mlsanityData to 14 Oct 2026

GenMem: Generative Symbolic Memory for Self-Evolving Harness

arXiv 2609.34633 ↗First posted 28 Sep 2026cs.LGcited by 0

On the map: LLM agents and tool use → Agent memory systems

also in 1 more
Abstract (arXiv)

Long-term memory supports the self-evolution of LLM agents by retaining experience and skills across tasks and enabling their retrieval, reuse, and revision in subsequent long-horizon decision-making. Yet existing memory management approaches remain limited to discriminative retrieval and to address the sparse, hierarchical, and highly redundant structure of reusable experience: only a small, task-dependent subset of trajectories and memories warrants retention, retrieval, or revision. Learning these operations is further complicated by sparse, delayed, and indirect task-level feedback, with weak supervision across the memory lifecycle. Moreover, continual memory evolution introduces an architectural tension as addressing invariance: stored experience is perpetually revised, yet the addressing interface consumed by learned retrieval policies must remain stable. To address, we present GenMem, which reformulates memory management as generative symbolic addressing. Its core mechanism is the Symbolic Identifier (SID), a multi-level discrete token tuple drawn from a Cartesian-product address space that factorizes a million-scale sparse memory space using fewer than one hundred discrete symbols. Instead of generating ever-changing raw content, the memory agent learns to generate SIDs, while memory evolution rewrites the payload at a fixed address without shifting the address itself. Architecturally, GenMem couples a MemRetriever and a MemEvolver within a multi-agent harness, trained via GRPO with dense process and outcome rewards with two-channels optimization. Under offline memory evolution, experiments spanning ALFWorld, WebShop, multi-hop QA, medical reasoning, and deep research evaluate GenMem against strong memory-augmented baselines...

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Agent memory systems, one of 16 directions in LLM agents and tool use. Builds on 52 of its 56 references that are in the map. First posted on arXiv 28 Sep 2026.

Builds on

52 of 56 references.

Closest ancestors · 9

Same direction, newest first

  1. 2026Fine-Mem: Fine-Grained Feedback Alignment for Long-Horizon Memory Management13 Jan
  2. 2026StackPlanner: A Centralized Hierarchical Multi-Agent System with Task-Experience Memory Management9 Jan
  3. 2026AgentOCR: Reimagining Agent History via Optical Self-Compression8 Jan
  4. 2026SimpleMem: Efficient Lifelong Memory for LLM Agents5 Jan
  5. 2025Memory as Action: Autonomous Context Curation for Long-Horizon Agentic Tasks14 Oct
  6. 2025MemGen: Weaving Generative Latent Memory for Self-Evolving Agents29 Sep
  7. 2025MEM1: Learning to Synergize Memory and Reasoning for Efficient Long-Horizon Agents18 Jun
  8. 2025Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory28 Apr
  9. 2024A Survey on the Memory Mechanism of Large Language Model based Agents21 Apr

Other references · 43

  1. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,380
  2. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  3. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,695
  4. 2017Neural Discrete Representation Learningcited by 4,712
  5. 2023Reflexion: Language Agents with Verbal Reinforcement Learningcited by 4,276
  6. 2023SWE-bench: Can Language Models Resolve Real-World GitHub Issues?cited by 3,373
  7. 2021Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsitycited by 3,073
  8. 2023GPQA: A Graduate-Level Google-Proof Q&A Benchmarkcited by 2,879
  9. 2023Voyager: An Open-Ended Embodied Agent with Large Language Modelscited by 2,191
  10. 2023MetaGPT: Meta Programming for A Multi-Agent Collaborative Frameworkcited by 1,808
  11. 2025Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learningcited by 1,444
  12. 2022WebShop: Towards Scalable Real-World Web Interaction with Grounded Language Agentscited by 1,251
31 more
  1. 2020ALFWorld: Aligning Text and Embodied Environments for Interactive Learningcited by 1,205
  2. 2022Interleaving Retrieval with Chain-of-Thought Reasoning for Knowledge-Intensive Multi-Step Questionscited by 923
  3. 2023ExpeL: LLM Agents Are Experiential Learnerscited by 876
  4. 2022Autoregressive Image Generation using Residual Quantizationcited by 722
  5. 2023Can LLM Already Serve as A Database Interface? A BIg Bench for Large-Scale Database Grounded Text-to-SQLscited by 688
  6. 2023Cognitive Architectures for Language Agentscited by 429
  7. 2020Autoregressive Entity Retrievalcited by 334
  8. 2026SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learningcited by 278
  9. 2025DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agentscited by 248
  10. 2022Transformer Memory as a Differentiable Search Indexcited by 247
  11. 2025ZeroSearch: Incentivize the Search Capability of LLMs without Searchingcited by 208
  12. 2025BrowseComp-Plus: A More Fair and Transparent Evaluation Benchmark of Deep-Research Agentcited by 200
  13. 2023Recommender Systems with Generative Retrievalcited by 198
  14. 2026Trace2Skill: Distill Trajectory-Local Lessons into Transferable Agent Skillscited by 187
  15. 2025OneRec: Unifying Retrieve and Rank with Generative Recommender and Iterative Preference Alignmentcited by 173
  16. 2025Agentic Reinforced Policy Optimizationcited by 172
  17. 2025ReSearch: Learning to Reason with Search for LLMs via Reinforcement Learningcited by 161
  18. 2025EvolveR: Self-Evolving LLM Agents through an Experience-Driven Lifecyclecited by 137
  19. 2026MemRL: Self-Evolving Agents via Runtime Reinforcement Learning on Episodic Memorycited by 129
  20. 2026SKILL0: In-Context Agentic Reinforcement Learning for Skill Internalizationcited by 94
  21. 2026Skill-SD: Skill-Conditioned Self-Distillation for Multi-turn LLM Agentscited by 90
  22. 2026Conditional Memory via Scalable Lookup: A New Axis of Sparsity for Large Language Modelscited by 86
  23. 2026SkillX: Automatically Constructing Skill Knowledge Bases for Agentscited by 76
  24. 2026SkillRouter: Skill Routing for LLM Agents at Scalecited by 61
  25. 2026Self-Distilled Agentic Reinforcement Learningcited by 55
  26. 2025Agentic Entropy-Balanced Policy Optimizationcited by 40
  27. 2024TC-RAG:Turing-Complete RAG's Case study on Medical LLM Systemscited by 28
  28. 2026SkillGraph: Skill-Augmented Reinforcement Learning for Agents via Evolving Skill Graphscited by 15
  29. 2026Co-Evolving Skill Generation and Policy Optimizationcited by 7
  30. 2026AgenticRag-R1: Agentic Reinforcement Learning with Stack Memory for Multi-Step Reasoning, Retrieval and Memorizingcited by 2
  31. 2026Harness-RL: Black-Box Reinforcement Learning with Action-Args Decoupling for Central-Agent Multi-Agent Harnessescited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026When Continual Learning Moves to Memory: A Study of Experience Reuse in LLM Agents
  2. 2026Rethinking Memory as Continuously Evolving Connectivity
  3. 2026Retrieval-Driven Memory Reconsolidation for Long-Term LLM Agents
  4. 2026Memory Beyond Recall: A Dual-Process Cognitive Memory System for Self-Evolving LLM Agents
  5. 2026Memory as a Controlled Process: Learned Adaptive Memory Management for LLM Agents
  6. 2026CONTRAMEM: Learning Self-Evolving Procedural Memory from Contrasting Multi-Model Trajectories
  7. 2026EvolveMem:Self-Evolving Memory Architecture via AutoResearch for LLM Agents
  8. 2025MemEvolve: Meta-Evolution of Agent Memory Systems