Awesome Agent Memory Papers

A curated list of papers on memory for LLM / multimodal agents — methods, benchmarks, and surveys — covering episodic, semantic, procedural, and multimodal memory, with both parametric (internal) and retrieval-based (external) storage, learned via prompting, supervised finetuning, or reinforcement learning.

90 papers · 7 surveys · 31 benchmarks · 52 methods · last updated 2026-04-21

Interactive dashboard with multi-tag filtering: https://yyyujintang.github.io/Awesome-Agent-Memory-Papers/

Contributions welcome — open an issue or PR with new papers.

Contents

• Surveys
• Benchmarks
  • QA-based Memory Evaluation (5)
  • Web Navigation (7)
  • Desktop / Mobile GUI (6)
  • Embodied & Game Environments (6)
  • General Long-Horizon / Office (7)
• Methods
  • Multimodal Memory (16)
  • Procedural Memory (10)
  • Episodic Memory (18)
  • Semantic Memory (2)
  • Internal / Parametric Memory (4)
  • Other Methods (2)
• Tag Legend

Surveys

• Rethinking Memory Mechanisms of Foundation Agents in the Second Half
  2026-01-14 · Jiawei Han, Philip Yu
  Survey
• AI Meets Brain: Memory Systems from Cognitive Neuroscience to Autonomous Agents
  2025-12-29 · [code]
  Survey
• Memory in the Age of AI Agents
  2025-12-15 · Shuicheng Yan, Guibin Zhang · [code]
  Survey
• Measuring Agents in Production
  2025-12-02 · Shuicheng Yan, Guibin Zhang
  Survey
• Retrieval Augmented Generation and Understanding in Vision: A Survey and New Outlook
  2025-03-23 · Xuming Hu
  Survey
• Episodic memory in AI agents poses risks that should be studied and mitigate
  2025-01-20
  Survey
• A Survey on the Memory Mechanism of Large Language Model based Agents
  2024-04-21
  Survey

Benchmarks

Evaluation suites for agent memory, split by interaction mode.

QA-based Memory Evaluation

• KnowMe-Bench: Benchmarking Person Understanding for Lifelong Digital Companions
  2026-01-08
  Benchmark QA
• Mem-Gallery: Benchmarking Multimodal Long-Term Conversational Memory for MLLM Agents
  2026-01-07
  Benchmark QA
• Evaluating Memory in LLM Agents via Incremental Multi-Turn Interactions
  2025-07-07 · ICLR26 · Yuanzhe Hu · [code]
  Benchmark QA
• LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory
  2024-10-14
  Benchmark QA
• (LoCoMo) Evaluating Very Long-Term Conversational Memory of LLM Agents
  2024-02-27
  Benchmark QA

Web Navigation

• WebChoreArena: Evaluating Web Browsing Agents on Realistic Tedious Web Tasks
  2025-06-02
  Benchmark Web
• RealWebAssist: A Benchmark for Long-Horizon Web Assistance with Real-World Users
  2025-04-14
  Benchmark Web
• The BrowserGym Ecosystem for Web Agent Research
  2024-12-06
  Benchmark Web
• VisualWebArena: Evaluating Multimodal Agents on Realistic Visual Web Tasks
  2024-01-24 · ACL24 · [code]
  Benchmark Web
• WebArena: A Realistic Web Environment for Building Autonomous Agents
  2023-07-25 · ICLR24 · Shuyan Zhou, Duke
  Benchmark Web
• Mind2Web: Towards a Generalist Agent for the Web
  2023-06-09 · NeurIPS23, Spotlight
  Benchmark Web
• WebShop: Towards Scalable Real-World Web Interaction with Grounded Language Agents
  2022-07-04 · NeurIPS22 · [code]
  Benchmark Web

