F\u002Fgithub.com\u002Fyuyuz\u002FMetaQA) |\n| CURD |  Large-scale Complex QA | [[arXiv 2024]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2401.17043) | [[Github]](https:\u002F\u002Fgithub.com\u002FIAAR-Shanghai\u002FCRUD_RAG) |\n| KQAPro | Large-scale Complex QA | [[ACL 2022]](https:\u002F\u002Faclanthology.org\u002F2022.acl-long.422\u002F) | [[Github]](https:\u002F\u002Fgithub.com\u002Fshijx12\u002FKQAPro_Baselines) |\n| LC-QuAD v2 | Large-scale Complex QA | [[ISWC 2019]](https:\u002F\u002Flink.springer.com\u002Fchapter\u002F10.1007\u002F978-3-030-30796-7_5) | [[figshare]](https:\u002F\u002Ffigshare.com\u002Fprojects\u002FLCQuAD_2_0\u002F62270) |\n| LC-QuAD | Large-scale Complex QA | [[ISWC 2017]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1007\u002F978-3-319-68204-4_22) | [[Github]](https:\u002F\u002Fgithub.com\u002FAskNowQA\u002FLC-QuAD) |\n| UltraDomain | Domain-specific QA | [[arXiv 2024]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2409.05591) | [[Github]](https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002FMemoRAG#dataset) |\n| TutorQA | Domain-specific QA | [[arXiv 2024]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2407.10794) | [[Github]](https:\u002F\u002Fgithub.com\u002FIreneZihuiLi\u002FCGPrompt) |\n| FACTKG  | Domain-specific QA | [[ACL 2023]](https:\u002F\u002Faclanthology.org\u002F2023.acl-long.895.pdf) | [[Github]](https:\u002F\u002Fgithub.com\u002Fjiho283\u002FFactKG) |\n| Mintaka | Domain-specific QA | [[ACL 2022]](https:\u002F\u002Faclanthology.org\u002F2022.coling-1.138\u002F) | [[Github]](https:\u002F\u002Fgithub.com\u002Famazon-science\u002Fmintaka) |\n| GrailQA | Domain-specific QA | [[WWW 2021]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3442381.3449992) | [[Github]](https:\u002F\u002Fgithub.com\u002Fdki-lab\u002FGrailQA) |\n| WebQSP | Domain-specific QA | [[ACL 2016]](https:\u002F\u002Faclanthology.org\u002FP16-2033.pdf) | [[Microsoft]](http:\u002F\u002Faka.ms\u002FWebQSP) |\n\n# 💻 Open-source Project\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002FHawksight-AI\u002Fsemantica) Semantica: an open-source, production-ready semantic layer and GraphRAG framework that sits between raw corpora and LLMs.\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Fbibinprathap\u002FVeritasGraph) Graph RAG pipeline that runs locally with ollama and has full source attribution \n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgraphrag-bench.github.io\u002F) GraphRAG-Bench: A Comprehensive Benchmark and Analysis for Graph Retrieval-Augmented Generation. \n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002FchensyCN\u002FAgentic-RAG) Agentic-RAG: A clean and extensible agentic RAG system. \n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Fapecloud\u002FApeRAG) ApeRAG: Production-ready GraphRAG with multi-modal indexing, AI agents, MCP support, and scalable K8s deployment\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Fgetzep\u002Fgraphiti) Graphiti: Build Real-Time Knowledge Graphs for AI Agents.\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002FJayLZhou\u002FGraphRAG) DIGIMON: A unified and modular graph-based RAG framework\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fgraphrag.git) Microsoft-GraphRAG: A modular graph-based Retrieval-Augmented Generation (RAG) system\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Fgusye1234\u002Fnano-graphrag) Nano-GraphRAG: A simple, easy-to-hack GraphRAG implementation\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Fcirclemind-ai\u002Ffast-graphrag) Fast GraphRAG: RAG that intelligently adapts to your use case, data, and queries\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002FHKUDS\u002FLightRAG) LightRAG: Simple and Fast Retrieval-Augmented Generation\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Ftpoisonooo\u002FHuixiangDou2) HuixiangDou2: A Robustly Optimized GraphRAG Approach\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002FFalkorDB\u002FGraphRAG-SDK) GraphRAG-SDK: a specialized toolkit for building GraphRAG systems.\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Fvitali87\u002Fcode-graph-rag) Code-Graph-RAG: A graph-based RAG system that analyzes multi-language codebases using Tree-sitter, builds knowledge graphs, and enables natural language querying and editing via MCP server.\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Fonestardao\u002FWFGY) WFGY Problem Map: a specialized toolkit that defines 16 recurring failure modes that show up in RAG and LLM pipelines.