Elamur Research⌕
☰
home / papers / 2026 / H-Mem: Hybrid Multi-Dimensional Memory Management for Long-Context Conversational Agents

H-Mem: Hybrid Multi-Dimensional Memory Management for Long-Context Conversational Agents

Paper · 2026 · Zihe Ye, Jingyuan Huang, Weixin Chen et al.

Summary

Conversational agents that run over long, multi-session dialogues need memory they can query reliably. Most existing memory systems index information along one axis only, usually semantic similarity and occasionally recency, so retrieval often misses evidence that depends on both when and what. H-Mem addresses this gap. First, an LLM-based Memory Curation Agent filters each dialogue turn and drops small talk. Each turn worth keeping is stored as a FactNode, which is attached to two parallel hierarchies: a Temporal Tree, ordered by date, and a Semantic Tree, grouped by topic. At query time, an LLM Mode Controller picks one of two retrieval strategies:

  • Mode 1 (sequential search): used when the question has no explicit timeframe. Vector-search "seed" facts are expanded through nearby temporal context via lowest-common-ancestor (LCA) expansion.
  • Mode 2 (intersective search): used when an absolute time is mentioned. Topic-based and time-based hits are retrieved separately and intersected.

On the LoCoMo benchmark, the authors report gains of more than 8.4% over state-of-the-art memory systems, with the largest benefits on temporal and multi-hop questions. The work suggests that structuring memory along several dimensions is a promising basis for stateful agents.

Key points

  • Venue and code: published at EACL 2026 (Long Papers) by Rutgers / Hong Kong Baptist University authors; code is available at github.com/rutgerswiselab/H-mem.
  • FactNode structure: each node holds a timestamp, speaker, source text, and links to its parents in both the temporal and semantic trees.
  • Ablation on LoCoMo (full H-Mem: F1 47.57, B1 41.40, J 72.47):
    • Removing the Temporal Tree lowers J to 71.04.
    • Replacing LCA expansion with a fixed window lowers J to 70.71, though F1/B1 rise slightly (48.14 / 41.91).
    • Forcing Mode 1 only lowers J to 71.30.
  • Seed-fact sensitivity: J-score rises as the number of top-k seed facts grows from 5 to 30, with diminishing gains near 30; k = 30 is used in the main experiments.
  • Stated limitations: the system depends heavily on strong LLMs, and its tree modelling is still basic. Future work includes support for smaller models and adding dimensions such as sentiment and conversational goals.

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, topics and concepts

📚 Shelves 4The 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 works

🧭 Topics 1Research subjects and languages the work is about.Dialogue systems · 654 works

Cite

Cite as Zihe Ye, Jingyuan Huang, Weixin Chen and Yongfeng Zhang. 2026. H-Mem: Hybrid Multi-Dimensional Memory Management for Long-Context Conversational Agents. Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.eacl-long.363

@inproceedings{ye2026h,
  title = {{H-Mem: Hybrid Multi-Dimensional Memory Management for Long-Context Conversational Agents}},
  author = {Zihe Ye, Jingyuan Huang, Weixin Chen and Yongfeng Zhang},
  booktitle = {Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.eacl-long.363},
  url = {https://doi.org/10.18653/v1/2026.eacl-long.363}
}