Desktop / Mobile GUI

• Gym-Anything: Turn any Software into an Agent Environment
  2026-04-07
  Benchmark GUI
• MemGUI-Bench: Benchmarking Memory of Mobile GUI Agents in Dynamic Environments
  2026-02-03 · [code]
  Benchmark GUI
• OS-Marathon: Benchmarking Computer-Use Agents on Long-Horizon Repetitive Tasks
  2026-01-28
  Benchmark GUI
• LongHorizonUI: A Unified Framework for Robust long-horizon Task Automation of GUI Agent
  2026-01-26 · ICLR26
  Benchmark GUI
• VisualAgentBench: Towards Large Multimodal Models as Visual Foundation Agents
  2024-08-12 · THU-Jie Tang · [code]
  Benchmark GUI
• OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments
  2024-04-11 · NeurIPS24 · [code]
  Benchmark GUI

Embodied & Game Environments

• AGENTVISTA: Evaluating Multimodal Agents in Ultra-Challenging Realistic Visual Scenarios
  2026-02-26 · Junxian He, May Fung
  Benchmark Embodied
• MentisOculi: Revealing the Limits of Reasoning with Mental Imagery
  2026-02-02
  Benchmark Embodied
• CAPTURe: Evaluating Spatial Reasoning in Vision Language Models via Occluded Object Counting
  2025-04-21 · ICCV25 · Mohit Bansal
  Benchmark Embodied
• ALFWorld: Aligning Text and Embodied Environments for Interactive Learning
  2020-10-08 · ICLR21
  Benchmark Embodied
• ALFRED: A Benchmark for Interpreting Grounded Instructions for Everyday Tasks
  2019-12-03 · CVPR20
  Benchmark Embodied
• TextWorld: A Learning Environment for Text-based Game
  2018-06-29 · IJCAI18
  Benchmark Embodied

General Long-Horizon / Office

• AMA-Bench: Evaluating Long-Horizon Memory for Agentic Applications
  2026-02-26
  Benchmark Long-Horizon
• MemoryArena: Benchmarking Agent Memory in Interdependent Multi-Session Agentic Tasks
  2026-02-12 · Yu Wang, Yuanzhe Hu
  Benchmark Long-Horizon
• A Framework for Studying AI Agent Behavior: Evidence from Consumer Choice Experiments
  2025-09-30 · ICLR26 · Nikhil, ABxLab
  Benchmark Long-Horizon
• OdysseyBench: Evaluating LLM Agents on Long-Horizon Complex Office Application Workflows
  2025-08-12
  Benchmark Long-Horizon
• TheAgentCompany: Benchmarking LLM Agents on Consequential Real World Tasks
  2024-12-18
  Benchmark Long-Horizon
• MemSim: A Bayesian Simulator for Evaluating Memory of LLM-based Personal Assistants
  2024-09-30 · NeurIPS25 · [code]
  Benchmark Long-Horizon
• AgentBench: Evaluating LLMs as Agents
  2023-08-07 · ICLR24 · [code]
  Benchmark Long-Horizon

Methods

Each paper is placed in exactly one primary section (Multimodal > Procedural > Episodic > Semantic > External > Internal). Tag badges on each entry show the full tag vector — use the website for true multi-axis filtering.

