{"id":30620481,"url":"https://github.com/Meirtz/Awesome-Context-Engineering","last_synced_at":"2025-08-30T13:37:53.244Z","repository":{"id":302936916,"uuid":"1012676456","full_name":"Meirtz/Awesome-Context-Engineering","owner":"Meirtz","description":" 🔥 Comprehensive survey on Context Engineering: from prompt engineering to production-grade AI systems. hundreds of papers, frameworks, and  implementation guides for LLMs and AI agents.","archived":false,"fork":false,"pushed_at":"2025-08-05T09:28:45.000Z","size":3322,"stargazers_count":1986,"open_issues_count":6,"forks_count":130,"subscribers_count":29,"default_branch":"main","last_synced_at":"2025-08-26T12:38:32.267Z","etag":null,"topics":["agent","agentic-ai","agi","awesome-list","cognitive-science","context-engineering","llm","rag"],"latest_commit_sha":null,"homepage":"","language":null,"has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/Meirtz.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2025-07-02T17:46:03.000Z","updated_at":"2025-08-26T11:31:59.000Z","dependencies_parsed_at":"2025-07-20T22:08:50.495Z","dependency_job_id":"64f796c1-6f4a-43ac-8ffa-ad36ff1b2ea2","html_url":"https://github.com/Meirtz/Awesome-Context-Engineering","commit_stats":null,"previous_names":["meirtz/awesome-context-engineering"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/Meirtz/Awesome-Context-Engineering","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Meirtz%2FAwesome-Context-Engineering","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Meirtz%2FAwesome-Context-Engineering/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Meirtz%2FAwesome-Context-Engineering/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Meirtz%2FAwesome-Context-Engineering/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Meirtz","download_url":"https://codeload.github.com/Meirtz/Awesome-Context-Engineering/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Meirtz%2FAwesome-Context-Engineering/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":272858486,"owners_count":25005092,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","status":"online","status_checked_at":"2025-08-30T02:00:09.474Z","response_time":77,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["agent","agentic-ai","agi","awesome-list","cognitive-science","context-engineering","llm","rag"],"created_at":"2025-08-30T13:37:51.104Z","updated_at":"2025-08-30T13:37:53.230Z","avatar_url":"https://github.com/Meirtz.png","language":null,"funding_links":[],"categories":["Learning \u0026 Reference","Related Awesome Lists","Agent Integration \u0026 Deployment Tools","Tutorials","Others","🛠️ Tools \u0026 Projects","Context Engineering","Machine Learning \u0026 AI","Curated Resource Lists","Learning \u0026 Documentation","4）安全、评测与相关列表","A01_文本生成_文本对话"],"sub_categories":["Adjacent Collections","AI Agent Development","Context Engineering","Comprehensive Resources","Prompt Attack \u0026 Defense","Comprehensive Guides","相关 Awesome 列表","大语言对话模型及数据"],"readme":"# Awesome Context Engineering\n\n\u003cdiv align=\"center\"\u003e\n  \u003cimg src=\"cover.png\" alt=\"Awesome Context Engineering Cover\" width=\"800\"/\u003e\n\u003c/div\u003e\n\n## 💬 Join Our Community\n\n\u003cdiv align=\"center\"\u003e\n  \u003cimg src=\"assets/wechat_group.png\" alt=\"WeChat Group\" width=\"200\"/\u003e\n  \u003cp\u003e\u003cstrong\u003eJoin our WeChat group for discussions and updates!\u003c/strong\u003e\u003c/p\u003e\n  \u003cp\u003e\u003ca href=\"https://discord.gg/fsqs3Ybh\"\u003e\u003cstrong\u003eJoin our Discord server\u003c/strong\u003e\u003c/a\u003e\u003c/p\u003e\n\u003c/div\u003e\n\n[![Awesome](https://awesome.re/badge.svg)](https://awesome.re)\n[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)\n[![PRs Welcome](https://img.shields.io/badge/PRs-welcome-brightgreen.svg?style=flat-square)](http://makeapullrequest.com)\n[![Paper](https://img.shields.io/badge/Paper-Published-green.svg)](https://arxiv.org/abs/2507.13334)\n\n\u003e 📄 **Our comprehensive survey paper on Context Engineering is now published!** Check out our latest academic insights and theoretical foundations.\n\nA comprehensive survey and collection of resources on **Context Engineering** - the evolution from static prompting to dynamic, context-aware AI systems.\n\n## 📧 Contact\n\nFor questions, suggestions, or collaboration opportunities, please feel free to reach out:\n\n**Lingrui Mei**  \n📧 Email:  [meilingrui25b@ict.ac.cn](mailto:meilingrui25b@ict.ac.cn) or [meilingrui22@mails.ucas.ac.cn](mailto:meilingrui22@mails.ucas.ac.cn)\n\n**I WROTE THE WRONG EMAIL ADDRESS IN THE FIRST VERSION OF MY PAPER!!** You can also open an issue in this repository for general discussions and suggestions.\n\n---\n\n## 📰 News\n\n- **[2025.07.17]** 🔥🔥 Our paper is now published! Check out [\"A Survey of Context Engineering for Large Language Models\"](https://arxiv.org/abs/2507.13334) on [arXiv](https://arxiv.org/abs/2507.13334) and [Hugging Face Papers](https://huggingface.co/papers/2507.13334)\n- **[2025.07.03]** Repository initialized with comprehensive outline\n- **[2025.07.03]** Survey structure established following modern context engineering paradigms\n\n---\n\n## 🎯 Introduction\n\nIn the era of Large Language Models (LLMs), the limitations of static prompting have become increasingly apparent. **Context Engineering** represents the natural evolution to address LLM uncertainty and achieve production-grade AI deployment. Unlike traditional prompt engineering, context engineering encompasses the complete information payload provided to LLMs at inference time, including all structured informational components necessary for plausible task completion.\n\nThis repository serves as a comprehensive survey of context engineering techniques, methodologies, and applications.\n\n---\n\n## 📚 Table of Contents\n\n- [Awesome Context Engineering](#awesome-context-engineering)\n  - [💬 Join Our Community](#-join-our-community)\n  - [📧 Contact](#-contact)\n  - [📰 News](#-news)\n  - [🎯 Introduction](#-introduction)\n  - [📚 Table of Contents](#-table-of-contents)\n  - [🔗 Related Survey](#-related-survey)\n  - [🏗️ Definition of Context Engineering](#️-definition-of-context-engineering)\n    - [LLM Generation](#llm-generation)\n    - [Definition of Context](#definition-of-context)\n    - [Definition of Context Engineering](#definition-of-context-engineering)\n    - [Dynamic Context Orchestration](#dynamic-context-orchestration)\n    - [Mathematical Principles](#mathematical-principles)\n    - [Theoretical Framework: Bayesian Context Inference](#theoretical-framework-bayesian-context-inference)\n    - [Comparison](#comparison)\n  - [🌐 Related Blogs](#-related-blogs)\n    - [Social Media \\\u0026 Talks](#social-media--talks)\n  - [🤔 Why Context Engineering?](#-why-context-engineering)\n    - [The Paradigm Shift: From Tactical to Strategic](#the-paradigm-shift-from-tactical-to-strategic)\n    - [1. Fundamental Challenges with Current Approaches](#1-fundamental-challenges-with-current-approaches)\n      - [Human Intent Communication Challenges](#human-intent-communication-challenges)\n      - [Complex Knowledge Requirements](#complex-knowledge-requirements)\n      - [Reliability and Trustworthiness Issues](#reliability-and-trustworthiness-issues)\n    - [2. Limitations of Static Prompting](#2-limitations-of-static-prompting)\n      - [From Strings to Systems](#from-strings-to-systems)\n      - [The \"Movie Production\" Analogy](#the-movie-production-analogy)\n    - [3. Enterprise and Production Requirements](#3-enterprise-and-production-requirements)\n      - [Context Failures Are the New Bottleneck](#context-failures-are-the-new-bottleneck)\n      - [Scalability Beyond Simple Tasks](#scalability-beyond-simple-tasks)\n      - [Reliability and Consistency](#reliability-and-consistency)\n      - [Economic and Operational Efficiency](#economic-and-operational-efficiency)\n    - [4. Cognitive and Information Science Foundations](#4-cognitive-and-information-science-foundations)\n      - [Artificial Embodiment](#artificial-embodiment)\n      - [Information Retrieval at Scale](#information-retrieval-at-scale)\n    - [5. The Future of AI System Architecture](#5-the-future-of-ai-system-architecture)\n  - [🔧 Components, Techniques and Architectures](#-components-techniques-and-architectures)\n    - [Context Scaling](#context-scaling)\n    - [Structured Data Integration](#structured-data-integration)\n    - [Self-Generated Context](#self-generated-context)\n  - [🛠️ Implementation and Challenges](#️-implementation-and-challenges)\n    - [1. Retrieval-Augmented Generation (RAG)](#1-retrieval-augmented-generation-rag)\n    - [2. Memory Systems](#2-memory-systems)\n    - [3. Agent Communication](#3-agent-communication)\n    - [4. Tool Use and Function Calling](#4-tool-use-and-function-calling)\n  - [📊 Evaluation Paradigms for Context-Driven Systems](#-evaluation-paradigms-for-context-driven-systems)\n    - [Context Quality Assessment](#context-quality-assessment)\n    - [Benchmarking Context Engineering](#benchmarking-context-engineering)\n  - [🚀 Applications and Systems](#-applications-and-systems)\n    - [Complex Research Systems](#complex-research-systems)\n    - [Production Systems](#production-systems)\n  - [🔮 Limitations and Future Directions](#-limitations-and-future-directions)\n    - [Current Limitations](#current-limitations)\n    - [Future Research Directions](#future-research-directions)\n  - [🤝 Contributing](#-contributing)\n    - [Paper Formatting Guidelines](#paper-formatting-guidelines)\n    - [Badge Colors](#badge-colors)\n  - [📄 License](#-license)\n  - [📑 Citation](#-citation)\n  - [⚠️ Disclaimer](#️-disclaimer)\n  - [📧 Contact](#-contact-1)\n  - [🙏 Acknowledgments](#-acknowledgments)\n  - [Star History](#star-history)\n  - [📖 Our Paper](#-our-paper)\n\n---\n\n## 🔗 Related Survey\n\n\u003cb\u003eGeneral AI Survey Papers\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eA Survey of Large Language Models\u003c/b\u003e\u003c/i\u003e, Zhao et al.,\u003ca href=\"https://arxiv.org/abs/2303.18223\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.03-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/RUCAIBox/LLMSurvey\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/RUCAIBox/LLMSurvey.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eThe Prompt Report: A Systematic Survey of Prompt Engineering Techniques\u003c/b\u003e\u003c/i\u003e, Schulhoff et al., \u003ca href=\"https://arxiv.org/abs/2406.06608\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/trigaten/The_Prompt_Report\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/trigaten/The_Prompt_Report.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eA Systematic Survey of Prompt Engineering in Large Language Models: Techniques and Applications\u003c/b\u003e\u003c/i\u003e, Sahoo et al., \u003ca href=\"https://arxiv.org/abs/2402.07927\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.03-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eA Systematic Survey of Prompt Engineering on Vision-Language Foundation Models\u003c/b\u003e\u003c/i\u003e, Gao et al., \u003ca href=\"https://arxiv.org/abs/2307.12980\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.07-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/JindongGu/Awesome-Prompting-on-Vision-Language-Model\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/JindongGu/Awesome-Prompting-on-Vision-Language-Model.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\n\u003cb\u003eContext and Reasoning\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eA Survey on In-context Learning\u003c/b\u003e\u003c/i\u003e, Dong et al., \u003ca href=\"https://doi.org/10.18653/v1/2024.emnlp-main.64\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/EMNLP-2024.11-blue\" alt=\"EMNLP Badge\"\u003e\u003c/a\u003e\n\u003ca href=\"https://github.com/dqxiu/ICL_PaperList\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/dqxiu/ICL_PaperList.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eThe Mystery of In-Context Learning: A Comprehensive Survey on Interpretation and Analysis\u003c/b\u003e\u003c/i\u003e, Zhou et al., \u003ca href=\"https://arxiv.org/abs/2311.00237\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.10-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003ca href=\"https://github.com/zyxnlp/ICL-Interpretation-Analysis-Resources\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/zyxnlp/ICL-Interpretation-Analysis-Resources.