Learning What to Share and What to Personalize: Hierarchical Strategy Co-Evolution for Agent Memory
cs.AI, cs.CL
Submitted: 2026-08-26
Updated: 2026-08-26
Comments: EMNLP'2026 Main Conference
Code: https://github.com/Hyp26cs/HiPS
License: http://creativecommons.org/licenses/by/4.0/
The gist: Memory-augmented agents maintain compact user profiles throughout extended conversations, enabling personalized and consistent responses without the need to process the entire dialogue history.
Terminology
Abstract
Memory-augmented agents maintain compact user profiles throughout extended conversations, enabling personalized and consistent responses without the need to process the entire dialogue history. The quality of these user profiles relies on the underlying memory management strategy: at each step, the agent must determine what to retain, compress, or discard. However, existing methods typically employ a static, one-size-fits-all strategy established before training. In practice, the optimal memory decision is inherently user-specific and dynamically evolves alongside policy optimization. To address this, we propose HiPS (Hierarchical Personalized Strategy), a framework that decouples memory management into a globally shared foundation and a user-specific adaptive tier. Specifically, HiPS employs Universal Strategy to extract shared principles from cross-persona trajectories, alongside Persona Delta Distillation to generate tailored rules for users whose behaviors diverge from general patterns. Cross-Level Rule Flow dynamically calibrates their boundary by promoting broadly validated personal rules and demoting contradicted global ones. The architecture establishes a co-evolution loop where a mechanism guarantees that all strategy refinements are anchored to task outcomes. Extensive experiments demonstrate consistent improvements over memory-augmented baselines.
Sources
- Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory
- LightMem: Lightweight and Efficient Memory-Augmented Generation
- EverMemOS: A Self-Organizing Memory Operating System for Structured Long-Horizon Reasoning
- Memory in the Age of AI Agents
- Know Me, Respond to Me: Benchmarking LLMs for Dynamic User Profiling and Personalized Responses at Scale
- MemOS: An Operating System for Memory-Augmented Generation (MAG) in Large Language Models
- PERMA: Benchmarking Personalized Memory Agents via Event-Driven Preference and Realistic Task Environments
- MemGPT: Towards LLMs as Operating Systems
- Qwen2.5 Technical Report
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- Agentic Memory: Learning Unified Long-Term and Short-Term Memory Management for Large Language Model Agents
- Breaking the Self-Confirming Loop: Diagnosing and Mitigating Systemic Reward Bias in Self-Rewarding RL
- MiniLM: Deep Self-Attention Distillation for Task-Agnostic Compression of Pre-Trained Transformers
- Learning How and What to Memorize: Cognition-Inspired Two-Stage Optimization for Evolving Memory
- Memory-R1: Enhancing Large Language Model Agents to Manage and Utilize Memories via Reinforcement Learning
- MemSkill: Learning and Evolving Memory Skills for Self-Evolving Agents
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems