icedpanda/awesome-personalized-graph-memory

[ICKG] Survey resources for personalized graph memory for llm agents.

50 commits

updated Sep 9, 2026

arxiv.org/abs/2609.08599

See the code

README

Awesome Graph-Based Personalized Memory for LLM Agents

GitHub stars GitHub forks

This repo is a curated list of papers, systems, and benchmarks on graph-based personalized memory for LLM agents, based on our survey, "Graph-Based Personalized Memory for LLM Agents: Representation, Evolution, Retrieval, and Evaluation". We organize the literature around four stages of the memory lifecycle: representation, evolution, retrieval, and evaluation.

We hope this repository can help researchers and practitioners navigate this emerging area.

This is a fast growing area so contributions of missing papers are welcome.

News

🎉 Our paper has been accepted to the IEEE International Conference on Knowledge Graphs (ICKG 2026)!

Graph-based personalized memory system

Overview of a graph-based personalized agent memory system.

📌 Contents

News
Related Surveys
- Core Related Surveys
- Broader Surveys
Personal Assistants
Memory Representation
Memory Evolution
Memory Retrieval
Evaluation
Applications
Contributing
Citation

Surveys on personalized agents, graph-based agent memory, and modular agent memory.

Work	Links
Graph-based Agent Memory: Taxonomy, Techniques, and Applications (arXiv'26)	[paper] [code]
Toward Personalized LLM-Powered Agents: Foundations, Evaluation, and Future Directions (arXiv'26)	[paper]
Memory in the LLM Era: Modular Architectures and Strategies in a Unified Framework (arXiv'26)	[paper]

Broader Surveys

Work	Links
A Survey on the Memory Mechanism of Large Language Model based Agents (arXiv'24)	[paper] [code]
From Storage to Experience: A Survey on the Evolution of LLM Agent Memory Mechanisms (ACL Findings'26)	[paper]
Graph-Augmented Large Language Model Agents: Current Progress and Future Prospects (IEEE Intelligent Systems'26)	[paper]
Large Language Model Agent: A Survey on Methodology, Applications and Challenges (arXiv'25)	[paper]
Memory in the Age of AI Agents (arXiv'25)	[paper]
Personalized Generation In Large Model Era: A Survey (ACL'25)	[paper]

Personal Assistants

Personal assistants and memory systems that motivate long-term agent memory.

Work	Links
Dreaming: Better Memory for ChatGPT (OpenAI Blog'26)	[link]
Hermes Agent	[link] [code]
OpenClaw	[link] [code]
Supermemory	[link]
Vellum Assistant: A Personal AI Assistant That Evolves With You	[link]

Memory Representation

How user facts, preferences, episodes, and relations are stored and structured with graphs. The papers below are grouped by structure.

Graph-based memory structure patterns

Representative graph-based memory patterns: flat, hierarchical, hypergraph, hybrid, and multiple disjoint graphs.

Flat Graph

User-related facts and relations as nodes and typed edges, without an explicit abstraction hierarchy. Heterogeneous types alone do not count as hierarchy.

Work	Links
Agentic Memory: Learning Unified Long-Term and Short-Term Memory Management for Large Language Model Agents (ACL'26)	[paper]
Beyond Dialogue Time: Temporal Semantic Memory for Personalized LLM Agents (ACL'26)	[paper]
EchoGuard: An Agentic Framework with Knowledge-Graph Memory for Detecting Manipulative Communication in Longitudinal Dialogue (arXiv'26)	[paper]
Knowledge Graph Tuning: Real-time Large Language Model Personalization based on Human Feedback (arXiv'24)	[paper]
Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory (ECAI'25)	[paper]
MemORAI: Memory Organization and Retrieval via Adaptive Graph Intelligence for LLM Conversational Agents (ACL Findings'26)	[paper]
Memory Matters More: Event-Centric Memory as a Logic Map for Agent Searching and Reasoning (ACL'26)	[paper]
Time is Not a Label: Continuous Phase Rotation for Temporal Knowledge Graphs and Agentic Memory (arXiv'26)	[paper]
User Memory Reasoning for Conversational Recommendation (COLING'20)	[paper]

Hierarchical Graph

Evidence in layers (turns, facts, topics, personas) so user state is available at more than one granularity.

Work	Links
A Simple Yet Strong Baseline for Long-Term Conversational Memory of LLM Agents (arXiv'25)	[paper]
Agentic Recommender System with Hierarchical Belief-State Memory (arXiv'26)	[paper]
Amory: Building Coherent Narrative-Driven Agent Memory through Agentic Reasoning (EACL'26)	[paper]
AriGraph: Learning Knowledge Graph World Models with Episodic Memory for LLM Agents (IJCAI'25)	[paper]
Bi-Mem: Bidirectional Construction of Hierarchical Memory for Personalized LLMs via Inductive-Reflective Agents (arXiv'26)	[paper]
GAAMA: Graph Augmented Associative Memory for Agents (arXiv'26)	[paper]
GAM: Hierarchical Graph-based Agentic Memory for LLM Agents (ACL'26)	[paper]
G-Memory: Tracing Hierarchical Memory for Multi-Agent Systems (NeurIPS'25)	[paper] [code]
Hierarchical Long-Term Semantic Memory for LinkedIn's Hiring Agent (KDD'26)	[paper]
Hierarchical Memory Orchestration for Personalized Persistent Agents (arXiv'26)	[paper]
Hindsight is 20/20: Building Agent Memory that Retains, Recalls, and Reflects (arXiv'25)	[paper]
HyperMem: Hypergraph Memory for Long-Term Conversations (ACL'26)	[paper]
LEGOMem: Modular Procedural Memory for Multi-agent LLM Systems for Workflow Automation (arXiv'25)	[paper]
LOOM: Personalized Learning Informed by Daily LLM Conversations Toward Long-Term Mastery via a Dynamic Learner Memory Graph (PerFM@AAAI'26)	[paper]
SGMem: Sentence Graph Memory for Long-Term Conversational Agents (arXiv'25)	[paper]
Zep: A Temporal Knowledge Graph Architecture for Agent Memory (arXiv'25)	[paper]

Hypergraph

Joint context stored as hyperedges, representing more complex relationships.

Work	Links
HingeMem: Boundary Guided Long-Term Memory with Query Adaptive Retrieval for Scalable Dialogues (WWW'26)	[paper]
HyperMem: Hypergraph Memory for Long-Term Conversations (ACL'26)	[paper]

Hybrid Graph

A graph for relations, plus a complementary store (tree, summaries, passages, or evidence buffer).

Work	Links
H-Mem: A Novel Memory Mechanism for Evolving and Retrieving Agent Memory via a Hybrid Structure (arXiv'26)	[paper]
LiCoMemory: Lightweight and Cognitive Agentic Memory for Efficient Long-Term Reasoning (ACL'26)	[paper]
MemCog: From Memory-as-Tool to Memory-as-Cognition in Conversational Agents (arXiv'26)	[paper]
MemWeaver: Weaving Hybrid Memories for Traceable Long-Horizon Agentic Reasoning (ACL'26)	[paper]

Multiple Disjoint Graphs

Separate graph instances for different timescales or functions.

Work	Links
DEMENTIA-PLAN: An Agent-Based Framework for Multi-Knowledge Graph Retrieval-Augmented Generation in Dementia Care (KGPPH@AAAI'25)	[paper]

Memory Evolution

How the graph is updated: whether to write, how to attach, how to handle conflict, how to reorganize, whether to consolidate to avoid memory bloat, and how to forget or delete.

Evolution of graph-based personalized memory

Evolution of graph-based personalized memory: admission, integration, conflict resolution, consolidation, and removal.

Admission

Whether incoming evidence becomes persistent memory, and at what type or granularity.

