[论文解读] CLAG: Adaptive Memory Organization via Agent-Driven Clustering for Small Language Model Agents

Taeyun Roh, Wonhee Jang+2 more|arXiv (Cornell University)|Mar 16, 2026

Topic Modeling被引用 0

一句话总结

CLAG 引入基于代理的聚类来组织面向小型语言模型代理的长距离记忆，从而实现局部记忆演化和两阶段、聚类感知的检索，在全局记忆池上提升鲁棒性与效率。

ABSTRACT

Large language model agents heavily rely on external memory to support knowledge reuse and complex reasoning tasks. Yet most memory systems store experiences in a single global retrieval pool which can gradually dilute or corrupt stored knowledge. This problem is especially pronounced for small language models (SLMs), which are highly vulnerable to irrelevant context. We introduce CLAG, a CLustering-based AGentic memory framework where an SLM agent actively organizes memory by clustering. CLAG employs an SLM-driven router to assign incoming memories to semantically coherent clusters and autonomously generates cluster-specific profiles, including topic summaries and descriptive tags, to establish each cluster as a self-contained functional unit. By performing localized evolution within these structured neighborhoods, CLAG effectively reduces cross-topic interference and enhances internal memory density. During retrieval, the framework utilizes a two-stage process that first filters relevant clusters via their profiles, thereby excluding distractors and reducing the search space. Experiments on multiple QA datasets with three SLM backbones show that CLAG consistently improves answer quality and robustness over prior memory systems for agents, remaining lightweight and efficient.

研究动机与目标

• 为需要结构化记忆的长距离 LLM 代理提供动机，以降低跨主题干扰。
• 提出一个基于聚类的记忆框架（CLAG），使记忆自组织为主题一致的簇。
• 在簇内实现局部记忆演化，以维持语义密度和连贯性。
• 开发一个两阶段、聚类感知的检索过程，以降低搜索空间和检索噪声。
• 在多种骨干模型与数据集上展示鲁棒性与效率提升。

提出的方法

• 代理路由通过一个 SLM 驱动的路由器将每条新记忆分配到语义上连贯的簇。
• 簇维护具有关键词和标签的主题特定轮廓，用于概述内容。
• 局部演化在被路由的簇内执行记忆整合，以增强语义密度。
• 两阶段检索先按轮廓筛选簇，然后在选定簇内检索记忆。
• 自适应聚类将过大的簇分割，以防止语义漂移并保持适应性。

Figure 1: Conceptual comparison between existing global memory systems and CLAG. (Left) Traditional approaches manage memories in a single global pool, where topic-mixed updates and retrieval lead to high interference and noise accumulation. (Right) CLAG employs agent-driven clustering to organize m

实验结果

研究问题

• RQ1 代理驱动的聚类是否能在多个领域提升以 SLM 支撑的代理的鲁棒性与答案质量？
• RQ2 在簇邻域内的局部演化是否比全局记忆更新更有效地缓解跨主题干扰？
• RQ3 与全局检索基线相比，聚类感知的检索是否降低检索噪声并提高效率？
• RQ4 CLAG 在多种骨干模型和知识密集任务上的表现如何？

主要发现

• CLAG 在三个问答基准和三个 SLM 骨干上持续提升答案质量与鲁棒性。
• 在知识密集和领域迁移设置下，CLAG 获得更强的 F1/BLEU-1 分数。
• 两阶段的聚类感知检索降低了搜索空间并抑制干扰项，提升了检索质量。
• 簇内的局部演化提高了记忆密度与连贯性，帮助复杂任务中的推理。
• CLAG 维持有利的延迟/吞吐，在关键数据集上的端到端效率优于某些基线。

Figure 2: Overview of the proposed CLAG framework. Left: Agentic Routing. An SLM router assigns each incoming memory note $m_{\text{new}}$ to the most relevant cluster using semantic metadata, and updates the corresponding cluster profile $\mathcal{P}$ . Middle: Localized Evolution. An evolution age