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eA Comprehensive Survey of Retrieval-Augmented Generation (RAG): Evolution, Current Landscape and Future Directions\u003c/b\u003e\u003c/i\u003e, Gupta et al., \u003ca href=\"https://arxiv.org/abs/2410.12837\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.10-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRetrieval-Augmented Generation for Large Language Models: A Survey\u003c/b\u003e\u003c/i\u003e, Gao et al., \u003ca href=\"https://arxiv.org/abs/2312.10997\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.03-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003ca href=\"https://github.com/Tongji-KGLLM/RAG-Survey\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/Tongji-KGLLM/RAG-Survey.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eA Survey on Knowledge-Oriented Retrieval-Augmented Generation\u003c/b\u003e\u003c/i\u003e, Cheng et al., \u003ca href=\"https://arxiv.org/abs/2503.10677\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003ca href=\"https://github.com/USTCAGI/Awesome-Papers-Retrieval-Augmented-Generation\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/USTCAGI/Awesome-Papers-Retrieval-Augmented-Generation.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\n\u003cb\u003eMemory Systems and Context Persistence\u003c/b\u003e\n\n\u003cb\u003eSurvey\u003c/b\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eA Survey on the Memory Mechanism of Large Language Model based Agents\u003c/b\u003e\u003c/i\u003e, Zhang et al., \u003ca href=\"https://arxiv.org/abs/2404.13501\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.04-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/nuster1128/LLM_Agent_Memory_Survey\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/nuster1128/LLM_Agent_Memory_Survey.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eSurvey on Memory-Augmented Neural Networks: Cognitive Insights to AI Applications\u003c/b\u003e\u003c/i\u003e, Khosla et al., \u003ca href=\"https://arxiv.org/abs/2312.06141\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.12-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eFrom Human Memory to AI Memory: A Survey on Memory Mechanisms in the Era of LLMs\u003c/b\u003e\u003c/i\u003e, Wu et al., \u003ca href=\"https://arxiv.org/abs/2504.15965\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.04-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eSurvey on Evaluation of LLM-based Agents\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2503.16416\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.03-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eA Survey of Personalized Large Language Models: Progress and Future Directions\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2502.11528\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eAgentic Retrieval-Augmented Generation: A Survey\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2501.09136\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.01-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRetrieval-Augmented Generation with Graphs (GraphRAG)\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2501.00309\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.12-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/Graph-RAG/GraphRAG/\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/Graph-RAG/GraphRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\n\u003cb\u003eBenchmarks\u003c/b\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eEvaluating Very Long-Term Conversational Memory of LLM Agents (LOCOMO)\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2402.17753\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ACL-2024.02-blue\" alt=\"ACL Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://snap-research.github.io/locomo/\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/snap-research/locomo.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eEvaluating Memory in LLM Agents via Incremental Multi-Turn Interactions\u003c/b\u003e\u003c/i\u003e, Hu et al.,\u003ca href=\"https://arxiv.org/abs/2507.05257\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.07-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/HUST-AI-HYZ/MemoryAgentBench\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/HUST-AI-HYZ/MemoryAgentBench.svg?style=social\" alt=\"GitHub stars\"\u003e\u003c/a\u003e\n      \u003ca href=\"https://huggingface.co/datasets/ai-hyz/MemoryAgentBench\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://huggingface.co/datasets/huggingface/badges/resolve/main/dataset-on-hf-sm.svg\" alt=\"HF Dataset\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eEpisodic Memories Generation and Evaluation Benchmark for Large Language Models\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2501.13121\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.01-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eOn the Structural Memory of LLM Agents\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2412.15266\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.12-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eHotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering\u003c/b\u003e\u003c/i\u003e, Yang et al., \u003ca href=\"https://arxiv.org/abs/1809.09600\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/EMNLP-2018.09-blue\" alt=\"EMNLP Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://hotpotqa.github.io/\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/hotpotqa/hotpot.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cb\u003eNeural Memory Architectures\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eNeural Turing Machines\u003c/b\u003e\u003c/i\u003e, Graves et al., \u003ca href=\"https://arxiv.org/abs/1410.5401\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2014.10-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eDifferentiable Neural Computers\u003c/b\u003e\u003c/i\u003e, Graves et al., \u003ca href=\"https://arxiv.org/abs/1610.06258\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2016.10-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/google-deepmind/dnc\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/google-deepmind/dnc.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eA Brain-inspired Memory Transformation based Differentiable Neural Computer\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2301.02809\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.01-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eDifferentiable Neural Computers with Memory Demon\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2211.02987\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2022.11-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003c/ul\u003e\n\u003cb\u003eMemory-Augmented Transformers\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMemorizing Transformers\u003c/b\u003e\u003c/i\u003e, Wu et al., \u003ca href=\"https://arxiv.org/abs/2203.08913\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2022.03-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRecurrent Memory Transformer\u003c/b\u003e\u003c/i\u003e, Bulatov et al., \u003ca href=\"https://arxiv.org/abs/2207.06881\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/NeurIPS-2022.07-blue\" alt=\"NeurIPS Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/booydar/recurrent-memory-transformer\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/booydar/recurrent-memory-transformer.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLeave No Context Behind: Efficient Infinite Context Transformers with Infini-attention\u003c/b\u003e\u003c/i\u003e, Munkhdalai et al., \u003ca href=\"https://arxiv.org/abs/2404.07143\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.04-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMemformer: A Memory-Augmented Transformer for Sequence Modeling\u003c/b\u003e\u003c/i\u003e, Wu et al., \u003ca href=\"https://arxiv.org/abs/2010.06891\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2020.10-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eToken Turing Machines\u003c/b\u003e\u003c/i\u003e, Ryoo et al., \u003ca href=\"https://arxiv.org/abs/2211.09119\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2022.11-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eTransformerFAM: Feedback Attention is Working Memory\u003c/b\u003e\u003c/i\u003e, Irie et al., \u003ca href=\"https://arxiv.org/abs/2404.09173\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.04-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003c/ul\u003e\n\n\u003cb\u003eProduction Memory Systems\u003c/b\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMemGPT: Towards LLMs as Operating Systems\u003c/b\u003e\u003c/i\u003e, Packer et al., \u003ca href=\"https://arxiv.org/abs/2310.08560\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://research.memgpt.ai\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/letta-ai/letta.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMemoryBank: Enhancing Large Language Models with Long-Term Memory\u003c/b\u003e\u003c/i\u003e, Zhong et al., \u003ca href=\"https://arxiv.org/abs/2305.10250\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.05-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/zhongwanjun/memorybank-siliconfriend\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/zhongwanjun/memorybank-siliconfriend.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMEM0: Building Production-Ready AI Agents with Scalable Long-Term Memory\u003c/b\u003e\u003c/i\u003e, Taranjeet et al., \u003ca href=\"https://arxiv.org/abs/2504.19413\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.04-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://mem0.ai/research\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/mem0ai/mem0.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMEM1: Learning to Synergize Memory and Reasoning for Efficient Long-Horizon Agents\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2506.15841\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.06-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/mannaandpoem/openmanus\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/mannaandpoem/openmanus.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eA-MEM: Agentic Memory for LLM Agents\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2502.12110\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/agiresearch/A-mem\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/agiresearch/A-mem.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory Agent\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2507.02259\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.07-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMemory OS of AI Agent\u003c/b\u003e\u003c/i\u003e, Kang et al., \u003ca href=\"https://arxiv.org/abs/2506.06326\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.05-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/BAI-LAB/MemoryOS\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/BAI-LAB/MemoryOS.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cb\u003eGraph-based Memory Systems\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003earigraph: learning knowledge graph world models with episodic memory for llm agents\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2407.04363\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.07-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eZep: A Temporal Knowledge Graph Architecture for Agent Memory\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2501.13956\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.01-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/getzep/graphiti\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/getzep/graphiti.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eKG-Agent: An Efficient Autonomous Agent Framework for Complex Reasoning over Knowledge Graph\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2402.11163\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eGraphReader: Building Graph-based Agent to Enhance Long-Context Abilities of Large Language Models\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2406.14550\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.06-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eFrom Local to Global: A GraphRAG Approach to Query-Focused Summarization\u003c/b\u003e\u003c/i\u003e, Edge et al., \u003ca href=\"https://arxiv.org/abs/2404.16130\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.04-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/microsoft/graphrag\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/microsoft/graphrag.