\n\n\n# 🍀 Citation\nIf you find this survey helpful, please cite our paper:\n```\n@article{zhang2025survey,\n  title={A Survey of Graph Retrieval-Augmented Generation for Customized Large Language Models},\n  author={Zhang, Qinggang and Chen, Shengyuan and Bei, Yuanchen and Yuan, Zheng and Zhou, Huachi and Hong, Zijin and Dong, Junnan and Chen, Hao and Chang, Yi and Huang, Xiao},\n  journal={arXiv preprint arXiv:2501.13958},\n  year={2025}\n}\n```\n","# 令人惊叹的GraphRAG（GraphRAG综述）\n\n\u003Cdiv align=\"center\">\n    \u003Ca href=\"https:\u002F\u002Fawesome.re\">\u003Cimg src=\"https:\u002F\u002Fawesome.re\u002Fbadge.svg\"\u002F>\u003C\u002Fa>\n    \u003Ca href=\"http:\u002F\u002Fmakeapullrequest.com\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FPRs-welcome-green.svg\"\u002F>\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Farxiv.org\u002Fabs\u002F2501.13958\" target=\"_blank\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FPaper-Arxiv-red?logo=arxiv&style=flat-square\" alt=\"arXiv:2506.08938\">\u003C\u002Fa>\n    \u003Ca href=\"http:\u002F\u002Fmakeapullrequest.com\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flast-commit\u002FDEEP-PolyU\u002FAwesome-GraphRAG?color=blue\"\u002F>\u003C\u002Fa>\n    \u003Ca href=\"http:\u002F\u002Fmakeapullrequest.com\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FDEEP-PolyU\u002FAwesome-GraphRAG\"\u002F>\u003C\u002Fa>\n\u003C\u002Fdiv>\n\n\n本仓库收录了基于图的检索增强生成（GraphRAG）相关资源的精选列表，这些资源按照“[**面向定制化大语言模型的图式检索增强生成综述**](https:\u002F\u002Farxiv.org\u002Fabs\u002F2501.13958)”进行分类。持续更新中，敬请关注！\n\n\n**📃 如果您觉得我们的综述或仓库有所帮助，请[引用我们的论文](#-citation)**！\n\n\n# 🎉 最新消息\n- **[2026-01-26]** 我们的**[GraphRAG基准测试](https:\u002F\u002Fgithub.com\u002FGraphRAG-Bench\u002FGraphRAG-Benchmark)**已被ICLR’26接收。\n- **[2026-01-26]** 我们的**[LinearRAG](https:\u002F\u002Fgithub.com\u002FDEEP-PolyU\u002FLinearRAG)**已被ICLR’26接收。\n- **[2025-11-08]** 我们的**[LogicRAG](https:\u002F\u002Fgithub.com\u002FchensyCN\u002FLogicRAG.git)**已被AAAI'26接收。\n- **[2025-10-27]** 我们发布了**[LinearRAG](https:\u002F\u002Fgithub.com\u002FDEEP-PolyU\u002FLinearRAG)**，这是一种用于高效GraphRAG的关系无关图构建方法。\n- **[2025-06-06]** 我们发布了**[GraphRAG基准测试](https:\u002F\u002Fgithub.com\u002FGraphRAG-Bench\u002FGraphRAG-Benchmark.git)**，用于评估GraphRAG模型。\n- **[2025-05-14]** 我们发布了[GraphRAG基准数据集](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002FGraphRAG-Bench\u002FGraphRAG-Bench)。\n- **[2025-01-21]** 我们发布了[GraphRAG综述](https:\u002F\u002Fgithub.com\u002FDEEP-PolyU\u002FAwesome-GraphRAG)。\n\n---\n\n\u003Cdiv>\n\u003Ch3 align=\"left\">\n       \u003Cp align=\"center\">\u003Cimg width=\"100%\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FDEEP-PolyU_Awesome-GraphRAG_readme_22be195ae39e.png\" \u002F>\u003C\u002Fp>\n    \u003Cp align=\"center\">\u003Cem>传统RAG与两种典型GraphRAG工作流的概述。\u003C\u002Fem>\u003C\u002Fp>\n\u003C\u002Fdiv>\n\n- **非图式RAG**将语料库组织成块，按相似度排序，并检索最相关的文本以生成响应。\n- **基于知识的GraphRAG**利用实体识别和关系抽取从语料库中提取详细的知识图谱，提供细粒度、领域特定的信息。\n- **基于索引的GraphRAG**将语料库概括为高层次的主题节点，这些节点通过链接形成索引图，而事实链接则将主题与文本关联起来。\n\n---\n\n# RAG与GraphRAG\nGraphRAG是RAG的一种新范式，它通过三大创新彻底革新了领域特定的LLM应用，从而克服了传统RAG的局限性：**(i) 图结构化的知识表示**，能够显式地捕捉实体关系和领域层次结构；**(ii) 图感知的检索机制**，支持多跳推理和上下文保留的知识获取；以及**(iii) 结构引导的知识搜索算法**，确保在大规模语料库中高效检索。\n    \n\n\u003Ch3 align=\"center\">\n   \u003Cp align=\"center\">\u003Cimg width=\"100%\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FDEEP-PolyU_Awesome-GraphRAG_readme_b0d1526bdca9.png\" \u002F>\u003C\u002Fp>\n    \u003Cp align=\"center\">\u003Cem>传统RAG与GraphRAG的比较。\u003C\u002Fem>\u003C\u002Fp>\n\n\n\n# 📫 联系我们\n我们欢迎研究人员分享相关工作，以丰富本列表，或对我们的综述提出有见地的意见。欢迎随时联系通讯作者之一：[Qinggang Zhang](https:\u002F\u002Fqing145.github.io\u002F)，[Shengyuan Chen](https:\u002F\u002Fchensycn.github.io\u002F)。\n\n\n## 目录\n- [🍀 引用](#-citation)\n- [📫 联系我们](#-contact-us)\n- [📈 GraphRAG研究趋势](#-trend-of-graphrag-research)\n- [📜 研究论文](#-research-papers)\n    - [知识组织](#knowledge-organization)\n        - [用于知识索引的图](#graphs-for-knowledge-indexing)\n        - [作为知识载体的图](#graphs-as-knowledge-carrier)\n            - [从语料库构建知识图谱](#knowledge-graph-construction-from-corpus)\n            - [使用现有KG的GraphRAG](#graphrag-with-existing-kgs)\n        - [混合式GraphRAG](#hybrid-graphrag)\n    - [知识检索](#knowledge-retrieval)\n        - [基于语义相似度的检索器](#semantics-similarity-based-retriever)\n        - [基于逻辑推理的检索器](#logical-reasoning-based-retriever)\n        - [基于LLM的检索器](#llm-based-retriever)\n        - [基于GNN的检索器](#gnn-based-retriever)\n        - [多轮检索器](#multi-round-retriever)\n        - [后检索](#post-retrieval)\n        - [混合式检索器](#hybrid-retriever)\n    - [知识整合](#knowledge-integration)\n        - [微调](#fine-tuning)\n            - [基于节点级知识的微调](#fine-tuning-with-node-level-knowledge)\n            - [基于路径级知识的微调](#fine-tuning-with-path-level-knowledge)\n            - [基于子图级知识的微调](#fine-tuning-with-subgraph-level-knowledge)\n        - [上下文学习](#in-context-learning)\n            - [图增强的思维链](#graph-enhanced-chain-of-thought)\n            - [协作式知识图谱精炼](#collaborative-knowledge-graph-refinement)\n- [📚 相关综述论文](#-related-survey-papers)\n- [🏆 基准测试](#-benchmarks)\n- [💻 开源项目](#-open-source-projects)\n\n\n# 📈 GraphRAG研究趋势\n\n\u003Ch3 align=\"center\">\n   \u003Cp align=\"center\">\u003Cimg width=\"100%\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FDEEP-PolyU_Awesome-GraphRAG_readme_4610c437cc87.png\" \u002F>\u003C\u002Fp>\n    \u003Cp align=\"center\">\u003Cem>GraphRAG领域的发展趋势及代表性工作。