References

46 references, 17 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, and 1 others. 2023. Gpt-4 technical report. arXiv preprint arXiv:2303.08774.
  2. John R Anderson. 1983. A spreading activation theory of memory. Journal of Verbal Learning and Verbal Behavior, 22(3):261–295. find
  3. Owen Bianchi, Mathew J Koretsky, Maya Willey, Chelsea X Alvarado, Tanay Nayak, Adi Asija, Nicole Kuznetsov, Mike A Nalls, Faraz Faghri, and Daniel Khashabi. 2025. Lost in the haystack: Smaller needles are more difficult for llms to find. arXiv preprint arXiv:2505.18148. find
  4. Weixin Chen, Yuhan Zhao, Jingyuan Huang, Zihe Ye, Clark Mingxuan Ju, Tong Zhao, Neil Shah, Li Chen, and Yongfeng Zhang. 2026. Memrec: Collaborative memory-augmented agentic recommender system. arXiv preprint arXiv:2601.08816. find
  5. 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.
  6. Tsz Ting Chung, Leyang Cui, Lemao Liu, Xinting Huang, Shuming Shi, and Dit-Yan Yeung. 2024. Selection-p: Self-supervised task-agnostic prompt compression for faithfulness and transferability. arXiv preprint arXiv:2410.11786. find
  7. Payel Das, Subhajit Chaudhury, Elliot Nelson, Igor Melnyk, Sarath Swaminathan, Sihui Dai, Aurélie Lozano, Georgios Kollias, Vijil Chenthamarakshan, Jiří, Navrátil, Soham Dan, and Pin-Yu Chen. 2024. Larimar: Large language models with episodic memory control. Preprint, arXiv:2403.11901. find
  8. Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. Bert: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 conference of the North American chapter of the association for computational linguistics: human language technologies,…
  9. Yiming Du, Wenyu Huang, Danna Zheng, Zhaowei Wang, Sebastien Montella, Mirella Lapata, Kam-Fai Wong, and Jeff Z Pan. 2025. Rethinking memory in ai: Taxonomy, operations, topics, and future directions. arXiv preprint arXiv:2505.00675. — [arXiv:2505.00675](https://arxiv.org/abs/2505.00675) find
  10. Minghao Guo, Xi Zhu, Haochen Xue, Chong Zhang, Shuhang Lin, Jingyuan Huang, Ziyi Ye, and Yongfeng Zhang. 2025. Reagan: Node-as-agent-reasoning graph agentic network. Preprint, arXiv:2508.00429. — [arXiv:2508.00429](https://arxiv.org/abs/2508.00429) find
  11. Bernal Jiménez Gutiérrez, Yiheng Shu, Yu Gu, Michihiro Yasunaga, and Yu Su. 2024. Hipporag: Neurobiologically inspired long-term memory for large language models. In The Thirty-eighth Annual Conference on Neural Information Processing Systems.
  12. Zifan He, Yingqi Cao, Zongyue Qin, Neha Prakriya, Yizhou Sun, and Jason Cong. 2024. Hmt: Hierarchical memory transformer for efficient long context language processing. arXiv preprint arXiv:2405.06067. find
  13. Sirui Hong, Mingchen Zhuge, Jonathan Chen, Xiawu Zheng, Yuheng Cheng, Ceyao Zhang, Jinlin Wang, Zili Wang, Steven Ka Shing Yau, Zijuan Lin, and 1 others. 2024. Metagpt: Meta programming for a multi-agent collaborative framework. International Conference on Learning Representations, ICLR.
  14. Jingyuan Huang, Dan Luo, Zihe Ye, Weixin Chen, Minghao Guo, and Yongfeng Zhang. 2026. From aggregation to selection: User-validated distributed social recommendation. Preprint, arXiv:2505.21388. — [arXiv:2505.21388](https://arxiv.org/abs/2505.21388) find
  15. Romuald A Janik. 2023. Aspects of human memory and large language models. arXiv preprint arXiv:2311.03839. — [arXiv:2311.03839](https://arxiv.org/abs/2311.03839) find
  16. Soyeong Jeong, Jinheon Baek, Sukmin Cho, Sung Ju Hwang, and Jong C Park. 2024. Adaptive-rag: Learning to adapt retrieval-augmented large language models through question complexity. arXiv preprint arXiv:2403.14403.
  17. Xia Jiang, Wei Xu, Ming Zhao, Li Chen, and Zhicheng Dou. 2024. Graphrag: Unlocking llms for graph-structured reasoning. arXiv preprint arXiv:2403.09236. — [arXiv:2403.09236](https://arxiv.org/abs/2403.09236) find
  18. Greg Kamradt. 2023. Llmtest_needleinahaystack: Doing simple retrieval from llm models at various context lengths to measure accuracy. find
  19. Charles Packer, Vivian Fang, Shishir_G Patil, Kevin Lin, Sarah Wooders, and Joseph_E Gonzalez. 2023. Memgpt: Towards llms as operating systems. find
  20. Jiazheng Kang, Mingming Ji, Zhe Zhao, and Ting Bai. 2025. Memory os of ai agent. arXiv preprint arXiv:2506.06326.
  21. Zhuoshi Pan, Qianhui Wu, Huiqiang Jiang, Xufang Luo, Hao Cheng, Dongsheng Li, Yuqing Yang, Chin-Yew Lin, H Vicky Zhao, Lili Qiu, and 1 others. 2025. Secom: On memory construction and retrieval for personalized conversational agents. In The Thirteenth International Conference on Learning Representati… find