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eKnowledge Graph-Guided Retrieval Augmented Generation\u003c/b\u003e\u003c/i\u003e, Zhu et al., \u003ca href=\"https://arxiv.org/abs/2502.06864\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003c/ul\u003e\n\u003cb\u003eEpisodic and Working Memory\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLarimar: Large Language Models with Episodic Memory Control\u003c/b\u003e\u003c/i\u003e, Goyal et al., \u003ca href=\"https://arxiv.org/abs/2403.11901\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ICML-2024.03-blue\" alt=\"ICML Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eEM-LLM: Human-like Episodic Memory for Infinite Context LLMs\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2407.09450\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ICLR-2024.07-blue\" alt=\"ICLR Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/em-llm/EM-LLM-model\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/em-llm/EM-LLM-model.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLarge Language Models with Controllable Working Memory\u003c/b\u003e\u003c/i\u003e, Goyal et al., \u003ca href=\"https://arxiv.org/abs/2211.05110\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2022.11-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eEmpowering Working Memory for Large Language Model Agents\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2312.17259\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.12-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003c/ul\u003e\n\u003cb\u003eConversational Memory\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMemoChat: Tuning LLMs to Use Memos for Consistent Long-Range Open-Domain Conversation\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2308.08239\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.08-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eThink-in-Memory: Recalling and Post-thinking Enable LLMs with Long-Term Memory\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2311.08719\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.11-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eGenerative Agents: Interactive Simulacra of Human Behavior\u003c/b\u003e\u003c/i\u003e, Park et al., \u003ca href=\"https://arxiv.org/abs/2304.03442\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.04-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eSelf-Controlled Memory Framework for Large Language Models\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2304.13343\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.04-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003c/ul\u003e\n\u003cb\u003eFoundational Survey Papers from Major Venues\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eAUTOPROMPT: Eliciting Knowledge from Language Models with Automatically Generated Prompts\u003c/b\u003e\u003c/i\u003e, Shin et al., \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/EMNLP-2020-blue\" alt=\"EMNLP Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/ucinlp/autoprompt\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/ucinlp/autoprompt.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eThe Power of Scale for Parameter-Efficient Prompt Tuning\u003c/b\u003e\u003c/i\u003e, Lester et al., \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/EMNLP-2021-blue\" alt=\"EMNLP Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/google-research/prompt-tuning\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/google-research/prompt-tuning.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003ePrefix-Tuning: Optimizing Continuous Prompts for Generation\u003c/b\u003e\u003c/i\u003e, Li et al., \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ACL-2021-blue\" alt=\"ACL Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/XiangLi1999/PrefixTuning\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/XiangLi1999/PrefixTuning.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eAn Explanation of In-context Learning as Implicit Bayesian Inference\u003c/b\u003e\u003c/i\u003e, Xie et al., \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ICLR-2022-blue\" alt=\"ICLR Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/p-lambda/incontext-learning\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/p-lambda/incontext-learning.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRethinking the Role of Demonstrations: What Makes In-context Learning Work?\u003c/b\u003e\u003c/i\u003e, Min et al., \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/EMNLP-2022-blue\" alt=\"EMNLP Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/Alrope123/rethinking-demonstrations\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/Alrope123/rethinking-demonstrations.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\n\n\u003cb\u003eAdditional RAG and Retrieval Surveys\u003c/b\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRetrieval-Augmented Generation for AI-Generated Content: A Survey\u003c/b\u003e\u003c/i\u003e, Various, \u003ca href=\"https://arxiv.org/abs/2402.19473\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/PKU-DAIR/RAG-Survey\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/PKU-DAIR/RAG-Survey.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRetrieval Augmented Generation (RAG) and Beyond: A Comprehensive Survey on How to Make your LLMs use External Data More Wisely\u003c/b\u003e\u003c/i\u003e, Various, \u003ca href=\"https://arxiv.org/abs/2409.14924\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.09-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLarge language models (LLMs): survey, technical frameworks, and future challenges\u003c/b\u003e\u003c/i\u003e, Various, \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/AIR-2024-blue\" alt=\"AIR Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003c/ul\u003e\n\n\n---\n\n## 🏗️ Definition of Context Engineering\n\n\u003e **Context is not just the single prompt users send to an LLM. Context is the complete information payload provided to a LLM at inference time, encompassing all structured informational components that the model needs to plausibly accomplish a given task.**\n\n### LLM Generation\n\nTo formally define Context Engineering, we must first mathematically characterize the LLM generation process. Let us model an LLM as a probabilistic function:\n\n$$P(\\text{output} | \\text{context}) = \\prod_{t=1}^T P(\\text{token}_t | \\text{previous tokens}, \\text{context})$$\n\nWhere:\n- $\\text{context}$ represents the complete input information provided to the LLM\n- $\\text{output}$ represents the generated response sequence\n- $P(\\text{token}_t | \\text{previous tokens}, \\text{context})$ is the probability of generating each token given the context\n\n### Definition of Context\n\nIn traditional prompt engineering, the context is treated as a simple string:\n$$\\text{context} = \\text{prompt}$$\n\nHowever, in Context Engineering, we decompose the context into multiple structured components:\n\n$$\\text{context} = \\text{Assemble}(\\text{instructions}, \\text{knowledge}, \\text{tools}, \\text{memory}, \\text{state}, \\text{query})$$\n\nWhere $\\text{Assemble}$ is a context assembly function that orchestrates:\n- $\\text{instructions}$: System prompts and rules\n- $\\text{knowledge}$: Retrieved relevant information\n- $\\text{tools}$: Available function definitions\n- $\\text{memory}$: Conversation history and learned facts\n- $\\text{state}$: Current world/user state\n- $\\text{query}$: User's immediate request\n\n### Definition of Context Engineering\n\n**Context Engineering** is formally defined as the optimization problem:\n\n$$\\text{Assemble}^* = \\arg\\max_{\\text{Assemble}} \\mathbb{E} [\\text{Reward}(\\text{LLM}(\\text{context}), \\text{target})]$$\n\nSubject to constraints:\n- $|\\text{context}| \\leq \\text{MaxTokens} \\text{(context window limitation)}$\n- $\\text{knowledge} = \\text{Retrieve}(\\text{query}, \\text{database})$\n- $\\text{memory} = \\text{Select}(\\text{history}, \\text{query})$\n- $\\text{state} = \\text{Extract}(\\text{world})$\n\nWhere:\n- $\\text{Reward}$ measures the quality of generated responses\n- $\\text{Retrieve}$, $\\text{Select}$, $\\text{Extract}$ are functions for information gathering\n\n### Dynamic Context Orchestration\n\nThe context assembly can be decomposed as:\n\n$$\\text{context} = \\text{Concat}(\\text{Format}(\\text{instructions}), \\text{Format}(\\text{knowledge}), \\text{Format}(\\text{tools}), \\text{Format}(\\text{memory}), \\text{Format}(\\text{query}))$$\n\nWhere $\\text{Format}$ represents component-specific structuring, and $\\text{Concat}$ assembles them respecting token limits and optimal positioning.\n\n**Context Engineering** is therefore the discipline of designing and optimizing these assembly and formatting functions to maximize task performance.\n\n### Mathematical Principles\n\nFrom this formalization, we derive four fundamental principles:\n\n1. **System-Level Optimization**: Context generation is a multi-objective optimization problem over assembly functions, not simple string manipulation.\n\n2. **Dynamic Adaptation**: The context assembly function adapts to each $\\text{query}$ and $\\text{state}$ at inference time: $\\text{Assemble}(\\cdot | \\text{query}, \\text{state})$.\n\n3. **Information-Theoretic Optimality**: The retrieval function maximizes relevant information: $\\text{Retrieve} = \\arg\\max \\text{Relevance}(\\text{knowledge}, \\text{query})$.\n\n4. **Structural Sensitivity**: The formatting functions encode structure that aligns with LLM processing capabilities.\n\n### Theoretical Framework: Bayesian Context Inference\n\nContext Engineering can be formalized within a Bayesian framework where the optimal context is inferred:\n\n$$P(\\text{context} | \\text{query}, \\text{history}, \\text{world}) \\propto P(\\text{query} | \\text{context}) \\cdot P(\\text{context} | \\text{history}, \\text{world})$$\n\nWhere:\n- $P(\\text{query} | \\text{context})$ models query-context compatibility\n- $P(\\text{context} | \\text{history}, \\text{world})$ represents prior context probability\n\nThe optimal context assembly becomes:\n\n$$\\text{context}^* = \\arg\\max_{\\text{context}} P(\\text{answer} | \\text{query}, \\text{context}) \\cdot P(\\text{context} | \\text{query}, \\text{history}, \\text{world})$$\n\nThis Bayesian formulation enables:\n- **Uncertainty Quantification**: Modeling confidence in context relevance\n- **Adaptive Retrieval**: Updating context beliefs based on feedback\n- **Multi-step Reasoning**: Maintaining context distributions across interactions\n\n### Comparison\n\n| Dimension | Prompt Engineering | Context Engineering |\n|-----------|-------------------|-------------------|\n| **Mathematical Model** | $\\text{context} = \\text{prompt}$ (static) | $\\text{context} = \\text{Assemble}(...)$ (dynamic) |\n| **Optimization Target** | $\\arg\\max_{\\text{prompt}} P(\\text{answer} \\mid \\text{query}, \\text{prompt})$ | $\\arg\\max_{\\text{Assemble}} \\mathbb{E}[\\text{Reward}(...)]$ |\n| **Complexity** | $O(1)$ context assembly | $O(n)$ multi-component optimization |\n| **Information Theory** | Fixed information content | Adaptive information maximization |\n| **State Management** | Stateless function | Stateful with $\\text{memory}(\\text{history}, \\text{query})$ |\n| **Scalability** | Linear in prompt length | Sublinear through compression/filtering |\n| **Error Analysis** | Manual prompt inspection | Systematic evaluation of assembly components |\n\n\n\n---\n\n## 🌐 Related Blogs\n\n- [The rise of \"context engineering\"](https://blog.langchain.com/the-rise-of-context-engineering/)\n- [The New Skill in AI is Not Prompting, It's Context Engineering](https://www.philschmid.de/context-engineering)\n- [davidkimai/Context-Engineering: \"Context engineering is the delicate art and science of filling the context window with just the right information for the next step.\" ](https://github.com/davidkimai/Context-Engineering)\n- [Context Engineering is Runtime of AI Agents | by Bijit Ghosh | Jun, 2025 | Medium](https://medium.com/@bijit211987/context-engineering-is-runtime-of-ai-agents-411c9b2ef1cb)\n- [Context Engineering](https://blog.langchain.com/context-engineering-for-agents/)\n- [Context Engineering for Agents](https://rlancemartin.github.io/2025/06/23/context_engineering/)\n- [Cognition | Don't Build Multi-Agents](https://cognition.ai/blog/dont-build-multi-agents)\n- [从Prompt Engineering到Context Engineering - 53AI-AI知识库|大模型知识库|大模型训练|智能体开发](https://www.53ai.com/news/tishicikuangjia/2025062727685.html)\n\n### Social Media \u0026 Talks\n\n- [Mastering Claude Code in 30 minutes](https://www.youtube.com/watch?v=6eBSHbLKuN0)\n- [Context Engineering for Agents](https://www.youtube.com/watch?v=4GiqzUHD5AA)\n- [Andrej Karpathy on X: \"+1 for \"context engineering\" over \"prompt engineering\"](https://x.com/karpathy/status/1937902205765607626?ref=blog.langchain.com)\n- [复旦大学/上海创智学院邱锡鹏：Context Scaling，通往AGI的下一幕](https://mp.weixin.qq.com/s/Knej0qbyr5j5KX_BO7FGew)\n\n---\n\n## 🤔 Why Context Engineering?