\u003C\u002Fem>\u003C\u002Fp>\n\n# 📜 研究论文\n## 知识组织\n\n### 知识图谱索引相关论文\n- (arXiv 2025) **LinearRAG：大规模语料上的线性图检索增强生成** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2510.10114)\n- (EMNLP 2025) **别忘了基础检索器！一种用于多跳问答的低资源图结构检索器** [[论文]](https:\u002F\u002Faclanthology.org\u002F2025.emnlp-industry.174\u002F)\n- (arXiv 2025) **以查询为中心的图检索增强生成** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2509.22009)\n- (arXiv 2025) **多智能体GraphRAG：面向标注属性图的文本到Cypher框架** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2510.09156)\n- (arXiv 2025) **基于经验的 grounded方法：结合层次化代理式检索的生成式医疗健康预测** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2511.13293)\n- (ICML 2025) **HippoRAG2：从RAG到记忆——大型语言模型的非参数持续学习** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2502.14802)\n- (arXiv 2025) **PersonaAgent与GraphRAG：面向个性化的社区感知知识图谱** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2511.17467)\n- (arXiv 2025) **E^2GraphRAG：简化图结构RAG以实现高效与高效果** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2505.24226)\n- (arXiv 2025) **DIGIMON：统一且模块化的图结构RAG框架** [[论文]](https:\u002F\u002Fgithub.com\u002FJayLZhou\u002FGraphRAG.git)\n- (arXiv 2025) **ArchRAG：基于属性的社区层次化检索增强生成** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2502.09891)\n- (arXiv 2025) **KET-RAG：面向GraphRAG的成本效益型多粒度索引框架** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2502.09304)\n- (arXiv 2025) **PIKE-RAG：专门的知识与推理增强生成** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2501.11551)\n- (EMNLP 2025 Findings) **基于层次化知识的检索增强生成** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2503.10150)\n- (arXiv 2024) **用于LLM问答任务的知识图神经网络增强检索** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2406.06572)\n- (arXiv 2024) **KAG：通过知识增强生成提升专业领域中的LLM表现** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2409.13731)\n- (arXiv 2024) **OG-RAG：面向大型语言模型的本体论驱动检索增强生成** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2412.15235)\n- (arXiv 2024) **GRAG：图检索增强生成** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2405.16506)\n- (arXiv 2024) **通过自学习赋能大型语言模型构建知识检索索引** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2405.16933)\n- (ICLR 2024) **RAPTOR：面向树状组织检索的递归抽象处理** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2401.18059)\n- (AAAI 2024) **面向多文档问答的知识图提示** [[论文]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1609\u002Faaai.v38i17.29889)\n- (arXiv 2024) **GraphCoder：基于代码上下文图检索与语言模型的仓库级代码补全增强** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2406.07003)\n- (NeurIPS 2023) **Avis：利用大型语言模型代理实现自主视觉信息搜索** [[论文]](https:\u002F\u002Fopenreview.net\u002Fforum?id=7EMphtUgCI&noteId=yGw4rbGozi)\n- (CoRL 2023) **Sayplan：使用3D场景图将大型语言模型接地，用于可扩展的机器人任务规划** [[论文]](https:\u002F\u002Fproceedings.mlr.press\u002Fv229\u002Frana23a\u002Frana23a.pdf)\n- (arXiv 2020) **通过多跳密集检索回答复杂开放域问题** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2009.12756)\n- (arXiv 2019) **面向开放域问答的知识引导文本检索与阅读** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F1911.03868)\n\n### 图作为知识载体\n#### 从语料库构建知识图谱\n- (AAAI 2026) **RAG 不需要预构建图：基于自适应推理结构的检索增强生成** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2508.06105)\n- (arXiv 2025) **AutoGraph-R1：面向知识图谱构建的端到端强化学习** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2510.15339)\n- (arXiv 2025) **AGRAG：面向大语言模型的高级图结构检索增强生成** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2511.05549)\n- (EMNLP 2025) **MaGiX：用于提升跨语言 RAG 的多粒度自适应图智能框架** [[论文]](https:\u002F\u002Faclanthology.org\u002Fanthology-files\u002Fanthology-files\u002Fpdf\u002Ffindings\u002F2025.findings-emnlp.279.pdf)\n- (CIKM 2025) **面向查询聚焦摘要的上下文感知细粒度图 RAG** [[论文]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3746252.3760935)\n- (CIKM 2025) **DocPolicyKG：基于轻量级大语言模型的中文政策文件知识图谱构建框架** [[论文]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fabs\u002F10.1145\u002F3746252.3760904)\n- (arXiv 2025) **SUBQRAG：子问题驱动的动态图 RAG** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2510.07718)\n- (arXiv 2025) **本体学习与知识图谱构建：方法比较及其对 RAG 性能的影响** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2511.05991)\n- (NeurIPS 2025) **GFM-RAG：用于检索增强生成的图基础模型** [论文](https:\u002F\u002Farxiv.org\u002Fabs\u002F2502.01113)\n- (arXiv 2025) **G-reasoner：面向图结构知识统一推理的基础模型** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2509.24276)\n- (CVPR 2025) **医学图 RAG：通过图结构检索增强生成迈向安全的医疗大语言模型** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2408.04187)\n- (arXiv 2025) **Youtu-GraphRAG：面向图结构检索增强复杂推理的垂直统一代理** [[论文]](https:\u002F\u002Fwww.arxiv.org\u002Fabs\u002F2508.19855)\n- (arXiv 2025) **基于层次化知识的检索增强生成** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2503.10150)\n- (arXiv 2025) **MedRAG：利用知识图谱引导的推理提升医疗助手的检索增强生成能力** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2502.04413)\n- (arXiv 2025) **PathRAG：基于关系路径剪枝的图结构检索增强生成** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2502.14902)\n- (EDBT 2025) **DBCopilot：通过模式路由实现对海量数据库的自然语言查询** [[论文]](https:\u002F\u002Fopenproceedings.org\u002F2025\u002Fconf\u002Fedbt\u002Fpaper-209.pdf)\n- (arXiv 2024) **从局部到全局：一种面向查询聚焦摘要的图 RAG 方法** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2404.16130)\n- (EMNLP 2024) **结构引导提示：通过探索文本的图结构指导大语言模型进行多步推理** [[论文]](https:\u002F\u002Faclanthology.org\u002F2024.emnlp-main.528.pdf)\n- (EMNLP 2024 Findings) **GraphReader：构建图结构代理以增强大语言模型的长上下文理解能力** [[论文]](https:\u002F\u002Faclanthology.org\u002F2024.findings-emnlp.746\u002F)\n- (SIGIR 2024) **面向客户服务问答的知识图谱检索增强生成** [[论文]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002Fabs\u002F10.1145\u002F3626772.3661370)\n- (arXiv 2024) **DynaGRAG：探索信息拓扑以推进图结构检索增强生成中的语言理解和生成** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2412.18644)\n- (arXiv 2024) **FastRAG：面向半结构化数据的检索增强生成** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2411.13773)\n- (TechRxiv 2024) **LuminiRAG：面向复杂多模态文档理解的视觉增强图 RAG** [[论文]](https:\u002F\u002Fwww.techrxiv.org\u002Fusers\u002F867713\u002Farticles\u002F1248304-luminirag-vision-enhanced-graph-rag-for-complex-multi-modal-document-understanding)\n- (BigData 2023) **AutoKG：面向语言模型的高效自动化知识图谱生成** [[论文]](https:\u002F\u002Fieeexplore.ieee.org\u002Fabstract\u002Fdocument\u002F10386454)\n- (ACL 2019) **利用局部知识图谱构建扩展序列到序列模型以处理多文档输入** [[论文]](https:\u002F\u002Faclanthology.org\u002FD19-1428.pdf)\n- (SIGIR 2019) **通过将多文档证据与准知识图谱结合来回答复杂问题** [[论文]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3331184.3331252)\n\n#### 基于现有知识图谱的GraphRAG\n- (arXiv 2025) **GraphSearch：面向图检索增强生成的代理式深度搜索工作流** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2509.22009)\n- (arXiv 2025) **通过注意力模式与语义对齐检测图检索增强生成中的幻觉现象** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2512.09148)\n- （arXiv 2025） **推理规模化的GraphRAG：提升知识图谱上的多跳问答能力** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2506.19967)\n- (AAAI 2025) **LightPROF：面向知识图谱的大语言模型轻量级推理框架** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2504.03137)\n- (ICLR 2025) **简单即有效：图与大语言模型在基于知识图谱的检索增强生成中的作用** [[论文]](https:\u002F\u002Fopenreview.net\u002Fforum?id=JvkuZZ04O7)\n- (arXiv 2025) **通过知识过滤与集成赋能GraphRAG** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2503.13804)\n- (arXiv 2024)**StructRAG：通过推理时混合信息结构化提升大语言模型的知识密集型推理能力** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2410.08815)\n- (ICLR 2024) **图上推理：忠实且可解释的大语言模型推理** [[论文]](https:\u002F\u002Fopenreview.net\u002Fforum?id=ZGNWW7xZ6Q)\n- (AAAI 2024) **通过自主知识图谱重校准缓解大语言模型幻觉问题** [[论文]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1609\u002Faaai.v38i16.29770)\n- (ICLR 2024) **Think-on-Graph：大语言模型在知识图谱上的深度与负责任推理** [[论文]](https:\u002F\u002Fopenreview.net\u002Fforum?id=nnVO1PvbTv)\n- (Bioinformatics 2024) **基于生物医学知识图谱的大语言模型提示词生成** [[论文]](https:\u002F\u002Facademic.oup.com\u002Fbioinformatics\u002Farticle\u002F40\u002F9\u002Fbtae560\u002F7759620)\n- (NeurIPS 2024) **KnowGPT：面向大语言模型的知识图谱提示技术** [[论文]](https:\u002F\u002Fopenreview.net\u002Fforum?id=PacBluO5m7&referrer=%5Bthe%20profile%20of%20Daochen%20Zha%5D(%2Fprofile%3Fid%3D~Daochen_Zha1))\n- (ACL 2024 Findings) **通过路径选择增强知识图谱的大语言模型** [[论文]](https:\u002F\u002Faclanthology.org\u002F2024.findings-acl.376\u002F)\n- (IEEE VIS 2024) **KNOWNET：通过知识图谱集成引导大语言模型进行健康信息检索** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2407.13598)\n- (CoLM 2024) **ProLLM：基于蛋白质思维链的大语言模型，用于预测蛋白质-蛋白质相互作用** [[论文]](https:\u002F\u002Fopenreview.net\u002Fforum?id=2nTzomzjjb#discussion)\n- (arXiv 2024) **LEGO-GraphRAG：模块化图基检索增强生成，用于设计空间探索** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2411.05844)\n- (arXiv 2024) **Think-on-Graph 2.0：基于知识引导的检索增强生成实现深度与忠实的大语言模型推理** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2407.10805)\n\n\n\n### 混合GraphRAG\n- (NAACL 2025) **知识图谱引导的检索增强生成** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2502.06864)\n- (ACL 2024 Findings) **HybGRAG：文本与关系型知识库上的混合检索增强生成**[[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2412.16311)\n- (arXiv 2024) **记录图：利用图结构提升长上下文摘要的检索增强生成能力** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2410.11001)\n- (arXiv 