  22. Vladimir Karpukhin, Barlas Oguz, Sewon Min, Patrick SH Lewis, Ledell Wu, Sergey Edunov, Danqi Chen, and Wen-tau Yih. 2020. Dense passage retrieval for open-domain question answering. In EMNLP (1), pages 6769–6781.
  23. LangChain. 2025. Langmem. //langchain-ai.github.io/langmem/. https: find
  24. 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–…
  25. Jiaqi Li, Mengmeng Wang, Zilong Zheng, and Muhan Zhang. 2023. Loogle: Can long-context language models understand long contexts? arXiv preprint arXiv:2311.04939.
  26. Zhiyu Li, Shichao Song, Chenyang Xi, Hanyu Wang, Chen Tang, Simin Niu, Ding Chen, Jiawei Yang, Chunyu Li, Qingchen Yu, and 1 others. 2025. Memos: A memory os for ai system. arXiv preprint arXiv:2507.03724. — [arXiv:2507.03724](https://arxiv.org/abs/2507.03724) find
  27. Nelson F Liu, Kevin Lin, John Hewitt, Ashwin Paranjape, Michele Bevilacqua, Fabio Petroni, and Percy Liang. 2023. Lost in the middle: How language models use long contexts. arXiv preprint arXiv:2307.03172.
  28. Adyasha Maharana, Dong-Ho Lee, Sergey Tulyakov, Mohit Bansal, Francesco Barbieri, and Yuwei Fang. 2024. Evaluating very long-term conversational memory of llm agents. arXiv preprint arXiv:2402.17753. find
  29. Kai Tzu-iunn Ong, Namyoung Kim, Minju Gwak, Hyungjoo Chae, Taeyoon Kwon, Yohan Jo, Seung-won Hwang, Dongha Lee, and Jinyoung Yeo. 2025. Towards lifelong dialogue agents via timeline-based memory management. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Associati… find
  30. OpenAI. 2024. Hello gpt-4o. OpenAI Blog. find
  31. OpenAI. 2024. Memory and new controls for chatgpt. https://openai.com/index/ memory-and-new-controls-for-chatgpt/. ↗
  32. Hongjin Qian, Zheng Liu, Peitian Zhang, Kelong Mao, Defu Lian, Zhicheng Dou, and Tiejun Huang. 2025. Memorag: Boosting long context processing with global memory-enhanced retrieval augmentation. In Proceedings of the ACM on Web Conference 2025, pages 2366–2377. find
  33. Preston Rasmussen, Pavlo Paliychuk, Travis Beauvais, Jack Ryan, and Daniel Chalef. 2025. Zep: a temporal knowledge graph architecture for agent memory. arXiv preprint arXiv:2501.13956.
  34. Alireza Rezazadeh, Zichao Li, Wei Wei, and Yujia Bao. 2025. From isolated conversations to hierarchical schemas: Dynamic tree memory representation for LLMs. In The Thirteenth International Conference on Learning Representations. find
  35. Parth Sarthi, Salman Abdullah, Aditi Tuli, Shubh Khanna, Anna Goldie, and Christopher D Manning. 2024. Raptor: Recursive abstractive processing for tree-organized retrieval. In The Twelfth International Conference on Learning Representations.
  36. Larry R Squire and Pablo Alvarez. 1995. Retrograde amnesia and memory consolidation: a neurobiological perspective. Current opinion in neurobiology, 5(2):169–177. find
  37. Larry R Squire, Lisa Genzel, John T Wixted, and Richard G Morris. 2015. Memory consolidation. Cold Spring Harbor perspectives in biology, 7(8):a021766. find
  38. Weihang Su, Qingyao Ai, Jingtao Zhan, Qian Dong, and Yiqun Liu. 2025. Dynamic and parametric retrieval-augmented generation. In Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval, pages 4118–4121. find
  39. Haoran Sun and Shaoning Zeng. 2025. Hierarchical memory for high-efficiency long-term reasoning in llm agents. arXiv preprint arXiv:2507.22925. — [arXiv:2507.22925](https://arxiv.org/abs/2507.22925) find
  40. Zhen Tan, Jun Yan, I Hsu, Rujun Han, Zifeng Wang, Long T Le, Yiwen Song, Yanfei Chen, Hamid Palangi, George Lee, and 1 others. 2025. In prospect and retrospect: Reflective memory management for long-term personalized dialogue agents. arXiv preprint arXiv:2503.08026.
  41. Marlieke TR Van Kesteren, Dirk J Ruiter, Guillén Fernández, and Richard N Henson. 2012. How schema and novelty augment memory formation. Trends in neurosciences, 35(4):211–219. find
  42. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. Advances in neural information processing systems, 30.
  43. Qingyue Wang, Yanhe Fu, Yanan Cao, Shuai Wang, Zhiliang Tian, and Liang Ding. 2025. Recursively summarizing enables long-term dialogue memory in large language models. Neurocomputing, 639:130193. find
  44. Di Wu, Hongwei Wang, Wenhao Yu, Yuwei Zhang, Kai-Wei Chang, and Dong Yu. 2025a. Longmemeval: Benchmarking chat assistants on long-term interactive memory. In The Thirteenth International Conference on Learning Representations.
  45. Qinggang Zhang, Shengyuan Chen, Yuanchen Bei, Zheng Yuan, Huachi Zhou, Zijin Hong, Junnan Dong, Hao Chen, Yi Chang, and Xiao Huang. 2025b. A survey of graph retrieval-augmented generation for customized large language models. arXiv preprint arXiv:2501.13958.
  46. 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/ye-2026-h-mem-hybrid-multi-dimensional/