\n\n### The Paradigm Shift: From Tactical to Strategic\n\nThe evolution from prompt engineering to context engineering represents a fundamental maturation in AI system design. As influential figures like Andrej Karpathy, Tobi Lutke, and Simon Willison have argued, the term \"prompt engineering\" has been diluted to mean simply \"typing things into a chatbot,\" failing to capture the complexity required for industrial-strength LLM applications.\n\n### 1. Fundamental Challenges with Current Approaches\n\n#### Human Intent Communication Challenges\n- **Unclear Human Intent Expression**: Human intentions are often unclear, incomplete, or ambiguous when expressed in natural language\n- **AI's Incomplete Understanding of Human Intent**: AI systems struggle to fully comprehend complex human intentions, especially those involving implicit context or cultural nuances\n- **Overly Literal AI Interpretation**: AI systems often interpret human instructions too literally, missing the underlying intent or contextual meaning\n\n#### Complex Knowledge Requirements\nSingle models alone cannot solve complex problems that require:\n- **(1) Large-scale External Knowledge**: Vast amounts of external knowledge that exceed model capacity\n- **(2) Accurate External Knowledge**: Precise, up-to-date information that models may not possess\n- **(3) Novel External Knowledge**: Emerging knowledge that appears after model training\n\n**Static Knowledge Limitations:**\n- **Static Knowledge Problem**: Pre-trained models contain static knowledge that becomes outdated\n- **Knowledge Cutoff**: Models cannot access information beyond their training data\n- **Domain-Specific Gaps**: Models lack specialized knowledge for specific industries or applications\n\n#### Reliability and Trustworthiness Issues\n- **AI Hallucination**: LLMs generate plausible but factually incorrect information when lacking proper context\n- **Lack of Provenance**: Absence of clear source attribution for generated information\n- **Confidence Calibration**: Models often appear confident even when generating false information\n- **Transparency Gaps**: Inability to trace how conclusions were reached\n- **Accountability Issues**: Difficulty in verifying the reliability of AI-generated content\n\n### 2. Limitations of Static Prompting\n\n#### From Strings to Systems\nTraditional prompting treats context as a static string, but enterprise applications require:\n- **Dynamic Information Assembly**: Context created on-the-fly, tailored to specific users and queries\n- **Multi-Source Integration**: Combining databases, APIs, documents, and real-time data\n- **State Management**: Maintaining conversation history, user preferences, and workflow status\n- **Tool Orchestration**: Coordinating external function calls and API interactions\n\n#### The \"Movie Production\" Analogy\nIf prompt engineering is writing a single line of dialogue for an actor, context engineering is the entire process of building the set, designing lighting, providing detailed backstory, and directing the scene. The dialogue only achieves its intended impact because of the rich, carefully constructed environment surrounding it.\n\n### 3. Enterprise and Production Requirements\n\n#### Context Failures Are the New Bottleneck\nMost failures in modern agentic systems are no longer attributable to core model reasoning capabilities but are instead **\"context failures\"**. The true engineering challenge lies not in what question to ask, but in ensuring the model has all necessary background, data, tools, and memory to answer meaningfully and reliably.\n\n#### Scalability Beyond Simple Tasks\nWhile prompt engineering suffices for simple, self-contained tasks, it breaks down when scaled to:\n- **Complex, multi-step applications**\n- **Data-rich enterprise environments** \n- **Stateful, long-running workflows**\n- **Multi-user, multi-tenant systems**\n\n#### Reliability and Consistency\nEnterprise applications demand:\n- **Deterministic Behavior**: Predictable outputs across different contexts and users\n- **Error Handling**: Graceful degradation when information is incomplete or contradictory\n- **Audit Trails**: Transparency in how context influences model decisions\n- **Compliance**: Meeting regulatory requirements for data handling and decision making\n\n#### Economic and Operational Efficiency\nContext Engineering enables:\n- **Cost Optimization**: Strategic choice between RAG and long-context approaches\n- **Latency Management**: Efficient information retrieval and context assembly\n- **Resource Utilization**: Optimal use of finite context windows and computational resources\n- **Maintenance Scalability**: Systematic approaches to updating and managing knowledge bases\n\nContext Engineering provides the architectural foundation for managing state, integrating diverse data sources, and maintaining coherence across these demanding scenarios.\n\n### 4. Cognitive and Information Science Foundations\n\n#### Artificial Embodiment\nLLMs are essentially \"brains in a vat\" - powerful reasoning engines lacking connection to specific environments. Context Engineering provides:\n- **Synthetic Sensory Systems**: Retrieval mechanisms as artificial perception\n- **Proxy Embodiment**: Tool use as artificial action capabilities  \n- **Artificial Memory**: Structured information storage and retrieval\n\n#### Information Retrieval at Scale\nContext Engineering addresses the fundamental challenge of information retrieval where the \"user\" is not human but an AI agent. This requires:\n- **Semantic Understanding**: Bridging the gap between intent and expression\n- **Relevance Optimization**: Ranking and filtering vast knowledge bases\n- **Query Transformation**: Converting ambiguous requests into precise retrieval operations\n\n### 5. The Future of AI System Architecture\n\nContext Engineering elevates AI development from a collection of \"prompting tricks\" to a rigorous discipline of systems architecture. It applies decades of knowledge in operating system design, memory management, and distributed systems to the unique challenges of LLM-based applications.\n\nThis discipline is foundational for unlocking the full potential of LLMs in production systems, enabling the transition from one-off text generation to autonomous agents and sophisticated AI copilots that can reliably operate in complex, dynamic environments.\n\n---\n\n## 🔧 Components, Techniques and Architectures\n\n### Context Scaling\n\n\u003cb\u003ePosition Interpolation and Extension Techniques\u003c/b\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eExtending Context Window of Large Language Models via Position Interpolation\u003c/b\u003e\u003c/i\u003e, Chen et al., \u003ca href=\"https://arxiv.org/abs/2306.15595\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.06-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/Math1019/Extend_Context_Window_Position_Interpolation\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/Math1019/Extend_Context_Window_Position_Interpolation.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eYaRN: Efficient Context Window Extension of Large Language Models\u003c/b\u003e\u003c/i\u003e, Peng et al., \u003ca href=\"https://arxiv.org/abs/2309.00071\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ICLR-2024.01-blue\" alt=\"ICLR Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/jquesnelle/yarn\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/jquesnelle/yarn.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLongRoPE: Extending LLM Context Window Beyond 2 Million Tokens\u003c/b\u003e\u003c/i\u003e, Ding et al., \u003ca href=\"https://arxiv.org/abs/2402.13753\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ICML-2024.02-blue\" alt=\"ICML Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/microsoft/LongRoPE\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/microsoft/LongRoPE.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLongRoPE2: Near-Lossless LLM Context Window Scaling\u003c/b\u003e\u003c/i\u003e, Shang et al., \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ICML-2025.05-blue\" alt=\"ICML Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/microsoft/LongRoPE\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/microsoft/LongRoPE.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\n\u003cb\u003eMemory-Efficient Attention Mechanisms\u003c/b\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eFast Multipole Attention: A Divide-and-Conquer Attention Mechanism for Long Sequences\u003c/b\u003e\u003c/i\u003e, Kang et al., \u003ca href=\"https://arxiv.org/abs/2310.11960\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ICLR-2024.02-blue\" alt=\"ICLR Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/yanmingk/FMA\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/yanmingk/FMA.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLeave No Context Behind: Efficient Infinite Context Transformers with Infini-attention\u003c/b\u003e\u003c/i\u003e, Munkhdalai et al., \u003ca href=\"https://arxiv.org/abs/2404.07143\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.04-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/jlamprou/Infini-Attention\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/jlamprou/Infini-Attention.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eDuoAttention: Efficient Long-Context LLM Inference with Retrieval and Streaming Heads\u003c/b\u003e\u003c/i\u003e, Xiao et al., \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ICLR-2025.01-blue\" alt=\"ICLR Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/mit-han-lab/duo-attention\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/mit-han-lab/duo-attention.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eStar Attention: Efficient LLM Inference over Long Sequences\u003c/b\u003e\u003c/i\u003e, Acharya et al., \u003ca href=\"https://arxiv.org/abs/2411.17116\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.11-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/NVIDIA/Star-Attention\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/NVIDIA/Star-Attention.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\n\u003cb\u003eUltra-Long Sequence Processing (100K+ Tokens)\u003c/b\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eTokenSwift: Lossless Acceleration of Ultra Long Sequence Generation\u003c/b\u003e\u003c/i\u003e, Wu et al., \u003ca href=\"https://arxiv.org/abs/2502.18890\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ICML-2025.02-blue\" alt=\"ICML Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/bigai-nlco/TokenSwift\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/bigai-nlco/TokenSwift.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLongHeads: Multi-Head Attention is Secretly a Long Context Processor\u003c/b\u003e\u003c/i\u003e, Lu et al., \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/EMNLP-2024.11-blue\" alt=\"EMNLP Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/LuLuLuyi/LongHeads\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/LuLuLuyi/LongHeads.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003e∞Bench: Extending Long Context Evaluation Beyond 100K Tokens\u003c/b\u003e\u003c/i\u003e, Bai et al., \u003ca href=\"https://arxiv.org/abs/2412.00359\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ACL-2024.06-blue\" alt=\"ACL Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/OpenBMB/InfiniteBench\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/OpenBMB/InfiniteBench.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\n\u003cb\u003eComprehensive Extension Surveys and Methods\u003c/b\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eBeyond the Limits: A Survey of Techniques to Extend the Context Length in Large Language Models\u003c/b\u003e\u003c/i\u003e, Various, \u003ca href=\"https://arxiv.org/abs/2402.02244\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eA Controlled Study on Long Context Extension and Generalization in LLMs\u003c/b\u003e\u003c/i\u003e, Various, \u003ca href=\"https://arxiv.org/abs/2409.12181\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.09-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/Leooyii/LCEG\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/Leooyii/LCEG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eSelective Attention: Enhancing Transformer through Principled Context Control\u003c/b\u003e\u003c/i\u003e, Various, \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/NeurIPS-2024-blue\" alt=\"NeurIPS Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/umich-sota/selective_attention\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/umich-sota/selective_attention.