mlsanityData to 7 Oct 2026

Remember Before You're Asked: MemDream for Self-Probing Memory Evolution

arXiv 2609.34545 ↗First posted 28 Sep 2026cs.AIcited by 0

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

Abstract (arXiv)

Memory is essential for enabling LLM-based agents to maintain coherent, personalized behavior over long-horizon interactions. However, existing memory systems share a fundamental limitation: they never proactively test their own memory, repairing it only after real queries expose weaknesses. This reactive paradigm means every retrieval failure corresponds to a real interaction in which the cost has already been paid. We propose MemDream, a framework that enables self-probing memory evolution for LLM agents. Our framework periodically enters offline dream cycles where three specialized agents (Dreamer, Analyst, Consolidator) collaboratively probe, diagnose, and repair the memory graph before failures occur. A policy trained via Group Relative Policy Optimization learns which repair operations produce durable retrieval improvements, while a soft decay mechanism provides reversible forgetting driven by the same anticipatory signal. Experiments on LoCoMo and MemoryAgentBench demonstrate that MemDream improves answer F1 by 4.5 points on LoCoMo and achieves a 9.1-point higher overall score on MAB over the strongest reactive-evolution baselines.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Agent memory systems, one of 16 directions in LLM agents and tool use. Builds on 51 of its 55 references that are in the map. First posted on arXiv 28 Sep 2026.

Builds on

51 of 55 references.

Closest ancestors · 18

Same direction, newest first

  1. 2026TRUSTMEM: Learning Trustworthy Memory Consolidation for LLM Agents with Long-Term Memory23 Jun
  2. 2026RecMem: Recurrence-based Memory Consolidation for Efficient and Effective Long-Running LLM Agents15 May
  3. 2026EvolveMem:Self-Evolving Memory Architecture via AutoResearch for LLM Agents13 May
  4. 2026FSFM: A Biologically-Inspired Framework for Selective Forgetting of Agent Memory22 Apr
  5. 2026Choosing How to Remember: Adaptive Memory Structures for LLM Agents15 Feb
  6. 2026Beyond Static Summarization: Proactive Memory Extraction for LLM Agents8 Jan
  7. 2026MAGMA: A Multi-Graph based Agentic Memory Architecture for AI Agents6 Jan
  8. 2025MemR3: Memory Retrieval via Reflective Reasoning for LLM Agents23 Dec
  9. 2025MemEvolve: Meta-Evolution of Agent Memory Systems21 Dec
  10. 2025O-Mem: Omni Memory System for Personalized, Long Horizon, Self-Evolving Agents17 Nov
  11. 2025Memory-R1: Enhancing Large Language Model Agents to Manage and Utilize Memories via Reinforcement Learning27 Aug
  12. 2025Evaluating Memory in LLM Agents via Incremental Multi-Turn Interactions7 Jul
6 more
  1. 2025Memory OS of AI Agent30 May
  2. 2025Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory28 Apr
  3. 2025A-MEM: Agentic Memory for LLM Agents17 Feb
  4. 2025Zep: A Temporal Knowledge Graph Architecture for Agent Memory20 Jan
  5. 2024A Survey on the Memory Mechanism of Large Language Model based Agents21 Apr
  6. 2024Evaluating Very Long-Term Conversational Memory of LLM Agents27 Feb

Other references · 33

  1. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,160
  2. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  3. 2016Overcoming catastrophic forgetting in neural networkscited by 4,718
  4. 2023Reflexion: Language Agents with Verbal Reinforcement Learningcited by 4,064
  5. 2023Generative Agents: Interactive Simulacra of Human Behaviorcited by 3,500
  6. 2023A Survey on Large Language Model based Autonomous Agentscited by 2,203
  7. 2023Retrieval-Augmented Generation for Large Language Models: A Surveycited by 2,185
  8. 2023Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflectioncited by 1,800
  9. 2024From Local to Global: A Graph RAG Approach to Query-Focused Summarizationcited by 1,411
  10. 2018Continual Lifelong Learning with Neural Networks: A Reviewcited by 1,313
  11. 2023MemGPT: Towards LLMs as Operating Systemscited by 1,295
  12. 2017Continual Learning with Deep Generative Replaycited by 1,152
21 more
  1. 2023MemoryBank: Enhancing Large Language Models with Long-Term Memorycited by 539
  2. 2024RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrievalcited by 519
  3. 2024Personal LLM Agents: Insights and Survey about the Capability, Efficiency and Securitycited by 257
  4. 2025From RAG to Memory: Non-Parametric Continual Learning for Large Language Modelscited by 223
  5. 2025MemOS: A Memory OS for AI Systemcited by 129
  6. 2024Machine Unlearning of Pre-trained Large Language Modelscited by 118
  7. 2026MemRL: Self-Evolving Agents via Runtime Reinforcement Learning on Episodic Memorycited by 113
  8. 2025On Memory Construction and Retrieval for Personalized Conversational Agentscited by 101
  9. 2025AgentEvolver: Towards Efficient Self-Evolving Agent Systemcited by 77
  10. 2024Larimar: Large Language Models with Episodic Memory Controlcited by 53
  11. 2025FLEX: Continuous Agent Evolution via Forward Learning from Experiencecited by 51
  12. 2025Mem-α: Learning Memory Construction via Reinforcement Learningcited by 49
  13. 2020Energy-Based Models for Continual Learningcited by 32
  14. 2025MemSearcher: Training LLMs to Reason, Search and Manage Memory via End-to-End Reinforcement Learningcited by 26
  15. 2025From Newborn to Impact: Bias-Aware Citation Predictioncited by 9
  16. 2026Auto-Dreamer: Learning Offline Memory Consolidation for Language Agentscited by 9
  17. 2026From Query to Counsel: Structured Reasoning with a Multi-Agent Framework and Dataset for Legal Consultationcited by 3
  18. 2026SEER: Long-Context Reasoning via Selective Visual-Text Compressioncited by 3
  19. 2026Beyond Sparse Supervision: Diffusion-Guided Learning for Few-Shot Graph Fraud Detectioncited by 2
  20. 2026PREF-Gate: Provenance-Constrained Relational Evidence Fusion with Validation-Gated Selection for Graph Fraud Detectioncited by 2
  21. 2026Diagnose Before You Compress: Prediction-Independent Bottleneck Witness Refinement for LLM Serving Tracescited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026MemMA: Coordinating the Memory Cycle through Multi-Agent Reasoning and In-Situ Self-Evolution
  2. 2026CoEvo-Mem: Co-Evolving Retrieval Policy and Memory Bank for LLM Agents
  3. 2026MemGym: a Long-Horizon Memory Environment for LLM Agents
  4. 2026Retrieval-Driven Memory Reconsolidation for Long-Term LLM Agents
  5. 2026MemForest: An Efficient Agent Memory System with Hierarchical Temporal Indexing
  6. 2026ERSkill: Evolving for Skill-Guided Adaptive Memory Retrieval
  7. 2026All-Mem: Agentic Lifelong Memory via Dynamic Topology Evolution
  8. 2026MemTrain: Self-Supervised Context Memory Training