Work	Links
Knowledge Graph Tuning: Real-time Large Language Model Personalization based on Human Feedback (arXiv'24)	[paper]
MemGuard: Preventing Memory Contamination in Long-Term Memory-Augmented Large Language Models (arXiv'26)	[paper]
MemORAI: Memory Organization and Retrieval via Adaptive Graph Intelligence for LLM Conversational Agents (ACL Findings'26)	[paper]
What Deserves Memory: Adaptive Memory Distillation for LLM Agents (ACL'26)	[paper]

Integration

Local writes: append-and-link, merge, or buffer before later consolidation.

Work	Links
Emotion-Attended Stateful Memory (EASM): The Architecture for Hyper-Personalization at Scale (arXiv'26)	[paper]
GAM: Hierarchical Graph-based Agentic Memory for LLM Agents (ACL'26)	[paper]
HingeMem: Boundary Guided Long-Term Memory with Query Adaptive Retrieval for Scalable Dialogues (WWW'26)	[paper]
LiCoMemory: Lightweight and Cognitive Agentic Memory for Efficient Long-Term Reasoning (ACL'26)	[paper]
Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory (ECAI'25)	[paper]

Conflict Resolution

Temporal supersession, user correction, or context-dependent coexistence — not a single newest-wins rule.

Work	Links
AriadneMem: Threading the Maze of Lifelong Memory for LLM Agents (arXiv'26)	[paper] [code]
GRAVITY: Architecture-Agnostic Structured Anchoring for Long-Horizon Conversational Memory (arXiv'26)	[paper]
Zep: A Temporal Knowledge Graph Architecture for Agent Memory (arXiv'25)	[paper]

Consolidation

Reorganization after many writes: abstraction, deduplication, topology edits, archival.

Work	Links
All-Mem: Agentic Lifelong Memory via Dynamic Topology Evolution (arXiv'26)	[paper]
EXG: Self-Evolving Agents with Experience Graphs (arXiv'26)	[paper]
GAM: Hierarchical Graph-based Agentic Memory for LLM Agents (ACL'26)	[paper]
HAGE: Harnessing Agentic Memory via RL-Driven Weighted Graph Evolution (arXiv'26)	[paper]
Managing Procedural Memory in LLM Agents: Control, Adaptation, and Evaluation (arXiv'26)	[paper]
Memory-R1: Enhancing Large Language Model Agents to Manage and Utilize Memories via Reinforcement Learning (ACL'26)	[paper]
Remember Me, Refine Me: A Dynamic Procedural Memory Framework for Experience-Driven Agent Evolution (ACL Findings'26)	[paper]
Rethinking How to Remember: Beyond Atomic Facts in Lifelong LLM Agent Memory (arXiv'26)	[paper]
RGMem: Renormalization Group-inspired Memory Evolution for Language Agents (ICML'26)	[paper] [code]
SAGE: A Self-Evolving Agentic Graph-Memory Engine for Structure-Aware Associative Memory (arXiv'26)	[paper]

Removal

Forgetting (lower exposure or resolution while something remains stored) versus deletion or archival of the active state.

Work	Links
Graph-Native Cognitive Memory for AI Agents: Formal Belief Revision Semantics for Versioned Memory Architectures (arXiv'26)	[paper]
ScrapMem: A Bio-inspired Framework for On-device Personalized Agent Memory via Optical Forgetting (arXiv'26)	[paper]

Memory Retrieval

How a request maps to a bounded subset of the user graph.

Similarity-Based Retrieval

Lexical, keyword, or dense matching to generate candidates or graph anchors.

Work	Links
Beyond RAG for Agent Memory: Retrieval by Decoupling and Aggregation (arXiv'26)	[paper]
Crafting Personalized Agents through Retrieval-Augmented Generation on Editable Memory Graphs (EMNLP'24)	[paper]
Does Memory Need Graphs? A Unified Framework and Empirical Analysis for Long-Term Dialog Memory (ACL'26)	[paper]
GRAVITY: Architecture-Agnostic Structured Anchoring for Long-Horizon Conversational Memory (arXiv'26)	[paper]

Structure-Based Retrieval

Expand or refine candidates through relations, layers, or relational views.