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cb\u003eVision-Language Models with Sophisticated Context Understanding\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eTowards LLM-Centric Multimodal Fusion: A Survey on Integration Strategies and Techniques\u003c/b\u003e\u003c/i\u003e, An et al., \u003ca href=\"https://arxiv.org/abs/2506.04788\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.01-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eBrowse and Concentrate: Comprehending Multimodal Content via Prior-LLM Context Fusion\u003c/b\u003e\u003c/i\u003e, Wang et al., \u003ca href=\"https://doi.org/10.18653/v1/2024.acl-long.605\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ACL-2024.08-blue\" alt=\"ACL Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/THUNLP-MT/Brote\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/THUNLP-MT/Brote.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eV2PE: Improving Multimodal Long-Context Capability of Vision-Language Models with Variable Visual Position Encoding\u003c/b\u003e\u003c/i\u003e, Dai et al., \u003ca href=\"https://arxiv.org/abs/2412.09616\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.12-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/OpenGVLab/V2PE\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/OpenGVLab/V2PE.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eFlamingo: a Visual Language Model for Few-Shot Learning\u003c/b\u003e\u003c/i\u003e, Alayrac et al., \u003ca href=\"https://arxiv.org/abs/2204.14198\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/NeurIPS-2022.04-blue\" alt=\"NeurIPS Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/lucidrains/flamingo-pytorch\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/lucidrains/flamingo-pytorch.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\n\n\n\u003cb\u003eAudio-Visual Context Integration and Processing\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eAligned Better, Listen Better for Audio-Visual Large Language Models\u003c/b\u003e\u003c/i\u003e, Guo et al., \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ICLR-2025.01-blue\" alt=\"ICLR Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eAVicuna: Audio-Visual LLM with Interleaver and Context-Boundary Alignment for Temporal Referential Dialogue\u003c/b\u003e\u003c/i\u003e, Chen et al., \u003ca href=\"https://arxiv.org/abs/2403.16276\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.03-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eSonicVisionLM: Playing Sound with Vision Language Models\u003c/b\u003e\u003c/i\u003e, Xie et al., \u003ca href=\"https://arxiv.org/abs/2401.04394\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/CVPR-2024.01-blue\" alt=\"CVPR Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/Yusiissy/SonicVisionLM\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/Yusiissy/SonicVisionLM.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eSAVEn-Vid: Synergistic Audio-Visual Integration for Enhanced Understanding in Long Video Context\u003c/b\u003e\u003c/i\u003e, Li et al., \u003ca href=\"https://arxiv.org/abs/2411.16213\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.11-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/LJungang/SAVEn-Vid\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/LJungang/SAVEn-Vid.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\n\n\n\n\u003cb\u003eMulti-Modal Prompt Engineering and Context Design\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eCaMML: Context-Aware Multimodal Learner for Large Models\u003c/b\u003e\u003c/i\u003e, Chen et al., \u003ca href=\"https://arxiv.org/abs/2404.11406\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ACL-2024.08-blue\" alt=\"ACL Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eVisual In-Context Learning for Large Vision-Language Models\u003c/b\u003e\u003c/i\u003e, Zhou et al., \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ACL-2024.08-blue\" alt=\"ACL Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eCAMA: Enhancing Multimodal In-Context Learning with Context-Aware Modulated Attention\u003c/b\u003e\u003c/i\u003e, Li et al., \u003ca href=\"https://arxiv.org/abs/2505.17097\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.05-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003c/ul\u003e\n\n\n\n\n\u003cb\u003eCVPR 2024 Vision-Language Advances\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eCogAgent: A Visual Language Model for GUI Agents\u003c/b\u003e\u003c/i\u003e, Various, \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/CVPR-2024-blue\" alt=\"CVPR Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/THUDM/CogAgent\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/THUDM/CogAgent.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLISA: Reasoning Segmentation via Large Language Model\u003c/b\u003e\u003c/i\u003e, Various, \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/CVPR-2024-blue\" alt=\"CVPR Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/dvlab-research/LISA\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/dvlab-research/LISA.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eReproducible scaling laws for contrastive language-image learning\u003c/b\u003e\u003c/i\u003e, Various, \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/CVPR-2023-blue\" alt=\"CVPR Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/LAION-AI/scaling-laws-openclip\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/LAION-AI/scaling-laws-openclip.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\n\n\n\n\u003cb\u003eVideo and Temporal Understanding\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eVideo Understanding with Large Language Models: A Survey\u003c/b\u003e\u003c/i\u003e, Various, \u003ca href=\"https://arxiv.org/abs/2312.17432\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.12-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/yunlong10/Awesome-LLMs-for-Video-Understanding\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/yunlong10/Awesome-LLMs-for-Video-Understanding.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\n\n\n### Structured Data Integration\n\n\u003cb\u003eKnowledge Graph-Enhanced Language Models\u003c/b\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLearn Together: Joint Multitask Finetuning of Pretrained KG-enhanced LLM for Downstream Tasks\u003c/b\u003e\u003c/i\u003e, Martynova et al., \u003ca href=\"https://doi.org/10.18653/v1/2025.genaik-1.2\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ICCL-2025.01-blue\" alt=\"ICCL Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/Vloods/multitask_finetune\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/Vloods/multitask_finetune.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eKnowledge Graph Tuning: Real-time Large Language Model Personalization based on Human Feedback\u003c/b\u003e\u003c/i\u003e, Sun et al., \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ICLR-2025.02-blue\" alt=\"ICLR Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eKnowledge Graph-Guided Retrieval Augmented Generation\u003c/b\u003e\u003c/i\u003e, Zhu et al., \u003ca href=\"https://arxiv.org/abs/2502.06864\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/nju-websoft/KG2RAG\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/starsnju-websoft/KG2RAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eKGLA: Knowledge Graph Enhanced Language Agents for Customer Service\u003c/b\u003e\u003c/i\u003e, Anonymous et al., \u003ca href=\"https://arxiv.org/abs/2410.19627\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.10-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003c/ul\u003e\n\n\n\u003cb\u003eGraph Neural Networks Combined with Language Models\u003c/b\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eAre Large Language Models In-Context Graph Learners?\u003c/b\u003e\u003c/i\u003e, Li et al., \u003ca href=\"https://arxiv.org/abs/2502.13562\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/yunlong10/Awesome-LLMs-for-Video-Understanding\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/yunlong10/Awesome-LLMs-for-Video-Understanding.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLet's Ask GNN: Empowering Large Language Model for Graph In-Context Learning\u003c/b\u003e\u003c/i\u003e, Hu et al., \u003ca href=\"https://arxiv.org/abs/2410.07074\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/EMNLP-2024.11-blue\" alt=\"EMNLP Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/ppsmk388/AskGNN\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/ppsmk388/AskGNN.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eGL-Fusion: Rethinking the Combination of Graph Neural Network and Large Language model\u003c/b\u003e\u003c/i\u003e, Yang et al., \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ICLR-2025.02-blue\" alt=\"ICLR Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eNT-LLM: A Novel Node Tokenizer for Integrating Graph Structure into Large Language Models\u003c/b\u003e\u003c/i\u003e, Ji et al., \u003ca href=\"https://arxiv.org/abs/2410.10743\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.10-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003c/ul\u003e\n\n\n\u003cb\u003eStructured Data Integration\u003c/b\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eCoddLLM: Empowering Large Language Models for Data Analytics\u003c/b\u003e\u003c/i\u003e, Authors et al., \u003ca href=\"https://arxiv.org/abs/2502.00329\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eStructure-Guided Large Language Models for Text-to-SQL Generation\u003c/b\u003e\u003c/i\u003e, Authors et al., \u003ca href=\"https://arxiv.org/abs/2402.13284\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eStructuredRAG: JSON Response Formatting with Large Language Models\u003c/b\u003e\u003c/i\u003e, Authors et al., \u003ca href=\"https://arxiv.org/abs/2408.11061\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.08-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/weaviate/structured-rag\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/weaviate/structured-rag.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\n\u003cb\u003eFoundational KG-LLM Integration Methods\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eUnifying Large Language Models and Knowledge Graphs: A Roadmap\u003c/b\u003e\u003c/i\u003e, Various, \u003ca href=\"https://arxiv.org/abs/2306.08302\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.06-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/RManLuo/Awesome-LLM-KG?tab=readme-ov-file\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/RManLuo/Awesome-LLM-KG?tab=readme-ov-file.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eCombining Knowledge Graphs and Large Language Models\u003c/b\u003e\u003c/i\u003e, Various, \u003ca href=\"https://arxiv.org/abs/2407.06564\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.07-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eAll Against Some: Efficient Integration of Large Language Models for Message Passing in Graph Neural Networks\u003c/b\u003e\u003c/i\u003e, Various, \u003ca href=\"https://arxiv.org/abs/2407.14996\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.07-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLarge Language Models for Graph Learning\u003c/b\u003e\u003c/i\u003e, Various, \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/WWW-2024-blue\" alt=\"WWW Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003c/ul\u003e\n\n\n### Self-Generated Context\n\n\u003cb\u003eSelf-Supervised Context Generation and Augmentation\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eSelfCite: Self-Supervised Alignment for Context Attribution in Large Language Models\u003c/b\u003e\u003c/i\u003e, Chuang et al., \u003ca href=\"https://arxiv.org/abs/2502.09604\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/facebookresearch/SelfCite\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/facebookresearch/SelfCite.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eSelf-Supervised Prompt Optimization\u003c/b\u003e\u003c/i\u003e, Xiang et al., \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/CoRR-2025.01-orange\" alt=\"CoRR Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/FoundationAgents/MetaGPT/tree/main/examples/spo\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/FoundationAgents/MetaGPT/tree/main/examples/spo.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eSCOPE: A Self-supervised Framework for Improving Faithfulness in Conditional Text Generation\u003c/b\u003e\u003c/i\u003e, Duong et al., \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ICLR-2025.01-blue\" alt=\"ICLR Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/sngdng/scope-faithfulness\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/sngdng/scope-faithfulness.