mlsanityData to 14 Oct 2026

HasMem: Hard-Origin Adaptively Softened Memory for Long-Term LLM Agents

arXiv 2609.30797 ↗First posted 25 Sep 2026cs.AIcited by 0

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

also in 1 more
Abstract (arXiv)

Text-based memory and context compression support reuse of past interactions. Resizing continuous memory changes the input to a frozen LLM, coupling capacity allocation with readout. We propose Hard-Origin Adaptively Softened Memory (HasMem). Frozen hard-prompt embeddings provide a verifiable initial state. A controller adjusts memory widths, a Writer re-encodes resized entries, and Reader and Global provide readout adaptation and cross-turn state. On all 535 questions in a reconstruction probe derived from the Multi-Session Chat (MSC) development split, the main configuration achieves lexical F1 of 95.3 (+4.4 percentage points) at 93.6% of the hard reference's framed memory positions. With approximately matched per-question target body budgets, six configurations at mean per-entry retention around 0.83--0.91 exceed rule-based re-encoding by 8.0--23.6 exact-match (EM) percentage points. With fixed model parameters and rule target width ratio 0.75, Global's EM gain passes a user-level exact paired test with Bonferroni correction over eight comparisons. On all 500 LongMemEval-S questions, local lexical F1 rises from the hard reference's 3.4 to 8.9, and answer negative log-likelihood (NLL) falls from 12.257 to 5.274. F1 gains accompany lower EM on both evaluations.

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 49 of its 54 references that are in the map. First posted on arXiv 25 Sep 2026.

Builds on

49 of 54 references.

Closest ancestors · 7

Same direction, newest first

  1. 2026MemRefine: LLM-Guided Compression for Long-Term Agent Memory11 Jun
  2. 2026SimpleMem: Efficient Lifelong Memory for LLM Agents5 Jan
  3. 2025Memory-R1: Enhancing Large Language Model Agents to Manage and Utilize Memories via Reinforcement Learning27 Aug
  4. 2025Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory28 Apr
  5. 2025A-MEM: Agentic Memory for LLM Agents17 Feb
  6. 2024A Survey on the Memory Mechanism of Large Language Model based Agents21 Apr
  7. 2024A Human-Inspired Reading Agent with Gist Memory of Very Long Contexts15 Feb

Other references · 42

  1. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,380
  2. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,297
  3. 2019Parameter-Efficient Transfer Learning for NLPcited by 4,236
  4. 2016SQuAD: 100,000+ Questions for Machine Comprehension of Textcited by 4,148
  5. 2021Prefix-Tuning: Optimizing Continuous Prompts for Generationcited by 3,915
  6. 2021The Power of Scale for Parameter-Efficient Prompt Tuningcited by 3,795
  7. 2023Generative Agents: Interactive Simulacra of Human Behaviorcited by 3,568
  8. 2023Efficient Streaming Language Models with Attention Sinkscited by 2,186
  9. 2019Transformer-XL: Attentive Language Models Beyond a Fixed-Length Contextcited by 1,850
  10. 2023MemGPT: Towards LLMs as Operating Systemscited by 1,389
  11. 2021Improving language models by retrieving from trillions of tokenscited by 1,293
  12. 2024SnapKV: LLM Knows What You are Looking for Before Generationcited by 880
30 more
  1. 2023H2O: Heavy-Hitter Oracle for Efficient Generative Inference of Large Language Modelscited by 832
  2. 2019Compressive Transformers for Long-Range Sequence Modellingcited by 726
  3. 2021BARTScore: Evaluating Generated Text as Text Generationcited by 722
  4. 2024LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memorycited by 707
  5. 2023LongLLMLingua: Accelerating and Enhancing LLMs in Long Context Scenarios via Prompt Compressioncited by 453
  6. 2023Scissorhands: Exploiting the Persistence of Importance Hypothesis for LLM KV Cache Compression at Test Timecited by 439
  7. 2024PyramidKV: Dynamic KV Cache Compression based on Pyramidal Information Funnelingcited by 432
  8. 2023Learning to Compress Prompts with Gist Tokenscited by 390
  9. 2023GPTScore: Evaluate as You Desirecited by 361
  10. 2023Adapting Language Models to Compress Contextscited by 356
  11. 2024LLMLingua-2: Data Distillation for Efficient and Faithful Task-Agnostic Prompt Compressioncited by 348
  12. 2022Matryoshka Representation Learningcited by 282
  13. 2024HippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language Modelscited by 281
  14. 2021Beyond Goldfish Memory: Long-Term Open-Domain Conversationcited by 267
  15. 2024Ada-KV: Optimizing KV Cache Eviction by Adaptive Budget Allocation for Efficient LLM Inferencecited by 235
  16. 2022Recurrent Memory Transformercited by 203
  17. 2023In-context Autoencoder for Context Compression in a Large Language Modelcited by 199
  18. 2023Augmenting Language Models with Long-Term Memorycited by 189
  19. 2023Compressing Context to Enhance Inference Efficiency of Large Language Modelscited by 189
  20. 2023Focused Transformer: Contrastive Training for Context Scalingcited by 133
  21. 2024xRAG: Extreme Context Compression for Retrieval-augmented Generation with One Tokencited by 125
  22. 2025From Human Memory to AI Memory: A Survey on Memory Mechanisms in the Era of LLMscited by 124
  23. 2024Dynamic Memory Compression: Retrofitting LLMs for Accelerated Inferencecited by 116
  24. 2024MEMORYLLM: Towards Self-Updatable Large Language Modelscited by 93
  25. 2024Larimar: Large Language Models with Episodic Memory Controlcited by 53
  26. 2023Compressed Context Memory For Online Language Model Interactioncited by 43
  27. 2024Generative Adapter: Contextualizing Language Models in Parameters with A Single Forward Passcited by 26
  28. 2024Fundamental Limits of Prompt Compression: A Rate-Distortion Framework for Black-Box Language Modelscited by 19
  29. 2025Saliency-driven Dynamic Token Pruning for Large Language Modelscited by 5
  30. 2026FlexComp: One Model for Every Ratio in Context Compressioncited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026RPMem: Learning Long-Term Recurrent Parametric Memory Across Sessions for LLM Agents
  2. 2026Adaptive Memory Admission Control for LLM Agents
  3. 2026Can Agents Forget When Asked? A Benchmark for Memory Unlearning in LLM Agents
  4. 2026Towards Autonomous Memory Agents
  5. 2026Evaluating Memory Structure in LLM Agents
  6. 2025Look Back to Reason Forward: Revisitable Memory for Long-Context LLM Agents
  7. 2026CAMeR: Keyword-Gated Hybrid Activation for Adaptive Memory Retention in LLM Agents
  8. 2026Human-Inspired Memory Architecture for LLM Agents