Work	Links
AssoMem: Scalable Memory QA with Multi-Signal Associative Retrieval (arXiv'25)	[paper]
Beyond RAG for Agent Memory: Retrieval by Decoupling and Aggregation (arXiv'26)	[paper]
GAM: Hierarchical Graph-based Agentic Memory for LLM Agents (ACL'26)	[paper]
Graph Retrieval-Augmented LLM for Conversational Recommendation Systems (PAKDD'25)	[paper]
H-Mem: A Novel Memory Mechanism for Evolving and Retrieving Agent Memory via a Hybrid Structure (arXiv'26)	[paper]
HyperMem: Hypergraph Memory for Long-Term Conversations (ACL'26)	[paper]
MAGMA: A Multi-Graph based Agentic Memory Architecture for AI Agents (ACL'26)	[paper]
MemORAI: Memory Organization and Retrieval via Adaptive Graph Intelligence for LLM Conversational Agents (ACL Findings'26)	[paper]
Memory is Reconstructed, Not Retrieved: Graph Memory for LLM Agents (ICML'26)	[paper] [code]
MemWeaver: Weaving Hybrid Memories for Traceable Long-Horizon Agentic Reasoning (ACL'26)	[paper]
PersonaAgent with GraphRAG: Community-Aware Knowledge Graphs for Personalized LLM (arXiv'25)	[paper]
PersonalAI: A Systematic Comparison of Knowledge Graph Storage and Retrieval Approaches for Personalized LLM agents (IEEE Access'26)	[paper]
PersonalAI 2.0: Enhancing knowledge graph traversal/retrieval with planning mechanism for Personalized LLM Agents (arXiv'26)	[paper]
PRISM: Pareto-Efficient Retrieval over Intent-Aware Structured Memory for Long-Horizon Agents (arXiv'26)	[paper]
Reasoning over User Preferences: Knowledge Graph-Augmented LLMs for Explainable Conversational Recommendations (ICDM'25)	[paper]

Adaptive and Agentic Retrieval

Query interpretation, expansion, stopping, routing, or compression that depends on the request and intermediate evidence.

Work	Links
ReAct: Synergizing Reasoning and Acting in Language Models (ICLR'23)	[paper] [code]
ActMem: Bridging the Gap Between Memory Retrieval and Reasoning in LLM Agents (arXiv'26)	[paper]
APEX-MEM: Agentic Semi-Structured Memory with Temporal Reasoning for Long-Term Conversational AI (ACL'26)	[paper]
HingeMem: Boundary Guided Long-Term Memory with Query Adaptive Retrieval for Scalable Dialogues (WWW'26)	[paper]
MemORAI: Memory Organization and Retrieval via Adaptive Graph Intelligence for LLM Conversational Agents (ACL Findings'26)	[paper]
MemRouter: Memory-as-Embedding Routing for Long-Term Conversational Agents (arXiv'26)	[paper]
MemToolAgent: Leveraging Memory for Tool Using Agents Based on Environment and User Feedback (arXiv'26)	[paper]
Memory is Reconstructed, Not Retrieved: Graph Memory for LLM Agents (ICML'26)	[paper] [code]
PRISM: Pareto-Efficient Retrieval over Intent-Aware Structured Memory for Long-Horizon Agents (arXiv'26)	[paper]

Evaluation

Benchmarks and protocols.

Long-Horizon Recall

Multi-session conversational memory under long histories.