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\n\u003cb\u003eReasoning Models That Generate Their Own Context\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eSelf-Consistency Improves Chain of Thought Reasoning in Language Models\u003c/b\u003e\u003c/i\u003e, Wang et al., \u003ca href=\"https://arxiv.org/abs/2203.11171\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ICLR-2023.02-blue\" alt=\"ICLR Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eTree of Thoughts: Deliberate Problem Solving with Large Language Models\u003c/b\u003e\u003c/i\u003e, Yao et al., \u003ca href=\"https://arxiv.org/abs/2305.10601\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.05-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/princeton-nlp/tree-of-thought-llm\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/princeton-nlp/tree-of-thought-llm.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRethinking Chain-of-Thought from the Perspective of Self-Training\u003c/b\u003e\u003c/i\u003e, Wu et al., \u003ca href=\"https://arxiv.org/abs/2412.10827\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.12-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/zongqianwu/ST-COT\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/zongqianwu/ST-COT.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eAutonomous Tree-search Ability of Large Language Models\u003c/b\u003e\u003c/i\u003e, Authors et al., \u003ca href=\"https://arxiv.org/abs/2310.10686\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.10-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/ZheyuAqaZhang/Autonomous-Tree-search\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/ZheyuAqaZhang/Autonomous-Tree-search.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\n\n\u003cb\u003eIterative Context Refinement and Self-Improvement\u003c/b\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eSelf-Refine: Iterative Refinement with Self-Feedback\u003c/b\u003e\u003c/i\u003e, Madaan et al., \u003ca href=\"https://arxiv.org/abs/2303.17651\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.03-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/madaan/self-refine\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/madaan/self-refine.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eReflect, Retry, Reward: Self-Improving LLMs via Reinforcement Learning\u003c/b\u003e\u003c/i\u003e, Authors et al., \u003ca href=\"https://arxiv.org/abs/2505.24726\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.05-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLarge Language Models Can Self-Improve in Long-context Reasoning\u003c/b\u003e\u003c/i\u003e, Li et al., \u003ca href=\"https://arxiv.org/abs/2411.08147\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.11-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/SihengLi99/SEALONG\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/SihengLi99/SEALONG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eCode Generation with AlphaCodium: From Prompt Engineering to Flow Engineering\u003c/b\u003e\u003c/i\u003e, Oren et al., \u003ca href=\"https://arxiv.org/abs/2401.08500\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.04-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e \u003ca href=\"https://github.com/Codium-ai/alphacodium\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/github/stars/Codium-ai/alphacodium.svg?style=social\" alt=\"GitHub stars\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLanguage Agent Tree Search Unifies Reasoning Acting and Planning in Language Models\u003c/b\u003e\u003c/i\u003e, Zhou et al., \u003ca href=\"https://arxiv.org/abs/2310.04406\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.10-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e \u003ca href=\"https://github.com/andyz245/Language-Agent-Tree-Search\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/github/stars/andyz245/Language-Agent-Tree-Search.svg?style=social\" alt=\"GitHub stars\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003c/ul\u003e\n\n\n\u003cb\u003eMeta-Learning and Autonomous Context Evolution\u003c/b\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMeta-in-context learning in large language models\u003c/b\u003e\u003c/i\u003e, Coda-Forno et al., \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/NeurIPS-2023.12-blue\" alt=\"NeurIPS Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eEvoPrompt: Connecting LLMs with Evolutionary Algorithms Yields Powerful Prompt Optimizers\u003c/b\u003e\u003c/i\u003e, Guo et al., \u003ca href=\"https://arxiv.org/abs/2309.08532\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/ICLR-2024.01-blue\" alt=\"ICLR Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/beeevita/EvoPrompt\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/beeevita/EvoPrompt.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eAutoPDL: Automatic Prompt Optimization for LLM Agents\u003c/b\u003e\u003c/i\u003e, Spiess et al., \u003ca href=\"https://arxiv.org/abs/2504.04365\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/AutoML-2025.04-orange\" alt=\"AutoML Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eAgent-Pro: Learning to Evolve Coder Agents via Proposal-based Programming\u003c/b\u003e\u003c/i\u003e, Zhang et al., \u003ca href=\"https://arxiv.org/abs/2402.17574\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.05-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003c/ul\u003e\n\n\n\u003cb\u003eFoundational Chain-of-Thought Research\u003c/b\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eChain-of-thought prompting elicits reasoning in large language models\u003c/b\u003e\u003c/i\u003e, Wei et al., \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/NeurIPS-2022-blue\" alt=\"NeurIPS Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003c/ul\u003e\n\n\n---\n\n## 🛠️ Implementation and Challenges\n\n### 1. Retrieval-Augmented Generation (RAG)\n\n\u003cb\u003esurvey\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRetrieval-Augmented Generation for Large Language Models: A Survey\u003c/b\u003e\u003c/i\u003e, Yunfan Gao et al., \u003ca href=\"https://arxiv.org/abs/2312.10997\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.12-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/Tongji-KGLLM/RAG-Survey\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/Tongji-KGLLM/RAG-Survey.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eA Survey of Graph Retrieval-Augmented Generation for Customized Large Language Models\u003c/b\u003e\u003c/i\u003e, Siyun Zhao et al., \u003ca href=\"https://arxiv.org/abs/2501.13958\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.01-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/DEEP-PolyU/Awesome-GraphRAG\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/DEEP-PolyU/Awesome-GraphRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRetrieval Augmented Generation (RAG) and Beyond: A Comprehensive Survey on How to Make your LLMs use External Data More Wisely\u003c/b\u003e\u003c/i\u003e, Siyun Zhao et al., \u003ca href=\"https://arxiv.org/abs/2409.14924\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.09-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eEvaluation of Retrieval-Augmented Generation: A Survey\u003c/b\u003e\u003c/i\u003e, Hao Yu et al., \u003ca href=\"https://arxiv.org/abs/2405.07437\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.07-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/YHPeter/Awesome-RAG-Evaluation\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/YHPeter/Awesome-RAG-Evaluation.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRetrieval-Augmented Generation for Knowledge-Intensive NLP Tasks\u003c/b\u003e\u003c/i\u003e, Lewis et al., \u003ca href=\"https://arxiv.org/abs/2005.11401\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2020.05-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/costadev00/RAG-paper-implementation-from-scratch\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/costadev00/RAG-paper-implementation-from-scratch.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eA Survey on Knowledge-Oriented Retrieval-Augmented Generation\u003c/b\u003e\u003c/i\u003e, Cheng et al., \u003ca href=\"https://arxiv.org/abs/2503.10677\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/USTCAGI/Awesome-Papers-Retrieval-Augmented-Generation\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/USTCAGI/Awesome-Papers-Retrieval-Augmented-Generation.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eA Survey on RAG Meeting LLMs: Towards Retrieval-Augmented Large Language Models\u003c/b\u003e\u003c/i\u003e, Ding et al., \u003ca href=\"https://arxiv.org/abs/2405.06211\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.06-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003c/ul\u003e\n\n\n\n\u003cb\u003eNaive RAG\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eBeyond the Limits: A Survey of Techniques to Extend the Context Length in Large Language Models\u003c/b\u003e\u003c/i\u003e, Xindi Wang et al., \u003ca href=\"https://arxiv.org/abs/2402.02244\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eIn-context Examples Selection for Machine Translation\u003c/b\u003e\u003c/i\u003e, Sweta Agrawal et al., \u003ca href=\"https://arxiv.org/abs/2212.02437\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2022.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eIn Defense of RAG in the Era of Long-Context Language Models\u003c/b\u003e\u003c/i\u003e, Tan Yu et al., \u003ca href=\"https://arxiv.org/abs/2409.01666\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.09-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRetrieval-Augmented Generation for Knowledge-Intensive NLP Tasks\u003c/b\u003e\u003c/i\u003e, Patrick Lewis et al., \u003ca href=\"https://arxiv.org/abs/2005.11401\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2020.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLightRAG: Simple and Fast Retrieval-Augmented Generation\u003c/b\u003e\u003c/i\u003e, Zirui Guo et al., \u003ca href=\"https://arxiv.org/abs/2410.05779\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://anonymous.4open.science/r/LightRAG-2BEE\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/anonymous/LightRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eGenerate rather than Retrieve: Large Language Models are Strong Context Generators\u003c/b\u003e\u003c/i\u003e, Wenhao Yu et al., \u003ca href=\"https://arxiv.org/abs/2209.10063\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2022.09-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/wyu97/GenRead\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/wyu97/GenRead.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLarge language models can be easily distracted by irrelevant context\u003c/b\u003e\u003c/i\u003e, Freda Shi et al., \u003ca href=\"https://arxiv.org/abs/2302.00093\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/google-research-datasets/GSM-IC\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/google-research-datasets/GSM-IC.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eOld IR Methods Meet RAG\u003c/b\u003e\u003c/i\u003e, Oz Huly et al.\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eDense Passage Retrieval for Open-Domain Question Answering\u003c/b\u003e\u003c/i\u003e, Vladimir Karpukhin et al., \u003ca href=\"https://arxiv.org/abs/2004.04906\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2020.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/facebookresearch/DPR\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/facebookresearch/DPR.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\n\n\n\u003cb\u003eAdvanced RAG\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eAdaptive-RAG: Learning to Adapt Retrieval-Augmented Large Language Models through Question Complexity\u003c/b\u003e\u003c/i\u003e, Soyeong Jeong et al., \u003ca href=\"https://arxiv.org/abs/2403.14403\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/starsuzi/Adaptive-RAG\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/starsuzi/Adaptive-RAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eImproving language models by retrieving from trillions of tokens\u003c/b\u003e\u003c/i\u003e, Sebastian Borgeaud et al., \u003ca href=\"https://arxiv.org/abs/2112.04426\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2022.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eFoRAG: Factuality-optimized Retrieval Augmented Generation for Web-enhanced Long-form Question Answering\u003c/b\u003e\u003c/i\u003e, Tianchi Cai et al.