A Survey on the Memory Mechanism of Large Language Model based Agents
2024/04/21 by Zeyu Zhang, Zhang, Zeyu, Quanyu Dai +15 · 110 citations
Computer Science · #Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #Topic Modeling
paper · pdf · doi:10.48550/arxiv.2404.13501
openalex publication_date 2024/04/21 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Abstract
Large language model (LLM) based agents have recently attracted much attention from the research and industry communities. Compared with original LLMs, LLM-based agents are featured in their self-evolving capability, which is the basis for solving real-world problems that need long-term and complex agent-environment interactions. The key component to support agent-environment interactions is the memory of the agents. While previous studies have proposed many promising memory mechanisms, they are scattered in different papers, and there lacks a systematical review to summarize and compare these works from a holistic perspective, failing to abstract common and effective designing patterns for inspiring future studies. To bridge this gap, in this paper, we propose a comprehensive survey on the memory mechanism of LLM-based agents. In specific, we first discuss ''what is'' and ''why do we need'' the memory in LLM-based agents. Then, we systematically review previous studies on how to design and evaluate the memory module. In addition, we also present many agent applications, where the memory module plays an important role. At last, we analyze the limitations of existing work and show important future directions. To keep up with the latest advances in this field, we create a repository at \urlhttps://github.com/nuster1128/LLMAgentMemorySurvey.
Cited by
- Keep It InMind: Benchmarking the Implicit-Association Blind Spot in Agent Memory
- Memory for Large Language Models
- Agent Team Work Zone: An Automated, Persistent Workspace for Long-Lived Claude Code Agent Teams
- Agent-based simulation of online social networks and disinformation
- StoryMem: Multi-shot Long Video Storytelling with Memory
- MemEvolve: Meta-Evolution of Agent Memory Systems
- Learning Hierarchical Procedural Memory for LLM Agents through Bayesian Selection and Contrastive Refinement
- Hindsight is 20/20: Building Agent Memory that Retains, Recalls, and Reflects
- Beyond Task Completion: An Assessment Framework for Evaluating Agentic AI Systems
- Memoria: A Scalable Agentic Memory Framework for Personalized Conversational AI
- Remember Me, Refine Me: A Dynamic Procedural Memory Framework for Experience-Driven Agent Evolution
- PersonaMem-v2: Towards Personalized Intelligence via Learning Implicit User Personas and Agentic Memory
- The Vision Wormhole: Latent-Space Communication in Heterogeneous Multi-Agent Systems
- MemVerse: Multimodal Memory for Lifelong Learning Agents
- ThetaEvolve: Test-time Learning on Open Problems
- Agentic Learner with Grow-and-Refine Multimodal Semantic Memory
- EWE: An Agentic Framework for Extreme Weather Analysis
- General Agentic Memory Via Deep Research
- Cross-Disciplinary Knowledge Retrieval and Synthesis: A Compound AI Architecture for Scientific Discovery
- AnimAgents: Coordinating Multi-Stage Animation Pre-Production with Human-Multi-Agent Collaboration
- Agentifying Agentic AI
- Learning to Debug: LLM-Organized Knowledge Trees for Solving RTL Assertion Failures
- AccelOpt: A Self-Improving LLM Agentic System for AI Accelerator Kernel Optimization
- O-Mem: Omni Memory System for Personalized, Long Horizon, Self-Evolving Agents
- Scaling Graph Chain-of-Thought Reasoning: A Multi-Agent Framework with Efficient LLM Serving
- LiCoMemory: Lightweight and Cognitive Agentic Memory for Efficient Long-Term Reasoning
- Smarter Together: Creating Agentic Communities of Practice through Shared Experiential Learning
- The Imperfect Learner: Incorporating Developmental Trajectories in Memory-based Student Simulation
- LLMs as Judges: Toward The Automatic Review of GSN-compliant Assurance Cases
- Active Thinking Model: A Goal-Directed Self-Improving Framework for Real-World Adaptive Intelligence
- Dynamic Affective Memory Management for Personalized LLM Agents
- Remember When It Matters: Proactive Memory Agent for Long-Horizon Agents
- Metis: Memory Foundation Model
- Joint Agent Memory and Exploration Learning via Novelty Signals
- The Moltbook Files: A Harmless Slopocalypse or Humanity's Last Experiment
- State of the Art of LLM-Enabled Interaction with Visualization
- The Narrative Continuity Test: A Conceptual Framework for Evaluating Identity Persistence in AI Systems
- Evaluating Long-Term Memory for Long-Context Question Answering
- Co-Sight: Enhancing LLM-Based Agents via Conflict-Aware Meta-Verification and Trustworthy Reasoning with Structured Facts
- LightMem: Lightweight and Efficient Memory-Augmented Generation
- Empowering Real-World: A Survey on the Technology, Practice, and Evaluation of LLM-driven Industry Agents
- MemoryBench: A Benchmark for Memory and Continual Learning in LLM Systems
- The Gatekeeper Knows Enough
- AlphaQuanter: An End-to-End Tool-Orchestrated Agentic Reinforcement Learning Framework for Stock Trading
- Higher Satisfaction, Lower Cost: A Technical Report on How LLMs Revolutionize Meituan's Intelligent Interaction Systems