Work	Links
Evaluating Long-Horizon Memory for Multi-Party Collaborative Dialogues (arXiv'26)	[paper]
Evaluating Very Long-Term Conversational Memory of LLM Agents (ACL'24)	[paper]
Supermemory: LongMemEval - Supermemory Research (Research report)	[link]
LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (ICLR'25)	[paper]
LongMemEval-V2: Evaluating Long-Term Agent Memory Toward Experienced Colleagues (arXiv'26)	[paper]

Personalization and User State

Profile, preference, and persona tracking rather than generic recall.

Work	Links
Do LLMs Recognize Your Preferences? Evaluating Personalized Preference Following in LLMs (ICLR'25)	[paper]
Know Me, Respond to Me: Benchmarking LLMs for Dynamic User Profiling and Personalized Responses at Scale (COLM'25)	[paper] [code]
PerLTQA: A Personal Long-Term Memory Dataset for Memory Classification, Retrieval, and Synthesis in Question Answering (arXiv'24)	[paper]
PersonaMem-v2: Towards Personalized Intelligence via Learning Implicit User Personas and Agentic Memory (arXiv'25)	[paper]
RealMem: Benchmarking LLMs in Real-World Memory-Driven Interaction (ACL Findings'26)	[paper]
SocialMemBench: Are AI Memory Systems Ready for Social Group Settings? (arXiv'26)	[paper]

Structure and Evolution

Graph organization, update over time, and failure attribution.

Work	Links
According to Me: Long-Term Personalized Referential Memory QA (arXiv'26)	[paper]
ActMem: Bridging the Gap Between Memory Retrieval and Reasoning in LLM Agents (arXiv'26)	[paper]
Benchmarking Deep Search over Heterogeneous Enterprise Data (EMNLP Industry'25)	[paper]
EngramaBench: Evaluating Long-Term Conversational Memory with Structured Graph Retrieval (arXiv'26)	[paper]
Evaluating Memory Structure in LLM Agents (arXiv'26)	[paper]
EvoArena: Tracking Memory Evolution for Robust LLM Agents in Dynamic Environments (arXiv'26)	[paper]
EvoMemBench: Benchmarking Agent Memory from a Self-Evolving Perspective (arXiv'26)	[paper]
MemoryCD: Benchmarking Long-Context User Memory of LLM Agents for Lifelong Cross-Domain Personalization (Lifelong Agent@ICLR'26)	[paper]
MemTrace: Tracing and Attributing Errors in Large Language Model Memory Systems (arXiv'26)	[paper]
WildGraphBench: Benchmarking GraphRAG with Wild-Source Corpora (ACL Findings'26)	[paper] [code]

Applications

Same memory ideas in a specific task: personalized dialogue, recommendation, embodied or GUI agents, and auditing.

