Elamur Research⌕
☰
home / papers / 2026 / AnchorMem: Anchored Facts with Associative Contexts for Building Memory in Large Language Models

AnchorMem: Anchored Facts with Associative Contexts for Building Memory in Large Language Models

Paper · 2026 · Zhanyu Shen, Sijie Cheng, Zhicheng Guo et al.

Summary

LLM memory systems such as A-Mem and Mem0 repeatedly summarize and rewrite past interactions. This can wash out fine details. Entity-graph indexing, as in HippoRAG 2, can instead link unrelated memories through common, generic entities. AnchorMem addresses both problems by separating what is used for retrieval from what is passed to the model for generation. The idea comes from involuntary autobiographical memory, the "Proust Phenomenon".

An LLM extracts atomic facts from the dialogue history to act as retrieval anchors. Each fact points back to its original, unmodified context chunk. Similar facts are then grouped into higher-order event nodes, forming an Associative Event Graph that ties related memories together without generic entities acting as bridges. At query time, the system matches the query against facts and events, but answers using the linked raw chunks and event descriptions.

On the LoCoMo benchmark, with three open- and closed-source backbones, AnchorMem beat baselines including LightMem and HippoRAG 2. It also built its memory fastest among the compared methods. This suggests that precise retrieval and faithful context do not have to be traded against each other.

Key points

  • Graph structure: The heterogeneous graph has three node types (interaction contexts, atomic facts, associative events). Its edges map facts to their source contexts and connect semantically similar facts.
  • Main settings: The chosen configuration uses a fact-similarity threshold τ = 0.85 and N = 3 fact neighbors. Redundancy pruning discards a fusion candidate once ⌈N/2⌉ retrieved facts share its overlapping content.
  • Effect of top-k: Raising top-k from 10 to 50 increased average accuracy from 69.54% to 75.45%, with Single Hop gaining 8.21%. Temporal accuracy, however, fell from 62.31% to 59.50%.
  • Cost on LoCoMo: Offline fact extraction took 1,264 s and event integration took 3,810 s. Online retrieval for all 1,540 queries took 6.53 s, about 4 ms per query.
  • Pruning and limitations: Pruning dropped 4,821 redundant fact clusters while keeping 3,570 integrated events. The authors note that extraction and grouping depend on LLM judgments and that event integration currently runs in a single pass.

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

 Institutions

🎓Tsinghua UniversityBeijing · China415 works · 94,028 citations🎓Shenzhen UniversityShenzhen · China22 works · 95 citations

 Fields, topics and concepts

📚 Shelves 5The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 worksInformation Systems · 1,730 works

🧭 Topics 1Research subjects and languages the work is about.Language models & pretraining · 3,894 works

Cite

Cite as Zhanyu Shen, Sijie Cheng, Zhicheng Guo, Weiqin Wang, Yile Wang and Hui Huang. 2026. AnchorMem: Anchored Facts with Associative Contexts for Building Memory in Large Language Models. Findings of the Association for Computational Linguistics: ACL 2026.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.findings-acl.1736

@inproceedings{shen2026anchormem,
  title = {{AnchorMem: Anchored Facts with Associative Contexts for Building Memory in Large Language Models}},
  author = {Zhanyu Shen, Sijie Cheng, Zhicheng Guo, Weiqin Wang, Yile Wang and Hui Huang},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2026},
  year = {2026},
  doi = {10.18653/v1/2026.findings-acl.1736},
  url = {https://doi.org/10.18653/v1/2026.findings-acl.1736}
}

References

28 references, 15 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Dorthe Berntsen. 2010. The unbidden past: Involuntary autobiographical memories as a basic mode of remembering. Current Directions in Psychological Science, 19(3):138–142. find
  2. Dorthe Berntsen. 2021. Involuntary autobiographical memories and their relation to other forms of spontaneous thoughts. Philosophical Transactions of the Royal Society B, 376(1817):20190693. find
  3. Prateek Chhikara, Dev Khant, Saket Aryan, Taranjeet Singh, and Deshraj Yadav. 2025. Mem0: Building production-ready ai agents with scalable long-term memory. arXiv preprint arXiv:2504.19413.
  4. Martin A Conway and Christopher W Pleydell-Pearce. 2000. The construction of autobiographical memories in the self-memory system. Psychological review, 107(2):261. find
  5. Deepseek-AI. 2025. Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning. arXiv preprint arXiv:2501.12948.
  6. André V. Duarte, João DS Marques, Miguel Graça, Miguel Freire, Lei Li, and Arlindo L. Oliveira. 2024. LumberChunker: Long-form narrative document segmentation. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 6473– 6486, Miami, Florida, USA. Association for Computation… find
  7. Nelson F. Liu, Kevin Lin, John Hewitt, Ashwin Paranjape, Michele Bevilacqua, Fabio Petroni, and Percy Liang. 2024. Lost in the middle: How language models use long contexts. Transactions of the Association for Computational Linguistics, 12:157–173.
  8. Darren Edge, Ha Trinh, Newman Cheng, Joshua Bradley, Alex Chao, Apurva Mody, Steven Truitt, Dasha Metropolitansky, Robert Osazuwa Ness, and Jonathan Larson. 2024. From local to global: A graph rag approach to query-focused summarization. arXiv preprint arXiv:2404.16130. find
  9. Adyasha Maharana, Dong-Ho Lee, Sergey Tulyakov, Mohit Bansal, Francesco Barbieri, and Yuwei Fang. 2024. Evaluating very long-term conversational memory of LLM agents. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 13851– 1387… find
  10. Jizhan Fang, Xinle Deng, Haoming Xu, Ziyan Jiang, Yuqi Tang, Ziwen Xu, Shumin Deng, Yunzhi Yao, Mengru Wang, Shuofei Qiao, and 1 others. 2025. Lightmem: Lightweight and efficient memory-augmented generation. arXiv preprint arXiv:2510.18866. — [arXiv:2510.18866](https://arxiv.org/abs/2510.18866) find
  11. Lingrui Mei, Jiayu Yao, Yuyao Ge, Yiwei Wang, Baolong Bi, Yujun Cai, Jiazhi Liu, Mingyu Li, Zhong-Zhi Li, Duzhen Zhang, and 1 others. 2025. A survey of context engineering for large language models. arXiv preprint arXiv:2507.13334. — [arXiv:2507.13334](https://arxiv.org/abs/2507.13334) find
  12. Yunfan Gao, Yun Xiong, Xinyu Gao, Kangxiang Jia, Jinliu Pan, Yuxi Bi, Yixin Dai, Jiawei Sun, Haofen Wang, and Haofen Wang. 2023. Retrieval-augmented generation for large language models: A survey. arXiv preprint arXiv:2312.10997, 2(1).
  13. Zirui Guo, Lianghao Xia, Yanhua Yu, Tu Ao, and Chao Huang. 2025. LightRAG: Simple and fast retrieval-augmented generation. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 10746–10761, Suzhou, China. Association for Computational Linguistics.
  14. Bernal Jiménez Gutiérrez, Yiheng Shu, Weijian Qi, Sizhe Zhou, and Yu Su. 2025. From RAG to memory: Non-parametric continual learning for large language models. In Forty-second International Conference on Machine Learning.
  15. Jiazheng Kang, Mingming Ji, Zhe Zhao, and Ting Bai. 2025. Memory OS of AI agent. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 25972–25981, Suzhou, China. Association for Computational Linguistics.
  16. Woosuk Kwon, Zhuohan Li, Siyuan Zhuang, Ying Sheng, Lianmin Zheng, Cody Hao Yu, Joseph Gonzalez, Hao Zhang, and Ion Stoica. 2023. Efficient memory management for large language model serving with pagedattention. In Proceedings of the 29th symposium on operating systems principles, pages 611–626.
  17. Philippe Laban, Hiroaki Hayashi, Yingbo Zhou, and Jennifer Neville. 2025. Llms get lost in multi-turn conversation. arXiv preprint arXiv:2505.06120.
  18. Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, and 1 others. 2020. Retrieval-augmented generation for knowledge-intensive nlp tasks. Advances in neural information processing systems, 33:9459–…
  19. OpenAI. 2023. Gpt-4 technical report. arXiv preprint arXiv:2303.08774.
  20. Charles Packer, Sarah Wooders, Kevin Lin, Vivian Fang, Shishir G. Patil, Ion Stoica, and Joseph E. Gonzalez. 2024. Memgpt: Towards llms as operating systems. arXiv preprint arXiv:2310.08560. find
  21. Renyi Qu, Ruixuan Tu, and Forrest Bao. 2025. Is semantic chunking worth the computational cost? In Findings of the Association for Computational Linguistics: NAACL 2025, pages 2155–2177. find
  22. Qwen. 2025. Qwen2.5 technical report. arXiv preprint arXiv:2412.15115. find
  23. Nils Reimers and Iryna Gurevych. 2019. Sentence-bert: Sentence embeddings using siamese bert-networks. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics.
  24. Di Wu, Hongwei Wang, Wenhao Yu, Yuwei Zhang, Kai-Wei Chang, and Dong Yu. 2025. Longmemeval: Benchmarking chat assistants on long-term interactive memory. In The Thirteenth International Conference on Learning Representations.
  25. Wujiang Xu, Zujie Liang, Kai Mei, Hang Gao, Juntao Tan, and Yongfeng Zhang. 2025. A-mem: Agentic memory for LLM agents. In The Thirty-ninth Annual Conference on Neural Information Processing Systems.
  26. Wenhao Yu, Hongming Zhang, Xiaoman Pan, Peixin Cao, Kaixin Ma, Jian Li, Hongwei Wang, and Dong Yu. 2024. Chain-of-note: Enhancing robustness in retrieval-augmented language models. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 14672–14685, Miami, Fl…
  27. Zeyu Zhang, Quanyu Dai, Xiaohe Bo, Chen Ma, Rui Li, Xu Chen, Jieming Zhu, Zhenhua Dong, and Ji-Rong Wen. 2025. A survey on the memory mechanism of large language model-based agents. ACM Trans. Inf. Syst., 43(6). find
  28. Wanjun Zhong, Lianghong Guo, Qiqi Gao, He Ye, and Yanlin Wang. 2024. Memorybank: Enhancing large language models with long-term memory. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 19724–19731. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/shen-2026-anchormem-anchored-facts-associative-contexts/