HiAgent：面向大语言模型长程智能体任务求解的分层工作记忆管理
HiAgent: Hierarchical Working Memory Management for Solving Long-Horizon Agent Tasks with Large Language Model
AI总结：
针对长程任务中现有LLM智能体工作记忆利用不足、信息冗余的问题，提出HiAgent分层工作记忆管理框架，以子目标为记忆块优化记忆，实现成功率翻倍、步骤数减少，兼具鲁棒性与泛化性。
摘要：
基于大语言模型（LLM）的智能体作为交互式系统，通过处理环境观测生成目标任务的可执行动作，在多个领域展现出巨大潜力。这类智能体的效能受其记忆机制的显著影响，该机制将历史经验记录为动作-观测对序列。我们将记忆分为两类：跨试次记忆（在多次尝试中积累）和试次内记忆（即工作记忆，在单次尝试中积累）。尽管已有大量研究通过跨试次记忆优化智能体性能，但通过提升工作记忆利用率来增强智能体表现的研究仍不充分。现有方法通常直接将全部历史动作-观测对输入LLM，导致长程任务中出现信息冗余。受人类问题解决策略启发，本文提出HiAgent框架，该框架以子目标为记忆块，对基于LLM的智能体的工作记忆进行分层管理。具体而言，HiAgent提示LLM在生成可执行动作前先制定子目标，并让LLM主动决定用总结后的观测替换先前的子目标，仅保留与当前子目标相关的动作-观测对。在五项长程任务上的实验结果表明，HiAgent将成功率提升了一倍，且平均所需步骤数减少了3.8。此外，分析显示HiAgent在不同步骤数下均能持续提升性能，凸显了其鲁棒性和泛化性。项目主页：https://github.com/HiAgent2024/HiAgent 。
英文摘要：
Large Language Model (LLM)-based agents exhibit significant potential across various domains, operating as interactive systems that process environmental observations to generate executable actions for target tasks. The effectiveness of these agents is significantly influenced by their memory mechanism, which records historical experiences as sequences of action-observation pairs. We categorize memory into two types: cross-trial memory, accumulated across multiple attempts, and in-trial memory (working memory), accumulated within a single attempt. While considerable research has optimized performance through cross-trial memory, the enhancement of agent performance through improved working memory utilization remains underexplored. Instead, existing approaches often involve directly inputting entire historical action-observation pairs into LLMs, leading to redundancy in long-horizon tasks. Inspired by human problem-solving strategies, this paper introduces HiAgent, a framework that leverages subgoals as memory chunks to manage the working memory of LLM-based agents hierarchically. Specifically, HiAgent prompts LLMs to formulate subgoals before generating executable actions and enables LLMs to decide proactively to replace previous subgoals with summarized observations, retaining only the action-observation pairs relevant to the current subgoal. Experimental results across five long-horizon tasks demonstrate that HiAgent achieves a twofold increase in success rate and reduces the average number of steps required by 3.8. Additionally, our analysis shows that HiAgent consistently improves performance across various steps, highlighting its robustness and generalizability. Project Page: https://github.com/HiAgent2024/HiAgent .
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
