H-Mem: A Novel Memory Mechanism for Evolving and Retrieving Agent Memory via a Hybrid Structure

2026-05-15

AI Summary

This work addresses the limited utilization of memory in existing large language model agents, which stems from inadequate modeling of memory evolution over time and inefficient retrieval mechanisms. To overcome these limitations, the authors propose H-Mem, a novel hybrid memory architecture that uniquely integrates a temporal semantic tree with a knowledge graph to simultaneously capture the temporal dynamics of memory and relational structures among entities. This design enables a progressive consolidation of short-term memories into long-term storage and is accompanied by a tailored efficient retrieval algorithm. Experimental results on three agent memory benchmarks demonstrate that H-Mem significantly outperforms current approaches in question-answering tasks, achieving state-of-the-art performance.

Abstract

Memory data are ubiquitous in Large Language Model (LLM)-based agents (e.g., OpenClaw and Manus). A few recent works have attempted to exploit agents'memory for improving their performance on the question-answering (QA) task, but they lack a principled mechanism for effectively modeling how memory data evolves over time and retrieving memory data effectively, leading to poor performance in memory utilization. To fill this gap, we present H-Mem, a novel memory mechanism via a hybrid structure that can not only effectively model the evolution of agent memory over a long period of time, but also provide an efficient memory retrieval approach. Particularly, H-Mem builds a temporal and semantic tree structure that allows the short-term memory data to evolve progressively into long-term memory data, where the latter provides summarized information about the former, while simultaneously constructing a knowledge graph to capture the relationships between entities in memory. Moreover, it offers an effective memory retrieval approach by exploiting the hybrid structure of the tree and graph structures. Extensive experiments on three agent memory benchmarks show that H-Mem achieves state-of-the-art performance on the QA task.

Problem

• agent memory
• memory evolution
• memory retrieval
• question-answering
• Large Language Model

Innovation

• H-Mem
• memory evolution
• hybrid structure
• knowledge graph
• memory retrieval