Elamur Research
☰
home / papers / 2024 / LightRAG: Simple and Fast Retrieval-Augmented Generation

LightRAG: Simple and Fast Retrieval-Augmented Generation

Written for the Research library (claude-opus-5-5) from the paper text, in our own words: the license "CC BY-NC-SA 4.0 (arXiv)" does not allow redistributing the text itself. Original — arxiv.org/pdf/2410.05779

Summary

Standard RAG pipelines index text as isolated chunks. When a question involves several linked concepts, the retrieved pieces often produce a disjointed answer. LightRAG (Guo et al., BUPT and University of Hong Kong) addresses this by building a knowledge graph at indexing time. Documents are split into chunks, an LLM extracts entities and their relations, writes a short key–value profile for every node and edge, and duplicates are merged. At query time, an LLM derives two kinds of keywords from the question. Specific ones are matched against entities, and broad thematic ones against relations. Graph neighbourhoods and vector search together assemble the context. New documents are merged incrementally instead of rebuilding the index. An LLM judge compares answers pairwise on comprehensiveness, diversity and empowerment. The authors report clear gains in answer quality and retrieval efficiency over existing RAG baselines. LightRAG matters as a lighter, update-friendly form of graph-based RAG, and its code is open source.

Key points

  • Indexing has three LLM-driven steps: entity/relation recognition, profiling into key–value pairs, and deduplication. Entities are keyed by name only, while relations can receive several theme-based keys.
  • Dual-level retrieval separates low-level keys (concrete entities and details) from high-level keys (themes). The keyword prompt returns JSON with high_level_keywords and low_level_keywords fields.
  • An incremental update algorithm adds new data without full re-indexing, which lowers compute cost in changing corpora.
  • Evaluation uses LLM-generated questions, with five high-level questions per (user, task) pair. A JSON-structured judge picks a winner per criterion plus an overall winner. A case study (Table 5) shows LightRAG beating NaiveRAG on all criteria.
  • The code is available at github.com/HKUDS/LightRAG (arXiv:2410.05779, v3 April 2025). The provided excerpt omits the experiments section, so specific benchmark figures are not included in this summary.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Other

Cited by

122 papers of our library cite it (of the 140673 whose bibliographies we read).

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.2410.05779 · arXiv 2410.05779

@misc{guo2024lightrag,
  title = {{LightRAG: Simple and Fast Retrieval-Augmented Generation}},
  author = {Guo, Zirui and Xia, Lianghao and Yu, Yanhua and Ao, Tu and Huang, Chao},
  year = {2024},
  doi = {10.48550/arxiv.2410.05779},
  eprint = {2410.05779},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2410.05779}
}