\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eIM-RAG: Multi-Round Retrieval-Augmented Generation Through Learning Inner Monologues\u003c/b\u003e\u003c/i\u003e, Diji Yang et al., \u003ca href=\"https://arxiv.org/abs/2405.13021\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRAGCache: Efficient Knowledge Caching for Retrieval-Augmented Generation\u003c/b\u003e\u003c/i\u003e, Chao Jin et al., \u003ca href=\"https://arxiv.org/abs/2404.12457\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eCorrective Retrieval Augmented Generation\u003c/b\u003e\u003c/i\u003e, Shi-Qi Yan et al., \u003ca href=\"https://arxiv.org/abs/2401.15884\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/HuskyInSalt/CRAG\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/HuskyInSalt/CRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRankRAG: Unifying Context Ranking with Retrieval-Augmented Generation in LLMs\u003c/b\u003e\u003c/i\u003e, Yue Yu et al., \u003ca href=\"https://arxiv.org/abs/2407.02485\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eAstute RAG: Overcoming Imperfect Retrieval Augmentation and Knowledge Conflicts for Large Language Models\u003c/b\u003e\u003c/i\u003e, Fei Wang et al., \u003ca href=\"https://arxiv.org/abs/2410.07176\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLearning to Filter Context for Retrieval-Augmented Generation\u003c/b\u003e\u003c/i\u003e, Zhiruo Wang et al., \u003ca href=\"https://arxiv.org/abs/2311.08377\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/zorazrw/filco\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/zorazrw/filco.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eQuery Rewriting in Retrieval-Augmented Large Language Models\u003c/b\u003e\u003c/i\u003e, Xinbei Ma et al., \u003ca href=\"https://arxiv.org/abs/2305.14283\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/qijimrc/ROBUST\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/qijimrc/ROBUST.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eUPRISE: Universal Prompt Retrieval for Improving Zero-Shot Evaluation\u003c/b\u003e\u003c/i\u003e, Daixuan Cheng et al., \u003ca href=\"https://arxiv.org/abs/2303.08518\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/MatthewKKai/SMRC2\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/MatthewKKai/SMRC2.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLongllmlingua: Accelerating and enhancing llms in long context scenarios via prompt compression\u003c/b\u003e\u003c/i\u003e, Huiqiang Jiang et al., \u003ca href=\"https://arxiv.org/abs/2310.06839\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/microsoft/LLMLingua\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/microsoft/LLMLingua.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eDocument-level event argument extraction by conditional generation\u003c/b\u003e\u003c/i\u003e, Sha Li et al., \u003ca href=\"https://arxiv.org/abs/2104.05919\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2021.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/raspberryice/gen-arg\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/raspberryice/gen-arg.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMulti-sentence Argument Linking\u003c/b\u003e\u003c/i\u003e, Seth Ebner et al., \u003ca href=\"https://arxiv.org/abs/1911.03766\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2019.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://nlp.jhu.edu/rams/\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/nlp-jhu/RAMS.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eFine-tuning or retrieval? comparing knowledge injection in llms\u003c/b\u003e\u003c/i\u003e, Oded Ovadia et al., \u003ca href=\"https://arxiv.org/abs/2312.05934\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eIAG: Induction-Augmented Generation Framework for Answering Reasoning Questions\u003c/b\u003e\u003c/i\u003e, Zhebin Zhang et al., \u003ca href=\"https://arxiv.org/abs/2311.18397\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRetrieval Meets Long Context Large Language Models\u003c/b\u003e\u003c/i\u003e, Peng Xu et al., \u003ca href=\"https://arxiv.org/abs/2310.03025\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eDense x retrieval: What retrieval granularity should we use?\u003c/b\u003e\u003c/i\u003e, Tong Chen et al., \u003ca href=\"https://arxiv.org/abs/2312.06648\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/ct123098/factoid-wiki\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/ct123098/factoid-wiki.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eInvestigating the Factual Knowledge Boundary of Large Language Models with Retrieval Augmentation\u003c/b\u003e\u003c/i\u003e, Ruiyang Ren et al., \u003ca href=\"https://arxiv.org/abs/2307.11019\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/RUCAIBox/LLM-Knowledge-Boundary\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/RUCAIBox/LLM-Knowledge-Boundary.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eThe Power of Noise: Redefining Retrieval for RAG Systems\u003c/b\u003e\u003c/i\u003e, Florin Cuconasu et al., \u003ca href=\"https://arxiv.org/abs/2401.14887\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/florin-git/The-Power-of-Noise\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/florin-git/The-Power-of-Noise.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRECITATION-AUGMENTED LANGUAGE MODELS\u003c/b\u003e\u003c/i\u003e, Zhiqing Sun et al., \u003ca href=\"https://arxiv.org/abs/2210.01296\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2022.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/Edward-Sun/RECITE\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/Edward-Sun/RECITE.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRobust Retrieval Augmented Generation for Zero-shot Slot Filling\u003c/b\u003e\u003c/i\u003e, Michael Glass et al., \u003ca href=\"https://arxiv.org/abs/2108.13934\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2021.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/IBM/kgi-slot-filling\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/IBM/kgi-slot-filling.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eIn-Context Retrieval-Augmented Language Models\u003c/b\u003e\u003c/i\u003e, Ori Ram et al., \u003ca href=\"https://arxiv.org/abs/2302.00083\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/AI21Labs/in-context-ralm\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/AI21Labs/in-context-ralm.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLearning to Retrieve In-Context Examples for Large Language Models\u003c/b\u003e\u003c/i\u003e, Liang Wang et al., \u003ca href=\"https://arxiv.org/abs/2307.07164\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.03-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/microsoft/LMOps/tree/main/llm_retriever\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/microsoft/LMOps.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\n\n\n\u003cb\u003eModular RAG\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eFlashRAG: A Modular Toolkit for Efficient Retrieval-Augmented Generation Research\u003c/b\u003e\u003c/i\u003e, Jiajie Jin et al., \u003ca href=\"https://arxiv.org/abs/2405.13576\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/RUC-NLPIR/FlashRAG\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/RUC-NLPIR/FlashRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMulti-Head RAG: Solving Multi-Aspect Problems with LLMs\u003c/b\u003e\u003c/i\u003e, Maciej Besta et al., \u003ca href=\"https://arxiv.org/abs/2406.05085\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/spcl/MRAG\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/spcl/MRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eStructRAG: Boosting Knowledge Intensive Reasoning of LLMs via Inference-time Hybrid Information Structurization\u003c/b\u003e\u003c/i\u003e, Zhuoqun Li et al., \u003ca href=\"https://arxiv.org/abs/2410.08815\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/Li-Z-Q/StructRAG\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/Li-Z-Q/StructRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRAFT: Adapting Language Model to Domain Specific RAG\u003c/b\u003e\u003c/i\u003e, Tianjun Zhang et al., \u003ca href=\"https://arxiv.org/abs/2403.10131\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/ShishirPatil/gorilla\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/ShishirPatil/gorilla.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRetrieval-Generation Alignment for End-to-End Task-Oriented Dialogue System\u003c/b\u003e\u003c/i\u003e, Weizhou Shen et al., \u003ca href=\"https://arxiv.org/abs/2310.08877\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/shenwzh3/MK-TOD\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/shenwzh3/MK-TOD.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eUniMS-RAG: A Unified Multi-source Retrieval-Augmented Generation for Personalized Dialogue Systems\u003c/b\u003e\u003c/i\u003e, Hongru Wang et al., \u003ca href=\"https://arxiv.org/abs/2401.13256\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRetrieve-and-Sample: Document-level Event Argument Extraction via Hybrid Retrieval Augmentation\u003c/b\u003e\u003c/i\u003e, Yubing Ren et al.\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRA-DIT: RETRIEVAL-AUGMENTED DUAL INSTRUCTION TUNING\u003c/b\u003e\u003c/i\u003e, Xi Victoria Lin et al., \u003ca href=\"https://arxiv.org/abs/2310.01352\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/facebookresearch/RA-DIT\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/facebookresearch/RA-DIT.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eSelf-Knowledge Guided Retrieval Augmentation for Large Language Models\u003c/b\u003e\u003c/i\u003e, Yile Wang et al., \u003ca href=\"https://arxiv.org/abs/2310.05002\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/THUNLP-MT/SKR\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/THUNLP-MT/SKR.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003ePrompt-Guided Retrieval Augmentation for Non-Knowledge-Intensive Tasks\u003c/b\u003e\u003c/i\u003e, Zhicheng Guo et al., \u003ca href=\"https://arxiv.org/abs/2305.17653\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/THUNLP-MT/PGRA\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/THUNLP-MT/PGRA.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eREPLUG: Retrieval-Augmented Black-Box Language Models\u003c/b\u003e\u003c/i\u003e, Weijia Shi et al., \u003ca href=\"https://arxiv.org/abs/2301.12652\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eQuery Rewriting for Retrieval-Augmented Large Language Models\u003c/b\u003e\u003c/i\u003e, Xinbei Ma et al., \u003ca href=\"https://doi.org/10.18653/v1/2023.emnlp-main.323\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/EMNLP-2023.00-blue\" alt=\"DOI Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/xbmxb/RAG-query-rewriting\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/xbmxb/RAG-query-rewriting.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLift Yourself Up: Retrieval-augmented Text Generation with Self-Memory\u003c/b\u003e\u003c/i\u003e, Xin Cheng et al., \u003ca href=\"https://arxiv.org/abs/2305.02437\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/Hannibal046/SelfMemory\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/Hannibal046/SelfMemory.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eImproving the Domain Adaptation of Retrieval Augmented Generation (RAG) Models for Open Domain Question Answering\u003c/b\u003e\u003c/i\u003e, Shamane Siriwardhana et al., \u003ca href=\"https://arxiv.org/abs/2210.02627\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\n\n\n\u003cb\u003eGraph-Based RAG\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eDon't Forget to Connect! Improving RAG with Graph-based Reranking\u003c/b\u003e\u003c/i\u003e, Jialin Dong et al., \u003ca href=\"https://arxiv.org/abs/2405.18414\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.05-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eFrom Local to Global: A Graph RAG Approach to Query-Focused Summarization\u003c/b\u003e\u003c/i\u003e, Darren Edge et al., \u003ca href=\"https://arxiv.org/abs/2404.16130\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eGRAG: Graph Retrieval-Augmented Generation\u003c/b\u003e\u003c/i\u003e, Yuntong Hu et al., \u003ca href=\"https://arxiv.org/abs/2405.16506\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/HuieL/GRAG\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/HuieL/GRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eIseeq: Information seeking question generation using dynamic meta-information retrieval and knowledge graphs\u003c/b\u003e\u003c/i\u003e, Manas Gaur et al., \u003ca href=\"https://arxiv.org/abs/2112.07622\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2022.06-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/manasgaur/AAAI-22\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/manasgaur/AAAI-22.