2024) **医疗图RAG：通过图检索增强生成迈向安全的医疗大语言模型** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2408.04187)\n- (arXiv 2024) **Codexgraph：通过代码图数据库连接大语言模型与代码仓库** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2408.03910)\n\n## 知识检索\n\n### 基于语义相似度的检索器\n- (AAAI 2024) **StructuGraphRAG：面向检索增强生成的结构化文档驱动知识图谱** [[论文]](https:\u002F\u002Fojs.aaai.org\u002Findex.php\u002FAAAI-SS\u002Farticle\u002Fview\u002F31798\u002F33965)\n- (arXiv 2024) **G-Retriever：面向文本图理解和问答的检索增强生成** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2402.07630)\n- (arXiv 2024) **CancerKG.ORG：一个面向网络规模、交互式且可验证的知识图谱-大语言模型混合系统，用于辅助最佳癌症治疗与护理** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2501.00223)\n- (arXiv 2024) **通过自学习赋能大语言模型构建知识检索索引** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2405.16933)\n- (arXiv 2024) **GraphCoder：通过代码上下文图检索与语言模型提升仓库级别的代码补全能力** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2406.07003)\n- (arXiv 2024) **医疗图RAG：通过图检索增强生成迈向安全的医疗大语言模型** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2408.04187)\n- (arXiv 2024) **如何让大语言模型成为强大的节点分类器？** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2410.02296)\n\n### 基于逻辑推理的检索器\n- (AAAI 2026) **RAG并不需要预构建的图：自适应推理结构支持的检索增强生成** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2508.06105)\n- (NeurIPS 2024) **KnowGPT：面向大语言模型的知识图谱提示技术** [[论文]](https:\u002F\u002Fopenreview.net\u002Fforum?id=PacBluO5m7&referrer=%5Bthe%20profile%20of%20Daochen%20Zha%5D(%2Fprofile%3Fid%3D~Daochen_Zha1))\n- (ACL 2024 Findings) **通过路径选择增强知识图谱的大语言模型** [[论文]](https:\u002F\u002Faclanthology.org\u002F2024.findings-acl.376\u002F)\n- (ICLR 2024) **Think-on-Graph：大语言模型在知识图谱上的深度与负责任推理** [[论文]](https:\u002F\u002Fopenreview.net\u002Fforum?id=nnVO1PvbTv)\n- (CIKM 2024) **RD-P：面向大语言模型的可信知识图谱检索增强提示器** [[论文]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3627673.3679659)\n- (arXiv 2024) **RuleRAG：面向问答任务的语言模型规则引导检索增强生成** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2410.22353)\n- (LHB 2024) **由知识图谱与大语言模型协作驱动的水利工程项目检查智能问答** [[论文]](https:\u002F\u002Fwww.tandfonline.com\u002Fdoi\u002Ffull\u002F10.1080\u002F27678490.2024.2397337)\n- (arXiv 2024) **RiTeK：面向大语言模型在文本知识图谱上复杂推理的数据集** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2410.13987)\n\n### 基于大语言模型的检索器\n- (AAAI 2024) **基于知识图谱提示的多文档问答** [[论文]](https:\u002F\u002Fwww.overleaf.com\u002Fproject\u002F667419080bc7191bc75f5880)\n- (EMNLP 2024) **结构引导式提示：通过探索文本的图结构指导大语言模型进行多步推理** [[论文]](https:\u002F\u002Faclanthology.org\u002F2024.emnlp-main.528.pdf)\n- (ACML 2024) **利用知识图谱增强的大语言模型提升教科书问答性能** [[论文]](https:\u002F\u002Fopenreview.net\u002Fforum?id=ATiIqCCqR2)\n- (ICLR 2024) **Think-on-Graph：大语言模型在知识图谱上的深度与负责任推理** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2307.07697)\n- (arXiv 2024) **LightRAG：简单快速的检索增强生成** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2410.05779)\n- (arXiv 2024) **MEG：用于问答任务的医学知识增强型大语言模型** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2411.03883)\n- (arXiv 2024) **从局部到全局：面向查询摘要的图RAG方法** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2404.16130)\n\n### 基于图神经网络的检索器\n- (arXiv 2025) **CG-RAG：基于引文图检索增强的大语言模型进行科研问答** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2501.15067)\n- (arXiv 2024) **具有图结构的高级RAG模型：优化复杂知识推理与文本生成** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2411.03572)\n- (arXiv 2024) **语言模型是图学习者** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2410.02296)\n- (arXiv 2024) **图神经网络增强的LLM问答检索** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2406.06572)\n- (arXiv 2024) **知识图谱增强的语言模型用于知识 grounded 对话生成** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2305.18846)\n\n\n### 多轮检索器\n- (arXiv 2024) **图思维链：通过在图上推理增强大语言模型** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2404.07103)\n- (arXiv 2024) **面向知识图谱 grounded 对话生成的生成式子图检索** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2410.09350)\n- (arXiv 2024) **记录图：利用图技术提升长上下文摘要的检索增强生成能力** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2410.11001)\n### 检索后处理\n- (ACL 2024) **通过知识链提示提升语言模型推理能力** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2306.06427)\n- (ACL 2024 Findings) **需要时再叫我：LLM 可以高效且忠实地在结构化环境中进行推理** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2403.08593)\n- (arXiv 2024) **图约束推理：使用大语言模型在知识图谱上进行忠实推理** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2410.13080)\n- (arXiv 2024) **通过自主知识图谱驱动的改造缓解大语言模型幻觉问题** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2311.13314)\n\n### 混合检索器\n- (arXiv 2024) **Think-on-Graph 2.0：结合知识引导的检索增强生成实现深度且忠实的大语言模型推理** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2407.10805)\n- (arXiv 2024) **StructRAG：通过推理时混合信息结构化提升LLM的知识密集型推理能力** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2410.08815)\n\n## 知识整合\n### 微调\n#### 基于节点级知识的微调\n- (arXiv 2025) **基于图卷积的大语言模型应用于文本属性网络？