Multimodal Memory

• Omni-SimpleMem: Autoresearch-Guided Discovery of Lifelong Multimodal Agent Memory
  2026-04-01
  Method External Prompt-based Episodic Multimodal Procedural Semantic
• Visual Generation Unlocks Human-Like Reasoning through Multimodal World Models
  2026-01-27 · Mingsheng Long Bytedance Seed
  Method Internal SFT Multimodal
• MemOCR: Layout-Aware Visual Memory for Efficient Long-Horizon Reasoning
  2026-01-26
  Method External Prompt-based Episodic Multimodal
• MemVerse: Multimodal Memory for Lifelong Learning Agents
  2025-12-03 · [code]
  Method External Prompt-based Episodic Multimodal Procedural Semantic
• ViLoMem: Agentic Learner with Grow-and-Refine Multimodal Semantic Memory
  2025-11-26 · CVPR26 · [code]
  Method External Multimodal Semantic
• LongVT: Incentivizing "Thinking with Long Videos" via Native Tool Calling
  2025-11-25 · CVPR26
  Method External Prompt-based Multimodal Procedural
• VisMem: Latent Vision Memory Unlocks Potential of Vision-Language Models
  2025-11-14 · Shuicheng Yan
  Method Internal SFT Multimodal
• VAGEN: Reinforcing World Model Reasoning for Multi-Turn VLM Agents
  2025-10-19 · NeurIPS25 · [code]
  Method Internal RL-based Episodic Multimodal
• VideoLucy: Deep Memory Backtracking for Long Video Understanding
  2025-10-14 · NeurIPS25
  Method External SFT Episodic Multimodal
• (M3-Agent) Seeing, Listening, Remembering, and Reasoning: A Multimodal Agent with Long-Term Memory
  2025-08-13 · ICLR26 · ByteDance Seed · [code]
  Method External SFT Episodic Multimodal Semantic
• MAViS: A Multi-Agent Framework for Long-Sequence Video Storytelling
  2025-08-11
  Method External Prompt-based Episodic Multimodal
• Machine Mental Imagery: Empower Multimodal Reasoning with Latent Visual Token
  2025-06-20 · Chuang Gan · [code]
  Method Internal Multimodal
• 3DLLM-Mem: Long-Term Spatial-Temporal Memory for Embodied 3D Large Language Model
  2025-05-28
  Method External Prompt-based Episodic Multimodal Semantic
• Towards General Continuous Memory for Vision-Language Models
  2025-05-23 · NeurIPS25
  Method External Internal SFT Episodic Multimodal Semantic
• SAM2Act: Integrating Visual Foundation Model with A Memory Architecture for Robotic Manipulation
  2025-01-30 · [code]
  Method External Prompt-based Episodic Multimodal
• Optimus-1: Hybrid Multimodal Memory Empowered Agents Excel in Long-Horizon Tasks
  2024-08-07 · NeurIPS24 · [code]
  Method External Prompt-based Multimodal

Procedural Memory

• A Subgoal-driven Framework for Improving Long-Horizon LLM Agents
  2026-03-20
  Method External Prompt-based Training-free Episodic Procedural
• Plan-MCTS: Plan Exploration for Action Exploitation in Web Navigation
  2026-02-15 · Weinan Zhang
  Method RL-based Procedural
• MemSkill: Learning and Evolving Memory Skills for Self-Evolving Agents
  2026-02-02
  Method External RL-based Procedural
• TokMem: Tokenized Procedural Memory for Large Language Models
  2025-10-01
  Method Internal SFT Procedural
• ReasoningBank: Scaling Agent Self-Evolving with Reasoning Memory
  2025-09-29 · ICLR26 · Siru Ouyang
  Method External Prompt-based Episodic Procedural
• Memory Management and Contextual Consistency for Long-Running Low-Code Agents
  2025-09-27
  Method External Prompt-based Episodic Procedural
• Memory OS of AI Agent
  2025-05-30 · EMNLIP25 Main
  Method External Prompt-based Episodic Procedural Semantic
• A-MEM: Agentic Memory for LLM Agents
  2025-02-17 · NeurIPS25 · [code]
  Method External Prompt-based Episodic Procedural Semantic
• Agent Workflow Memory (AWM)
  2024-09-11 · ICML26 · [code]
  Method External Prompt-based Procedural
• Synapse: Trajectory-as-Exemplar Prompting with Memory for Computer Control
  2023-06-13 · [code]
  Method External Prompt-based Episodic Procedural