Work	Links
AdaMem: Adaptive User-Centric Memory for Long-Horizon Dialogue Agents (arXiv'26)	[paper]
AMEM4Rec: Leveraging Cross-User Similarity for Memory Evolution in Agentic LLM Recommenders (arXiv'26)	[paper]
Agentic Recommender System with Hierarchical Belief-State Memory (arXiv'26)	[paper]
Beyond Dialogue Time: Temporal Semantic Memory for Personalized LLM Agents (ACL'26)	[paper]
Crafting Personalized Agents through Retrieval-Augmented Generation on Editable Memory Graphs (EMNLP'24)	[paper]
DEMENTIA-PLAN: An Agent-Based Framework for Multi-Knowledge Graph Retrieval-Augmented Generation in Dementia Care (KGPPH@AAAI'25)	[paper]
EchoGuard: An Agentic Framework with Knowledge-Graph Memory for Detecting Manipulative Communication in Longitudinal Dialogue (arXiv'26)	[paper]
Empowering Retrieval-based Conversational Recommendation with Contrasting User Preferences (NAACL'25)	[paper]
EXG: Self-Evolving Agents with Experience Graphs (arXiv'26)	[paper]
Graph Retrieval-Augmented LLM for Conversational Recommendation Systems (PAKDD'25)	[paper]
Hierarchical Long-Term Semantic Memory for LinkedIn's Hiring Agent (KDD'26)	[paper]
LOOM: Personalized Learning Informed by Daily LLM Conversations Toward Long-Term Mastery via a Dynamic Learner Memory Graph (PerFM@AAAI'26)	[paper]
MemAudit: Post-hoc Auditing of Poisoned Agent Memory via Causal Attribution and Structural Anomaly Detection (arXiv'26)	[paper]
MemoCRS: Memory-enhanced Sequential Conversational Recommender Systems with Large Language Models (CIKM'24)	[paper]
MemRec: Collaborative Memory-Augmented Agentic Recommender System (ACL'26)	[paper]
Open-Ended Instructable Embodied Agents with Memory-Augmented Large Language Models (EMNLP Findings'23)	[paper] [code]
Personalized Large Language Model Assistant with Evolving Conditional Memory (COLING'25)	[paper]
PersonaAgent with GraphRAG: Community-Aware Knowledge Graphs for Personalized LLM (arXiv'25)	[paper]
Personalizing Embodied Multimodal Large Language Model Agents over Long-term User Interactions (arXiv'26)	[paper]
Reasoning over User Preferences: Knowledge Graph-Augmented LLMs for Explainable Conversational Recommendations (ICDM'25)	[paper]
SE-GA: Memory-Augmented Self-Evolution for GUI Agents (ICML'26)	[paper]
Skill-Pro: Learning Reusable Skills from Experience via Non-Parametric PPO for LLM Agents (ICML'26)	[paper]
Temporal User Profiling with LLMs: Balancing Short-Term and Long-Term Preferences for Recommendations (arXiv'25)	[paper]
TraceMem: Weaving Narrative Memory Schemata from User Conversational Traces (arXiv'26)	[paper] [code]
User Memory Reasoning for Conversational Recommendation (COLING'20)	[paper]
User Simulator-Guided Multi-Turn Preference Optimization for Reasoning LLM-based Conversational Recommendation (arXiv'26)	[paper]

🤝 Contributing

PRs are welcome. Please:

Add entries to the relevant table using | **Paper Title** (Venue'YY) | [[paper]](url) [[code]](url) |. Omit the code link when unavailable.
List a paper under every category it illustrates (representation structure, evolution operation, retrieval mechanism, evaluation family, or application). Duplicate rows are intended; they make papers findable by category.

📃 Citation

@article{nguyen2026graph,
  title={Graph-Based Personalized Memory for LLM Agents: Representation, Evolution, Retrieval, and Evaluation},
  author={Nguyen, Dac Duy Anh and Qiu, Zhangchi and Chen, Shigeng and Liew, Alan Wee-Chung},
  journal={arXiv preprint arXiv:2609.08599},
  year={2026}
}

agent-memory

awesome-list

graph-memory

llm-agents

personalization

personalized-graph-memory

personalized-graph-memory-survey

survey

Similar repos

DEEP-PolyU/Awesome-GraphMemory

A survey of Graph-based Agent Memory | A curated list of resources (surveys, papers, benchmarks,…

355

TsinghuaC3I/Awesome-Memory-for-Agents

A Collection of Papers about Memory for Language Agents

663

Shichun-Liu/Agent-Memory-Paper-List

The paper list of "Memory in the Age of AI Agents: A Survey"

2,404

tfatykhov/awesome-agent-memory

Curated research on memory systems for LLM agents

Clare-Nie/Awesome-Personalization-in-MLLMs

Awesome papers on personalization in LLMs/MLLMs: memory, alignment, retrieval, and evaluation.

revectores/awesome-llm-memory

Curated papers on memory mechanisms for large language models: agent memory, RAG, KV cache,…

Elvin-Yiming-Du/Survey_Memory_in_AI

This repository introduce a comprehensive paper list, datasets, methods and tools for memory…

353

yxf203/Awesome-Efficient-Agents

Survey and paper list on efficiency-guided LLM agents (memory, tool use, planning).

309