H-Mem: A Novel Memory Mechanism for Evolving and Retrieving Agent Memory via a Hybrid Structure
The paper introduces H-Mem, a new memory mechanism designed for agents using Large Language Models. This mechanism aims to enhance the evolution and retrieval of memory data, addressing limitations in current approaches. H-Mem utilizes a hybrid structure that combines temporal and semantic trees with knowledge graphs to improve performance on question-answering tasks.
- ▪H-Mem is a novel memory mechanism for agents using Large Language Models.
- ▪It effectively models the evolution of memory data over time and improves retrieval efficiency.
- ▪Extensive experiments demonstrate that H-Mem achieves state-of-the-art performance on question-answering tasks.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Computation and Language arXiv:2605.15701 (cs) [Submitted on 15 May 2026] Title:H-Mem: A Novel Memory Mechanism for Evolving and Retrieving Agent Memory via a Hybrid Structure Authors:Jiawei Yu, Yixiang Fang, Xilin Liu, Yuchi Ma View a PDF of the paper titled H-Mem: A Novel Memory Mechanism for Evolving and Retrieving Agent Memory via a Hybrid Structure, by Jiawei Yu and 3 other authors View PDF HTML (experimental) Abstract:Memory data are ubiquitous in Large Language Model (LLM)-based agents (e.g., OpenClaw and Manus).
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