Episodic Memory

• Gated Memory Policy
  2026-04-21 · Shuran Song
  Method Internal RL-based Episodic
• HiGMem: A Hierarchical and LLM-Guided Memory System for Long-Term Conversational Agents
  2026-04-20
  Method External Prompt-based Training-free Episodic Semantic
• PlugMem: A Task-Agnostic Plugin Memory Module for LLM Agents
  2026-02-23 · [code]
  Method External Prompt-based Training-free Episodic Semantic
• Modeling Distinct Human Interaction in Web Agents
  2026-02-19
  Method External Prompt-based Episodic
• REMem: Reasoning with Episodic Memory in Language Agent
  2026-02-13 · Yu Su, Huan Sun
  Method External Prompt-based Episodic
• TraceMem: Weaving Narrative Memory Schemata from User Conversational Traces
  2026-02-10 · HKU
  Method External Prompt-based Episodic Semantic
• Learning to Continually Learn via Meta-learning Agentic Memory Designs
  2026-02-08 · [code]
  Method External RL-based Episodic
• Dep-Search: Learning Dependency-Aware Reasoning Traces with Persistent Memory
  2026-01-27
  Method External Prompt-based Episodic
• CAST: Character-and-Scene Episodic Memory for Agents
  2026-01-14
  Method External Prompt-based Episodic
• SimpleMem: Efficient Lifelong Memory for LLM Agents
  2026-01-05
  Method External Prompt-based Episodic Semantic
• Hindsight is 20/20: Building Agent Memory that Retains, Recalls, and Reflects
  2025-12-14
  Method External Prompt-based Episodic Semantic
• A neural network model of free recall learns multiple memory strategies
  2025-09-25 · [code]
  Method Internal Episodic
• PRIME: Large Language Model Personalization with Cognitive Dual-Memory and Personalized Thought Process
  2025-07-07 · EMNLP25, Main
  Method External Prompt-based Episodic Semantic
• Ella: Embodied Social Agents with Lifelong Memory
  2025-06-30 · Chuang Gan
  Method External Prompt-based Episodic Semantic
• Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory
  2025-04-28
  Method External Prompt-based Episodic Semantic
• R3Mem: Bridging Memory Retention and Retrieval via Reversible Compressio n
  2025-02-21
  Method External Prompt-based Episodic
• HippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language Models
  2024-05-23 · NeurIPS24 · Yu Su · [code]
  Method External Prompt-based Training-free Episodic Semantic
• MemoryBank: Enhancing Large Language Models with Long-Term Memory
  2023-05-17
  Method External Prompt-based Episodic Semantic

Semantic Memory

• Explicit v.s. Implicit Memory: Exploring Multi-hop Complex Reasoning Over Personalized Information
  2025-08-15 · SIGKDD 26 · Zeyu Zhang
  Method External Internal Prompt-based Semantic
• From RAG to Memory: Non-Parametric Continual Learning for Large Language Models (HippoRAG 2)
  2025-02-20 · ICML25 · [code]
  Method External Internal Prompt-based Semantic

Internal / Parametric Memory

• When to Memorize and When to Stop: Gated Recurrent Memory for Long-Context Reasoning
  2026-02-11 · Bytedance Seed
  Method Internal SFT
• QwenLong-L1.5: Post-Training Recipe for Long-Context Reasoning and Memory Management
  2025-12-25
  Method Internal SFT
• MemGen: Weaving Generative Latent Memory for Self-Evolving Agents
  2025-09-29 · Shuicheng Yan, Guibin Zhang · [code]
  Method Internal
• Scaling Test-time Compute for LLM Agents
  2025-06-15 · ICLR26
  Method Internal Prompt-based

Other Methods

• Agentic Reasoning for Large Language Models
  2026-01-18 · Heng Ji
  Method Prompt-based Training-free
• AgentRL: Scaling Agentic Reinforcement Learning with a Multi-Turn, Multi-Task Framework
  2025-10-05 · Jie Tang
  Method RL-based

Tag Legend

Citation

If this list is useful in your work, please consider starring the repo.