[论文解读] MAPLE: A Sub-Agent Architecture for Memory, Learning, and Personalization in Agentic AI Systems

Deepak Babu Piskala|arXiv (Cornell University)|Feb 3, 2026

Persona Design and Applications被引用 0

一句话总结

MAPLE 将记忆、学习与个性化划分为独立子代理，以实现实时个性化和异步学习，相对于无状态基线有显著提升。MAPLE-Personas 基准显示更强的个性化与特征融入。

ABSTRACT

Large language model (LLM) agents have emerged as powerful tools for complex tasks, yet their ability to adapt to individual users remains fundamentally limited. We argue this limitation stems from a critical architectural conflation: current systems treat memory, learning, and personalization as a unified capability rather than three distinct mechanisms requiring different infrastructure, operating on different timescales, and benefiting from independent optimization. We propose MAPLE (Memory-Adaptive Personalized LEarning), a principled decomposition where Memory handles storage and retrieval infrastructure; Learning extracts intelligence from accumulated interactions asynchronously; and Personalization applies learned knowledge in real-time within finite context budgets. Each component operates as a dedicated sub-agent with specialized tooling and well-defined interfaces. Experimental evaluation on the MAPLE-Personas benchmark demonstrates that our decomposition achieves a 14.6% improvement in personalization score compared to a stateless baseline (p < 0.01, Cohen's d = 0.95) and increases trait incorporation rate from 45% to 75% -- enabling agents that genuinely learn and adapt.

研究动机与目标

• 解释为何在 LLM 代理中，记忆、学习与个性化在体系架构上应当区分开来。
• 提出具有专用工具与接口的三子代理 MAPLE 架构。
• 展示带有实时个性化的异步学习如何改进对用户的特定适应。

提出的方法

• 将代理认知分解为记忆（存储/检索）、学习（异步洞察提取）与个性化（实时适应）。
• 以神经生物学启发的记忆与检索-生成文献为设计基础。
• 通过 MAPLE-Personas 基准，在受控消融对照的无状态基线下进行评估。
• 采用两阶段流程：请求时的个性化 + 背景学习循环。

Figure 1. Personalization in action: Sarah (senior ML engineer) and Marcus (product manager) ask the same question about transformers. MAPLE retrieves different user profiles and generates tailored responses—technical implementation details for Sarah, conceptual analogies for Marcus.

实验结果

研究问题

• RQ1 将记忆、学习与个性化分离是否会在个性化质量上优于无状态基线？
• RQ2 将异步学习与实时个性化结合是否能带来更高的特征融入与“完美个性化”得分？
• RQ3 MAPLE 设计在多样化用户画像和查询集上的表现如何？
• RQ4 子代理分解在延迟与模块化方面有哪些好处？

主要发现

指标	基线	MAPLE	Δ
Judge Score (1–5)	4.17	4.78	+0.61***
Trait Incorp.	45%	75%	+30pp
Perfect (5/5)	15%	88%	+73pp

• MAPLE 在个性化得分上实现相对提升 14.6%（p<0.01；Cohen’s d=0.95）。
• 特征融入率从 45% 提升至 75%。
• 达到“完美个性化”得分（5/5）的比例从 15% 提升至 88%。

Figure 2. The evolution from foundation models to personalized agents. Base LLMs provide general capabilities; RAG adds knowledge retrieval; MAPLE introduces memory, learning, and personalization as distinct architectural components, enabling agents that adapt to individual users over time.