** [[论文]](https:\u002F\u002Fopenreview.net\u002Fforum?id=x5FfUvsLIE)\n- (SIGIR 2024) **Graphgpt：面向大语言模型的图指令微调** [[论文]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3626772.3657775)\n#### 基于路径级知识的微调\n- (AAAI 2024) **探索大语言模型在在线职位推荐中对图数据的理解** [[论文]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1609\u002Faaai.v38i8.28769)\n- (arXiv 2024) **MuseGraph：面向通用图挖掘的大语言模型图导向指令微调** [[论文]](https:\u002F\u002FarXiv.org\u002Fpdf\u002F2403.04780)\n- (WWW 2023) **知识图谱迁移的结构预训练与提示微调** [[论文]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3543507.3583301)\n- (ICLR 2023) **在图上进行推理：忠实且可解释的大语言模型推理** [[论文]](https:\u002F\u002Fopenreview.net\u002Fforum?id=ZGNWW7xZ6Q)\n\n#### 基于子图级知识的微调\n- (ICML 2024) **Llaga：大型语言与图助手** [[论文]](https:\u002F\u002Fopenreview.net\u002Fpdf?id=B48Pzc4oKi)\n- (KDD 2024) **Graphwiz：面向图问题的指令遵循语言模型** [[论文]](https:\u002F\u002Fgraph-wiz.github.io\u002F)\n- (AAAI 2024) **大语言模型的图神经提示** [[论文]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1609\u002Faaai.v38i17.29875)\n- (ACL 2024 Findings) **Rho：通过知识 grounding 降低开放域对话中的幻觉现象** [[论文]](https:\u002F\u002Faclanthology.org\u002F2023.findings-acl.275\u002F)\n- (EACL 2024 Findings) **语言就是图所需的一切** [[论文]](https:\u002F\u002Faclanthology.org\u002F2024.findings-eacl.132.pdf)\n\n### 上下文学习\n#### 图增强思维链\n- (KBS 2025) **殊途同归：面向多跳开放域问答的多样化大模型生成** [[论文]](https:\u002F\u002Fwww.sciencedirect.com\u002Fscience\u002Farticle\u002Fabs\u002Fpii\u002FS0950705124014230)\n- (ICLR 2024) **图上的推理：忠实且可解释的大语言模型推理** [[论文]](https:\u002F\u002Fopenreview.net\u002Fforum?id=ZGNWW7xZ6Q)\n- (ICLR 2024) **图上思考：知识图谱上的深度与负责任的大语言模型推理** [[论文]](https:\u002F\u002Fopenreview.net\u002Fforum?id=nnVO1PvbTv)\n- (arXiv 2024) **图上思考 2.0：基于知识引导的检索增强生成的深度与忠实的大语言模型推理** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2407.10805)\n- (arXiv 2024) **图思维链：通过图上推理增强大语言模型** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2404.07103)\n- (ICLR 2024) **知识链：通过跨异构源的动态知识适配来 grounded 大语言模型** [[论文]](https:\u002F\u002Fopenreview.net\u002Fforum?id=cPgh4gWZlz)\n- (ACL 2024 Findings) **大型视觉-语言模型的视觉上下文学习** [[论文]](https:\u002F\u002Fwww.semanticscholar.org\u002FPaper\u002FVisual-In-Context-Learning-for-Large-Models-Zhou-Li\u002Fb00d1028291ae64e9d7485a34ec5f1b7b5a37909)\n- (NeurIPS 2023) **什么才是视觉上下文学习中好的示例？** [[论文]](https:\u002F\u002Fproceedings.neurips.cc\u002Fpaper_files\u002Fpaper\u002F2023\u002Fhash\u002F398ae57ed4fda79d0781c65c926d667b-Abstract-Conference.html)\n- (ACL 2023) **计划-求解提示：提升大语言模型零样本思维链推理能力** [[论文]](https:\u002F\u002Faclanthology.org\u002F2023.acl-long.147.pdf)\n- (AAAI 2024) **何时思维程序法适用于推理？** [[论文]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1609\u002Faaai.v38i16.29721)\n- (ICLR 2022) **将上下文学习解释为隐式贝叶斯推理** [[论文]](https:\u002F\u002Fopenreview.net\u002Fforum?id=RdJVFCHjUMI)\n- (EMNLP 2023) **KnowledGPT：通过知识库的检索与存储访问增强大语言模型** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2308.11761)\n\n#### 协作式知识图谱精炼\n- (AAAI 2024) **通过自主的知识图谱后处理缓解大语言模型幻觉问题** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2311.13314)\n- (ACL 2024 Findings) **通过路径选择增强知识图谱的大语言模型** [[论文]](https:\u002F\u002Faclanthology.org\u002F2024.findings-acl.376\u002F)\n- (NeurIPS 2024) **图上计划：大语言模型在知识图谱上的自我修正自适应规划** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2410.23875)\n- (arXiv 2024) **先探索再决策：基于 GNN-LLM 协同的知识图谱推理框架** [[论文]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2406.01145)\n- (ACL 2024) **CogMG：大语言模型与知识图谱之间的协作增强** [[论文]](https:\u002F\u002Faclanthology.org\u002F2024.acl-demos.35\u002F)\n\n\n\n# 📚 相关综述论文\n- (arXiv 2025) **基于图的检索增强生成（GraphRAG）** [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2501.00309)\n- (arXiv 2024) **图检索增强生成：综述** [[论文]](https:\u002F\u002FarXiv.org\u002Fpdf\u002F2408.08921)\n- (AIxSET 2024) **面向大语言模型的图检索增强生成：综述** [[论文]](https:\u002F\u002Fpapers.ssrn.com\u002Fsol3\u002FDelivery.cfm?abstractid=4895062)\n\n为了探索大语言模型在图任务中的应用，我们推荐以下仓库：\n- [Awesome-LLMs-in-Graph-tasks](https:\u002F\u002Fgithub.com\u002FyhLeeee\u002FAwesome-LLMs-in-Graph-tasks) 由香港科技大学（广州）的 [Yuhan Li](https:\u002F\u002Fyhleeee.github.io\u002F) 维护。\n- [Awesome-Graph-LLM](https:\u002F\u002Fgithub.com\u002FXiaoxinHe\u002FAwesome-Graph-LLM) 由新加坡国立大学的 [Xiaoxin He](https:\u002F\u002Fxiaoxinhe.github.io\u002F) 维护。\n- [Awesome-Graph-Prompt](https:\u002F\u002Fgithub.com\u002FWxxShirley\u002FAwesome-Graph-Prompt)，由香港中文大学的 [Xixi Wu](https:\u002F\u002Fwxxshirley.github.io\u002F) 创建。