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eG-retriever: Retrieval-augmented generation for textual graph understanding and question answering\u003c/b\u003e\u003c/i\u003e, Xiaoxin He et al., \u003ca href=\"https://arxiv.org/abs/2402.07630\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/XiaoxinHe/G-Retriever\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/XiaoxinHe/G-Retriever.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eKnowledge graph prompting for multi-document question answering\u003c/b\u003e\u003c/i\u003e, Yu Wang et al., \u003ca href=\"https://arxiv.org/abs/2402.08774\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/YuWVandy/KG-LLM-MDQA\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/YuWVandy/KG-LLM-MDQA.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eGNN-RAG: Graph Neural Retrieval for Large Language Model Reasoning\u003c/b\u003e\u003c/i\u003e, Costas Mavromatis et al., \u003ca href=\"https://arxiv.org/abs/2405.20139\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.05-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/cmavro/GNN-RAG\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/cmavro/GNN-RAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLightPROF: A Lightweight Reasoning Framework for Large Language Model on Knowledge Graph\u003c/b\u003e\u003c/i\u003e\n    \u003ca href=\"https://github.com/tsinghua-fib-lab/ACL24-EconAgent\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/tsinghua-fib-lab/ACL24-EconAgent.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eSimple Is Effective: The Roles of Graphs and Large Language Models in Knowledge-Graph-Based Retrieval-Augmented Generation\u003c/b\u003e\u003c/i\u003e\n    \u003ca href=\"https://github.com/Graph-COM/SubgraphRAG\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/Graph-COM/SubgraphRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eKnowledge Graph-Guided Retrieval Augmented Generation\u003c/b\u003e\u003c/i\u003e\n    \u003ca href=\"https://github.com/nju-websoft/KG2RAG\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/nju-websoft/KG2RAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMedRAG: Enhancing Retrieval-augmented Generation with Knowledge Graph-Elicited Reasoning for Healthcare Copilot\u003c/b\u003e\u003c/i\u003e\n    \u003ca href=\"https://github.com/SNOWTEAM2023/MedRAG\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/SNOWTEAM2023/MedRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMitigating Large Language Model Hallucinations via Autonomous Knowledge Graph-based Retrofitting\u003c/b\u003e\u003c/i\u003e, KGR et al., \u003ca href=\"https://arxiv.org/abs/2311.13314\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.11-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/mansicer/MAIC\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/mansicer/MAIC.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eIn-depth Analysis of Graph-based RAG in a Unified Framework\u003c/b\u003e\u003c/i\u003e\u003ca href=\"https://arxiv.org/abs/2503.04338\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.05-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/JayLZhou/GraphRAG\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/JayLZhou/GraphRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRAPTOR: Recursive Abstractive Processing for Tree-Organized Retrieval\u003c/b\u003e\u003c/i\u003e, Parth Sarthi et al., \u003ca href=\"https://arxiv.org/abs/2401.18059\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/parthsarthi03/raptor\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/parthsarthi03/raptor.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eTableRAG: Million-Token Table Understanding with Language Models\u003c/b\u003e\u003c/i\u003e, Si-An Chen et al., \u003ca href=\"https://arxiv.org/abs/2410.04739\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/google-research/google-research/tree/master/table_rag\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/google-research/google-research.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eKAG: Boosting LLMs in Professional Domains via Knowledge Augmented Generation\u003c/b\u003e\u003c/i\u003e, Lei Liang et al., \u003ca href=\"https://arxiv.org/abs/2409.13731\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/OpenSPG/KAG\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/OpenSPG/KAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eGFM-RAG: Graph Foundation Model for Retrieval Augmented Generation\u003c/b\u003e\u003c/i\u003e, Luo et al., \u003ca href=\"https://arxiv.org/abs/2502.01113\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.02-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/RManLuo/gfm-rag\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/RManLuo/gfm-rag.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eHybridRAG: A Hybrid Retrieval System for RAG Combining Vector and Graph Search\u003c/b\u003e\u003c/i\u003e, Sarabesh, \u003ca href=\"#\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/GitHub-2024.12-white\" alt=\"GitHub Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/sarabesh/HybridRAG\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/sarabesh/HybridRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\n\n\n\u003cb\u003eAgentic RAG\u003c/b\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eFrom RAG to Memory: Non-Parametric Continual Learning for Large Language Models\u003c/b\u003e\u003c/i\u003e, Bernal Jiménez Gutiérrez et al., \u003ca href=\"https://arxiv.org/abs/2502.14802\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/OSU-NLP-Group/HippoRAG\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/OSU-NLP-Group/HippoRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eHippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language Models\u003c/b\u003e\u003c/i\u003e, Bernal Jiménez Gutiérrez et al., \u003ca href=\"https://arxiv.org/abs/2405.14924\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/OSU-NLP-Group/HippoRAG\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/OSU-NLP-Group/HippoRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eGraphReader: Building Graph-based Agent to Enhance Long-Context Abilities of Large Language Models\u003c/b\u003e\u003c/i\u003e, Shilong Li et al., \u003ca href=\"https://arxiv.org/abs/2406.14550\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003ePlanRAG: A Plan-then-Retrieval Augmented Generation for Generative Large Language Models as Decision Makers\u003c/b\u003e\u003c/i\u003e, Myeonghwa Lee et al., \u003ca href=\"https://arxiv.org/abs/2406.12430\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/myeon9h/PlanRAG\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/myeon9h/PlanRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eSelf-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection\u003c/b\u003e\u003c/i\u003e, Akari Asai et al., \u003ca href=\"https://arxiv.org/abs/2402.08353\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/AkariAsai/self-rag\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/AkariAsai/self-rag.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eDeepRAG: Thinking to Retrieve Step by Step for Large Language Models\u003c/b\u003e\u003c/i\u003e, Xinyan Guan et al., \u003ca href=\"https://arxiv.org/abs/2502.01142\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003ePaperqa: Retrieval-augmented generative agent for scientific research\u003c/b\u003e\u003c/i\u003e, Jakub Lála et al., \u003ca href=\"https://arxiv.org/abs/2312.07559\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eLarge Language Models as Source Planner for Personalized Knowledge-grounded Dialogues\u003c/b\u003e\u003c/i\u003e, Hongru Wang et al., \u003ca href=\"https://arxiv.org/abs/2308.06181\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/hrwise-nlp/SAFARI\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/hrwise-nlp/SAFARI.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003ePRCA: Fitting Black-Box Large Language Models for Retrieval Question Answering via Pluggable Reward-Driven Contextual Adapter\u003c/b\u003e\u003c/i\u003e, Haoyan Yang et al., \u003ca href=\"https://arxiv.org/abs/2310.18347\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/xbmxb/RAG-query-rewriting\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/xbmxb/RAG-query-rewriting.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eSELF-RAG: LEARNING TO RETRIEVE, GENERATE, AND CRITIQUE THROUGH SELF-REFLECTION\u003c/b\u003e\u003c/i\u003e, Akari Asai et al., \u003ca href=\"https://arxiv.org/abs/2310.11511\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://selfrag.github.io/\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/selfrag/selfrag.github.io.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eRAT: Retrieval Augmented Thoughts Elicit Context-Aware Reasoning in Long-Horizon Generation\u003c/b\u003e\u003c/i\u003e, Zihao Wang et al., \u003ca href=\"https://arxiv.org/abs/2403.05313\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/CraftJarvis/RAT\" target=\"_blank\"\u003e\n        \u003cimg src=\"https://img.shields.io/github/stars/CraftJarvis/RAT.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eChain-of-verification reduces hallucination in large language models\u003c/b\u003e\u003c/i\u003e, Shehzaad Dhuliawala et al., \u003ca href=\"https://arxiv.org/abs/2309.11495\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.00-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eHM-RAG: Hierarchical Multi-Agent Multimodal Retrieval Augmented Generation\u003c/b\u003e\u003c/i\u003e, Liu et al., \u003ca href=\"https://arxiv.org/abs/2504.12330\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.04-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/ocean-luna/HMRAG\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/ocean-luna/HMRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries\u003c/b\u003e\u003c/i\u003e, Tang \u0026 Yang, \u003ca href=\"https://arxiv.org/abs/2401.15391\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.01-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/yixuantt/MultiHop-RAG\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/yixuantt/MultiHop-RAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMMOA-RAG: Improving Retrieval-Augmented Generation through Multi-Agent Reinforcement Learning\u003c/b\u003e\u003c/i\u003e, Chen et al., \u003ca href=\"https://arxiv.org/abs/2010.10110\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2021.01-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/chenyiqun/MMOA-RAG\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/chenyiqun/MMOA-RAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eSearch-in-the-Chain: Towards Accurate, Credible, and Up-to-Date Large Language Models\u003c/b\u003e\u003c/i\u003e, Menick et al., \u003ca href=\"https://arxiv.org/abs/2304.14732\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.04-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003c/ul\u003e\n\n\n\n\u003cb\u003eReal-Time and Streaming RAG\u003c/b\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eStreamingRAG: Real-time Contextual Retrieval and Generation Framework\u003c/b\u003e\u003c/i\u003e, Sankaradas et al., \u003ca href=\"https://arxiv.org/abs/2501.14101\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2024.01-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/video-db/StreamRAG\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shields.io/github/stars/video-db/StreamRAG.svg?style=social\" alt=\"GitHub stars\"\u003e\n    \u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMulti-task Retriever Fine-tuning for Domain-Specific and Efficient RAG\u003c/b\u003e\u003c/i\u003e, Authors, \u003ca href=\"https://arxiv.org/abs/2501.04652\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2025.01-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003c/li\u003e\n\u003c/ul\u003e\n\n\n### 2. Memory Systems\n\n\u003cb\u003ePersistent Memory Architecture\u003c/b\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ci\u003e\u003cb\u003eMemGPT: Towards LLMs as Operating Systems\u003c/b\u003e\u003c/i\u003e, Packer et al., \u003ca href=\"https://arxiv.org/abs/2310.08560\" target=\"_blank\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2023.10-red\" alt=\"arXiv Badge\"\u003e\u003c/a\u003e\n    \u003ca href=\"https://github.com/letta-ai/letta\" target=\"_blank\"\u003e\n  \t\t\u003cimg src=\"https://img.shiel","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FMeirtz%2FAwesome-Context-Engineering","html_url":"https://awesome.ecosyste.ms/projects/github.com%2FMeirtz%2FAwesome-Context-Engineering","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FMeirtz%2FAwesome-Context-Engineering/lists"}