- GenCellAgent: Generalizable, Training-Free Cellular Image Segmentation via Large Language Model Agents
- VizCopilot: Fostering Appropriate Reliance on Enterprise Chatbots with Context Visualization
- Attacks by Content: Automated Fact-checking is an AI Security Issue
- PaperArena: An Evaluation Benchmark for Tool-Augmented Agentic Reasoning on Scientific Literature
- D3MAS: Decompose, Deduce, and Distribute for Enhanced Knowledge Sharing in Multi-Agent Systems
- Traj-CoA: Patient Trajectory Modeling via Chain-of-Agents for Lung Cancer Risk Prediction
- MemDecay: Region-Aware KV Cache Eviction for Efficient LLM Agent Inference
- The Personalization Trap: How User Memory Alters Emotional Reasoning in LLMs
- Student Development Agent: Risk-free Simulation for Evaluating AIED Innovations
- Preference-Aware Memory Update for Long-Term LLM Agents
- Diffusion-Inspired Masked Fine-Tuning for Knowledge Injection in Autoregressive LLMs
- MemWeaver: A Hierarchical Memory from Textual Interactive Behaviors for Personalized Generation
- Intelligent AI Delegation
- A Goal Without a Plan Is Just a Wish: Efficient and Effective Global Planner Training for Long-Horizon Agent Tasks
- The New Quant: A Survey of Large Language Models in Financial Prediction and Trading
- CAM: A Constructivist View of Agentic Memory for LLM-Based Reading Comprehension
- PsycholexTherapy: Simulating Reasoning in Psychotherapy with Small Language Models in Persian
- JoyAgent-JDGenie: Technical Report on the GAIA
- MEMTRACK: Evaluating Long-Term Memory and State Tracking in Multi-Platform Dynamic Agent Environments
- From Trace to Line: LLM Agent for Real-World OSS Vulnerability Localization
- TVS Sidekick: Challenges and Practical Insights from Deploying Large Language Models in the Enterprise
- ReasoningBank: Scaling Agent Self-Evolving with Reasoning Memory
- MemGen: Weaving Generative Latent Memory for Self-Evolving Agents
- SGMem: Sentence Graph Memory for Long-Term Conversational Agents
- Embodied AI: From LLMs to World Models
- LoopMemGR: From Behavior Logs to Evolving Memory for Generative Recommendation
- Σ-Mem: An Online Reliability Memory for LLM-based Multi-Agent Systems
- ChronoMem: Version Control and Semantic Rollback for Large Language Model Agent Memory
- AutoMem: Automated Learning of Memory as a Cognitive Skill
- LLM-based Agents Suffer from Hallucinations: A Survey of Taxonomy, Methods, and Directions
- MemOrb: A Plug-and-Play Verbal-Reinforcement Memory Layer for E-Commerce Customer Service
- Temporal-Aware User Behaviour Simulation with Large Language Models for Recommender Systems
- Chinese Court Simulation with LLM-Based Agent System
- AgenticIE: An Adaptive Agent for Information Extraction from Complex Regulatory Documents
- SWE-Mirror: Scaling Issue-Resolving Datasets by Mirroring Issues Across Repositories
- Meta-Policy Reflexion: Reusable Reflective Memory and Rule Admissibility for Resource-Efficient LLM Agent
- Social World Models
- AI Compute Architecture and Evolution Trends
- Memory-R1: Enhancing Large Language Model Agents to Manage and Utilize Memories via Reinforcement Learning
- A Multi-Memory Segment System for Generating High-Quality Long-Term Memory Content in Agents
- Explicit v.s. Implicit Memory: Exploring Multi-hop Complex Reasoning Over Personalized Information
- A Comprehensive Review of AI Agents: Transforming Possibilities in Technology and Beyond
- Learn to Memorize: Optimizing LLM-based Agents with Adaptive Memory Framework
- Memory-Augmented Transformers: A Systematic Review from Neuroscience Principles to Enhanced Model Architectures
- Intrinsic Memory Agents: Heterogeneous Multi-Agent LLM Systems through Structured Contextual Memory
- BlindGuard: Safeguarding LLM-based Multi-Agent Systems under Unknown Attacks
- SHIELDA: Structured Handling of Exceptions in LLM-Driven Agentic Workflows
- Measuring Stereotype and Deviation Biases in Large Language Models
- Memp: Exploring Agent Procedural Memory
- RoboTron-Sim: Improving Real-World Driving via Simulated Hard-Case
- OS Agents: A Survey on MLLM-based Agents for General Computing Devices Use
- Probing the Gaps in ChatGPT Live Video Chat for Real-World Assistance for People who are Blind or Visually Impaired
- Polymath: A Self-Optimizing Agent with Dynamic Hierarchical Workflow
- AgentArmor: Enforcing Program Analysis on Agent Runtime Trace to Defend Against Prompt Injection
- Sari Sandbox: A Virtual Retail Store Environment for Embodied AI Agents
- MemoCue: Empowering LLM-Based Agents for Human Memory Recall via Strategy-Guided Querying
- Towards Interpretable Renal Health Decline Forecasting via Multi-LMM Collaborative Reasoning Framework
- Evaluation and Benchmarking of LLM Agents: A Survey
- A Novel Self-Evolution Framework for Large Language Models
- Efficient Agents: Building Effective Agents While Reducing Cost
- Are We Ready For An Agent-Native Memory System?
- Confident RAG: Enhancing the Performance of LLMs for Mathematics Question Answering through Multi-Embedding and Confidence Scoring
- Hierarchical Memory for High-Efficiency Long-Term Reasoning in LLM Agents
- Aime: Towards Fully-Autonomous Multi-Agent Framework
- Agentic Large Language Models for Conceptual Systems Engineering and Design
Related