\n\n\n# 🏆 基准测试\n| 数据集 | 任务 | 论文 | 仓库 |\n| --- | --- | --- | --- |\n| GraphRAG-Bench | GraphRAG 评估 | [[arXiv 2025]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2506.05690) | [[Github]](https:\u002F\u002Fgithub.com\u002FGraphRAG-Bench\u002FGraphRAG-Benchmark) |\n| DIGIMON | 大规模 GraphRAG | [[arXiv 2025]](https:\u002F\u002Fwww.arxiv.org\u002Fpdf\u002F2503.04338) | [[Github]](https:\u002F\u002Fgithub.com\u002FJayLZhou\u002FGraphRAG) |\n| PolyG | GraphRAG 评估 | [[arXiv 2025]](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2504.02112) | [[Github]](https:\u002F\u002Fgithub.com\u002FLiu-rj\u002FPolyG) |\n| SimpleQuestion | 简单问答 | [[arXiv 2015]](https:\u002F\u002FarXiv.org\u002Fabs\u002F1506.02075) | [[Github]](https:\u002F\u002Fgithub.com\u002FJerryzhao-z\u002Fsimple-question-answering-with-memory-networks) |\n| WebQ | 简单问答 | [[EMNLP 2013]](https:\u002F\u002Fnlp.stanford.edu\u002Fpubs\u002FsemparseEMNLP13.pdf) | [[CodaLab]](https:\u002F\u002Fworksheets.codalab.org\u002Fworksheets\u002F0xba659fe363cb46e7a505c5b6a774dc8a) |\n| Multihop-RAG | 多跳推理 | [[COLING 2024]](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2401.15391) | [[Github]](https:\u002F\u002Fgithub.com\u002Fyixuantt\u002FMultiHop-RAG\u002F) |\n| CWQ | 多跳推理 | [[NAACL 2018]](https:\u002F\u002Faclanthology.org\u002FN18-1059\u002F) | [[TAU-NLP]](https:\u002F\u002Fwww.tau-nlp.org\u002Fcompwebq) |\n| MetaQA | 多跳推理 | [[AAAI 2018]](https:\u002F\u002FarXiv.org\u002Fabs\u002F1709.04071) | [[Github]](https:\u002F\u002Fgithub.com\u002Fyuyuz\u002FMetaQA) |\n| MetaQA-3 | 多跳推理 | [[AAAI 2018]](https:\u002F\u002FarXiv.org\u002Fabs\u002F1709.04071) | [[Github]](https:\u002F\u002Fgithub.com\u002Fyuyuz\u002FMetaQA) |\n| CURD | 大规模复杂问答 | [[arXiv 2024]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2401.17043) | [[Github]](https:\u002F\u002Fgithub.com\u002FIAAR-Shanghai\u002FCRUD_RAG) |\n| KQAPro | 大规模复杂问答 | [[ACL 2022]](https:\u002F\u002Faclanthology.org\u002F2022.acl-long.422\u002F) | [[Github]](https:\u002F\u002Fgithub.com\u002Fshijx12\u002FKQAPro_Baselines) |\n| LC-QuAD v2 | 大规模复杂问答 | [[ISWC 2019]](https:\u002F\u002Flink.springer.com\u002Fchapter\u002F10.1007\u002F978-3-030-30796-7_5) | [[figshare]](https:\u002F\u002Ffigshare.com\u002Fprojects\u002FLCQuAD_2_0\u002F62270) |\n| LC-QuAD | 大规模复杂问答 | [[ISWC 2017]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1007\u002F978-3-319-68204-4_22) | [[Github]](https:\u002F\u002Fgithub.com\u002FAskNowQA\u002FLC-QuAD) |\n| UltraDomain | 领域特定问答 | [[arXiv 2024]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2409.05591) | [[Github]](https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002FMemoRAG#dataset) |\n| TutorQA | 领域特定问答 | [[arXiv 2024]](https:\u002F\u002FarXiv.org\u002Fabs\u002F2407.10794) | [[Github]](https:\u002F\u002Fgithub.com\u002FIreneZihuiLi\u002FCGPrompt) |\n| FACTKG | 领域特定问答 | [[ACL 2023]](https:\u002F\u002Faclanthology.org\u002F2023.acl-long.895.pdf) | [[Github]](https:\u002F\u002Fgithub.com\u002Fjiho283\u002FFactKG) |\n| Mintaka | 领域特定问答 | [[ACL 2022]](https:\u002F\u002Faclanthology.org\u002F2022.coling-1.138\u002F) | [[Github]](https:\u002F\u002Fgithub.com\u002Famazon-science\u002Fmintaka) |\n| GrailQA | 领域特定问答 | [[WWW 2021]](https:\u002F\u002Fdl.acm.org\u002Fdoi\u002F10.1145\u002F3442381.3449992) | [[Github]](https:\u002F\u002Fgithub.com\u002Fdki-lab\u002FGrailQA) |\n| WebQSP | 领域特定问答 | [[ACL 2016]](https:\u002F\u002Faclanthology.org\u002FP16-2033.pdf) | [[Microsoft]](http:\u002F\u002Faka.ms\u002FWebQSP) |\n\n# 💻 开源项目\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002FHawksight-AI\u002Fsemantica) Semantica：一个开源、生产就绪的语义层和GraphRAG框架，位于原始语料库与大语言模型之间。\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Fbibinprathap\u002FVeritasGraph) 在本地使用Ollama运行的Graph RAG流水线，具备完整的来源引用功能。\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgraphrag-bench.github.io\u002F) GraphRAG-Bench：面向图检索增强生成的全面基准测试与分析平台。\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002FchensyCN\u002FAgentic-RAG) Agentic-RAG：一个简洁且可扩展的代理式RAG系统。\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Fapecloud\u002FApeRAG) ApeRAG：具备多模态索引、AI代理、MCP支持及可扩展K8s部署的生产级GraphRAG。\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Fgetzep\u002Fgraphiti) Graphiti：为AI代理构建实时知识图谱。\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002FJayLZhou\u002FGraphRAG) DIGIMON：一个统一且模块化的基于图的RAG框架。\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fgraphrag.git) Microsoft-GraphRAG：一个基于图的模块化检索增强生成（RAG）系统。\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Fgusye1234\u002Fnano-graphrag) Nano-GraphRAG：一个简单易用、便于二次开发的GraphRAG实现。\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Fcirclemind-ai\u002Ffast-graphrag) Fast GraphRAG：能够智能适应用户场景、数据及查询的RAG系统。\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002FHKUDS\u002FLightRAG) LightRAG：简单快速的检索增强生成系统。\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Ftpoisonooo\u002FHuixiangDou2) HuixiangDou2：一种经过稳健优化的GraphRAG方法。\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002FFalkorDB\u002FGraphRAG-SDK) GraphRAG-SDK：用于构建GraphRAG系统的专用工具包。\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Fvitali87\u002Fcode-graph-rag) Code-Graph-RAG：一个基于图的RAG系统，利用Tree-sitter分析多语言代码库，构建知识图谱，并通过MCP服务器实现自然语言查询与编辑功能。\n- [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub-100000?style=for-the-badge&logo=github&logoColor=white)](https:\u002F\u002Fgithub.com\u002Fonestardao\u002FWFGY) WFGY问题地图：一个专门的工具包，定义了在RAG和LLM流水线中反复出现的16种失败模式。\n\n\n# 🍀 引用\n如果您觉得本综述有所帮助，请引用我们的论文：\n```\n@article{zhang2025survey,\n  title={针对定制化大型语言模型的图检索增强生成综述},\n  author={张庆刚、陈圣元、贝元辰、袁正、周华驰、洪子衿、董俊楠、陈浩、常毅、黄晓},\n  journal={arXiv预印本 arXiv:2501.13958},\n  year={2025}\n}\n```","# Awesome-GraphRAG 快速上手指南\n\n## 环境准备\n- **系统要求**：任意支持浏览器的系统（Windows\u002FmacOS\u002FLinux）\n- **前置依赖**：无需安装额外依赖，仅需网络访问能力\n- **中国加速方案**：推荐使用 [ghproxy.com](https:\u002F\u002Fghproxy.com) 加速 GitHub 访问（例如将 `https:\u002F\u002Fgithub.com` 替换为 `https:\u002F\u002Fghproxy.com\u002Fhttps:\u002F\u002Fgithub.com`）\n\n## 安装步骤\n无需安装！直接访问仓库：\n```bash\n# 通过国内加速镜像访问（推荐）\nhttps:\u002F\u002Fghproxy.com\u002Fhttps:\u002F\u002Fgithub.com\u002FDEEP-PolyU\u002FAwesome-GraphRAG\n```\n\n## 基本使用\n1. **访问仓库**：打开上述加速链接，或直接访问 [Awesome-GraphRAG 仓库](https:\u002F\u002Fgithub.com\u002FDEEP-PolyU\u002FAwesome-GraphRAG)\n2. **浏览资源**：按目录结构快速定位所需内容：\n   - `📜 Research Papers`：最新论文分类（如 `Knowledge Organization`\u002F`Knowledge Retrieval`）\n   - `🏆 Benchmarks`：GraphRAG 评测基准（如 [GraphRAG-Benchmark](https:\u002F\u002Fgithub.com\u002FGraphRAG-Bench\u002FGraphRAG-Benchmark)）\n   - `💻 Open-source Projects`：开源项目（如 [LinearRAG](https:\u002F\u002Fgithub.com\u002FDEEP-PolyU\u002FLinearRAG)）\n3. **获取论文\u002F项目**：点击任意条目中的 `[[Paper]]` 或 `[[GitHub]]` 链接直达资源\n\n> 💡 示例：查看 **Graphs for Knowledge Indexing** 中的最新论文  \n> 路径：`📜 Research Papers` → `Knowledge Organization` → `Graphs for Knowledge Indexing`  \n> 例如：[LinearRAG (arXiv 2025)](https:\u002F\u002Farxiv.org\u002Fabs\u002F2510.10114)","某医疗科技公司正开发慢性病管理AI助手，需整合数万篇医学文献以提供精准用药建议，但传统RAG模型常因忽略药物-疾病-副作用的关联关系导致回答错误。\n\n### 没有 Awesome-GraphRAG 时\n- 开发团队需在arXiv、GitHub等平台手动搜索分散的GraphRAG论文，平均每周花费15小时，常遗漏关键资源如ICLR'26的LinearRAG方法。\n- 无法系统对比知识图谱构建（如实体关系抽取）与索引图方法的优劣，初期误选低效方案，模型准确率仅68%。\n- 缺乏统一评估基准，团队仅靠人工测试模型性能，迭代周期长达8周，难以量化改进。\n- 研究进展迅猛（如2025年AAAI论文），团队跟不上最新趋势，技术方案滞后3个月。\n- 重复实现关系抽取模块，浪费40%开发时间在基础功能上。\n\n### 使用 Awesome-GraphRAG 后\n- 通过官方资源列表快速定位ICLR'26的LinearRAG论文及代码，10分钟内完成技术选型，节省90%搜索时间。\n- 基于分类资源（如知识组织章节）选择知识图谱构建方案，模型准确率提升至87%。\n- 直接调用GraphRAG Benchmark进行自动化测试，1周内完成模型优化，迭代周期缩短至2周。\n- 实时获取最新动态（如2026年Benchmark论文），确保技术栈始终领先。\n- 复用LinearRAG开源项目，减少60%基础开发工作量，专注业务逻辑优化。\n\nAwesome-GraphRAG将领域AI应用的开发效率提升3倍，让复杂知识关系处理从“摸索”变为“精准落地”。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FDEEP-PolyU_Awesome-GraphRAG_4610c437.png","DEEP-PolyU","PolyU X Lab","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002FDEEP-PolyU_51768cd2.png","",null,"https:\u002F\u002Fwww4.comp.polyu.edu.hk\u002F~xiaohuang\u002F","https:\u002F\u002Fgithub.com\u002FDEEP-PolyU",2260,194,"2026-04-04T19:11:34","MIT",1,"未说明",{"notes":87,"python":87,"dependencies":89},[],[54,13,51,26],[92,93,94,95,96,97,98],"knowledge-graph","large-language-models","retrieval-augmented-generation","graphrag","rag","graphrag-survey","graphrag-paper","2026-03-27T02:49:30.150509","2026-04-06T08:18:32.073314",[102,107,110,113,116],{"id":103,"question_zh":104,"answer_zh":105,"source_url":106},5519,"如何引用 Awesome-GraphRAG 的 survey 论文？","在论文中引用以下 BibTeX 代码：\n@article{zhang2025survey,\n  title={A Survey of Graph Retrieval-Augmented Generation for Customized Large Language Models},\n  author={Zhang, Qinggang and Chen, Shengyuan and Bei, Yuanchen and Yuan, Zheng and Zhou, Huachi and Hong, Zijin and Dong, Junnan and Chen, Hao and Chang, Yi and Huang, Xiao},\n  journal={arXiv preprint arXiv:2501.13958},\n  year={2025}\n}","https:\u002F\u002Fgithub.com\u002FDEEP-PolyU\u002FAwesome-GraphRAG\u002Fissues\u002F17",{"id":108,"question_zh":104,"answer_zh":105,"source_url":109},5520,"https:\u002F\u002Fgithub.com\u002FDEEP-PolyU\u002FAwesome-GraphRAG\u002Fissues\u002F16",{"id":111,"question_zh":104,"answer_zh":105,"source_url":112},5521,"https:\u002F\u002Fgithub.com\u002FDEEP-PolyU\u002FAwesome-GraphRAG\u002Fissues\u002F15",{"id":114,"question_zh":104,"answer_zh":105,"source_url":115},5522,"https:\u002F\u002Fgithub.com\u002FDEEP-PolyU\u002FAwesome-GraphRAG\u002Fissues\u002F11",{"id":117,"question_zh":118,"answer_zh":119,"source_url":112},5523,"如何引用 Awesome-GraphRAG 的 benchmark 论文？","在论文中引用以下 BibTeX 代码：\n@article{xiang2025use,\n  title={When to use Graphs in RAG: A Comprehensive Analysis for Graph Retrieval-Augmented Generation},\n  author={Xiang, Zhishang and Wu, Chuanjie and Zhang, Qinggang and Chen, Shengyuan and Hong, Zijin and Huang, Xiao and Su, Jinsong},\n  journal={arXiv preprint arXiv:2506.05690},\n  year={2025}\n}",[]]