[2605.15156] MeMo: Memory as a Model

Abstract:Large language models (LLMs) achieve strong performance across a wide range of tasks, but remain frozen after pretraining until subsequent updates. Many real-world applications require timely, domain-specific information, motivating the need for efficient mechanisms to incorporate new knowledge. In this paper, we introduce MeMo (Memory as a Model), a modular framework that encodes new knowledge into a dedicated memory model while keeping the LLM parameters unchanged. Compared to existing methods, MeMo offers several advantages: (a) it captures complex cross-document relationships, (b) it is robust to retrieval noise, (c) it avoids catastrophic forgetting in the LLM, (d) it does not require access to the LLM's weights or output logits, enabling plug-and-play integration with both open and proprietary closed-source LLMs, and (e) its retrieval cost is independent of corpus size at inference time. Our experimental results on three benchmarks, BrowseComp-Plus, NarrativeQA, and MuSiQue, show that MeMo achieves strong performance compared to existing methods across diverse settings.

# link_2cws8vnjy4p.pdf ## Metadata - PDFFormatVersion=1.7 - IsLinearized=false - IsAcroFormPresent=false - IsXFAPresent=false - IsCollectionPresent=false - IsSignaturesPresent=false - Author=Ryan Wei Heng Quek; Sanghyuk Lee; Alfred Wei Lun Leong; Arun Verma; Alok Prakash; Nancy F. Chen; Bryan Kian Hsiang Low; Daniela Rus; Armando Solar-Lezama - Creator=arXiv GenPDF (tex2pdf:a6404ea) - Custom.DOI=https://doi.org/10.48550/arXiv.2605.15156 - Custom.License=http://creativecommons.org/licenses/by/4.0/ - Custom.PTEX.Fullbanner=This is pdfTeX, Version 3.141592653-2.6-1.40.28 (TeX Live 2025) kpathsea

saved by

• Christina Lee

related reading

• [2606.10346] Reasoning or Memorization? Direction-Aware Diversity Exploration in LLM Reinforcement Learningarxiv.org
• [2506.06266] Cartridges: Lightweight and general-purpose long context representations via self-studyarxiv.org
• [2005.11401] Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasksarxiv.org
• [2501.00663] Titans: Learning to Memorize at Test Timearxiv.org
• [2603.23516] MSA: Memory Sparse Attention for Efficient End-to-End Memory Model Scaling to 100M Tokensarxiv.org
• arxiv.org/pdf/2505.24832arxiv.org
• Self-Adapting Language Modelsarxiv.org
• [2304.13343] Unleashing Infinite-Length Input Capacity for Large-scale Language Models with Self-Controlled Memory Systemarxiv.org
• Unlocking the Working Memory of Large Language Models for Latent Reasoningarxiv.org
• [2604.18002] Neural Garbage Collection: Learning to Forget while Learning to Reasonarxiv.org
• [2605.15220] Always Learning, Always Mixing: Efficient and Simple Data Mixing All The Timearxiv.org
• GitHub - mem0ai/mem0: The Memory Layer for AI Agents - Drop-in memory infrastructure for AI agents and apps. Context that persists. Built for production.github.com