References

21 references, 4 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. William Brannon, Suyash Fulay, Hang Jiang, Wonjune Kang, Brandon Roy, Jad Kabbara, and Deb Roy. Congrat: Self-supervised contrastive pretraining for joint graph and text embeddings. arXiv preprint arXiv:2305.14321, 2023. find
  2. Chi-Min Chan, Chunpu Xu, Ruibin Yuan, Hongyin Luo, Wei Xue, Yike Guo, and Jie Fu. Rq-rag: Learning to refine queries for retrieval augmented generation. arXiv preprint arXiv:2404.00610, 2024. find
  3. Runjin Chen, Tong Zhao, AJAY KUMAR JAISWAL, Neil Shah, and Zhangyang Wang. Llaga: Large language and graph assistant. In International Conference on Machine Learning (ICML), 2024. find
  4. Darren Edge, Ha Trinh, Newman Cheng, Joshua Bradley, Alex Chao, Apurva Mody, Steven Truitt, and Jonathan Larson. From local to global: A graph rag approach to query-focused summarization. arXiv preprint arXiv:2404.16130, 2024. find
  5. Shahul Es, Jithin James, Luis Espinosa Anke, and Steven Schockaert. Ragas: Automated evaluation of retrieval augmented generation. In International Conference of the European Chapter of the Association for Computational Linguistics (EACL), pp. 150–158, 2024. find our card
  6. Wenqi Fan, Yujuan Ding, Liangbo Ning, Shijie Wang, Hengyun Li, Dawei Yin, Tat-Seng Chua, and Qing Li. A survey on rag meeting llms: Towards retrieval-augmented large language models. In International Conference on Knowledge Discovery and Data Mining (KDD), pp. 6491–6501, 2024. find our card
  7. Luyu Gao, Xueguang Ma, Jimmy Lin, and Jamie Callan. Precise zero-shot dense retrieval without relevance labels. arXiv preprint arXiv:2212.10496, 2022. find our card
  8. Yunfan Gao, Yun Xiong, Xinyu Gao, Kangxiang Jia, Jinliu Pan, Yuxi Bi, Yi Dai, Jiawei Sun, and Haofen Wang. Retrieval-augmented generation for large language models: A survey. arXiv preprint arXiv:2312.10997, 2023. find our card
  9. Yichuan Li, Kaize Ding, and Kyumin Lee. Grenade: Graph-centric language model for self-supervised representation learning on text-attributed graphs. In International Conference on Empirical Methods in Natural Language Processing (EMNLP), pp. 2745–2757, 2023. find
  10. Hao Liu, Jiarui Feng, Lecheng Kong, Ningyue Liang, Dacheng Tao, Yixin Chen, and Muhan Zhang. One for all: Towards training one graph model for all classification tasks. In International Conference on Learning Representations (ICLR), 2024. find
  11. Yuanjie Lyu, Zhiyu Li, Simin Niu, Feiyu Xiong, Bo Tang, Wenjin Wang, Hao Wu, Huanyong Liu, Tong Xu, and Enhong Chen. Crud-rag: A comprehensive chinese benchmark for retrieval-augmented generation of large language models. arXiv preprint arXiv:2401.17043, 2024. — [arXiv:2401.17043](https://arxiv.org/abs/2401.17043) find
  12. Hongjin Qian, Peitian Zhang, Zheng Liu, Kelong Mao, and Zhicheng Dou. Memorag: Moving towards next-gen rag via memory-inspired knowledge discovery, 2024. URL https://arxiv. org/abs/2409.05591. — [arXiv:2409.05591](https://arxiv.org/abs/2409.05591) find
  13. Ori Ram, Yoav Levine, Itay Dalmedigos, Dor Muhlgay, Amnon Shashua, Kevin Leyton-Brown, and Yoav Shoham. In-context retrieval-augmented language models. Transactions of the Association for Computational Linguistics (TACL), 11:1316–1331, 2023. find
  14. Ladislav Rampášek, Michael Galkin, Vijay Prakash Dwivedi, Anh Tuan Luu, Guy Wolf, and Dominique Beaini. Recipe for a general, powerful, scalable graph transformer. International Conference on Neural Information Processing Systems (NeurIPS), 35:14501–14515, 2022. find
  15. Alireza Salemi and Hamed Zamani. Evaluating retrieval quality in retrieval-augmented generation. In ACM International Conference on Research and Development in Information Retrieval (SIGIR), pp. 2395–2400, 2024. find
  16. Viju Sudhi, Sinchana Ramakanth Bhat, Max Rudat, and Roman Teucher. Rag-ex: A generic framework for explaining retrieval augmented generation. In ACM International Conference on Research and Development in Information Retrieval (SIGIR), pp. 2776–2780, 2024. find
  17. Jiabin Tang, Yuhao Yang, Wei Wei, Lei Shi, Lixin Su, Suqi Cheng, Dawei Yin, and Chao Huang. Graphgpt: Graph instruction tuning for large language models. In ACM International Conference on Research and Development in Information Retrieval (SIGIR), pp. 491–500, 2024. find
  18. Shangqing Tu, Yuanchun Wang, Jifan Yu, Yuyang Xie, Yaran Shi, Xiaozhi Wang, Jing Zhang, Lei Hou, and Juanzi Li. R-eval: A unified toolkit for evaluating domain knowledge of retrieval augmented large language models. In International Conference on Knowledge Discovery and Data Mining (KDD), pp. 5813–5… find
  19. Han Xie, Da Zheng, Jun Ma, Houyu Zhang, Vassilis N Ioannidis, Xiang Song, Qing Ping, Sheng Wang, Carl Yang, Yi Xu, et al. Graph-aware language model pre-training on a large graph corpus can help multiple graph applications. In International Conference on Knowledge Discovery and Data Mining (KDD), pp… find
  20. Yue Yu, Wei Ping, Zihan Liu, Boxin Wang, Jiaxuan You, Chao Zhang, Mohammad Shoeybi, and Bryan Catanzaro. Rankrag: Unifying context ranking with retrieval-augmented generation in llms. arXiv preprint arXiv:2407.02485, 2024. — [arXiv:2407.02485](https://arxiv.org/abs/2407.02485) find
  21. Penghao Zhao, Hailin Zhang, Qinhan Yu, Zhengren Wang, Yunteng Geng, Fangcheng Fu, Ling Yang, Wentao Zhang, and Bin Cui. Retrieval-augmented generation for ai-generated content: A survey. arXiv preprint arXiv:2402.19473, 2024. find