_version_ 1866911298476834816
author Li, Zhiyu
Xi, Chenyang
Li, Chunyu
Chen, Ding
Chen, Boyu
Song, Shichao
Niu, Simin
Wang, Hanyu
Yang, Jiawei
Tang, Chen
Yu, Qingchen
Zhao, Jihao
Wang, Yezhaohui
Liu, Peng
Lin, Zehao
Wang, Pengyuan
Huo, Jiahao
Chen, Tianyi
Chen, Kai
Li, Kehang
Tao, Zhen
Lai, Huayi
Wu, Hao
Tang, Bo
Wang, Zhengren
Fan, Zhaoxin
Zhang, Ningyu
Zhang, Linfeng
Yan, Junchi
Yang, Mingchuan
Xu, Tong
Xu, Wei
Chen, Huajun
Wang, Haofen
Yang, Hongkang
Zhang, Wentao
Xu, Zhi-Qin John
Chen, Siheng
Xiong, Feiyu
author_facet Li, Zhiyu
Xi, Chenyang
Li, Chunyu
Chen, Ding
Chen, Boyu
Song, Shichao
Niu, Simin
Wang, Hanyu
Yang, Jiawei
Tang, Chen
Yu, Qingchen
Zhao, Jihao
Wang, Yezhaohui
Liu, Peng
Lin, Zehao
Wang, Pengyuan
Huo, Jiahao
Chen, Tianyi
Chen, Kai
Li, Kehang
Tao, Zhen
Lai, Huayi
Wu, Hao
Tang, Bo
Wang, Zhengren
Fan, Zhaoxin
Zhang, Ningyu
Zhang, Linfeng
Yan, Junchi
Yang, Mingchuan
Xu, Tong
Xu, Wei
Chen, Huajun
Wang, Haofen
Yang, Hongkang
Zhang, Wentao
Xu, Zhi-Qin John
Chen, Siheng
Xiong, Feiyu
contents Large Language Models (LLMs) have become an essential infrastructure for Artificial General Intelligence (AGI), yet their lack of well-defined memory management systems hinders the development of long-context reasoning, continual personalization, and knowledge consistency.Existing models mainly rely on static parameters and short-lived contextual states, limiting their ability to track user preferences or update knowledge over extended periods.While Retrieval-Augmented Generation (RAG) introduces external knowledge in plain text, it remains a stateless workaround without lifecycle control or integration with persistent representations.Recent work has modeled the training and inference cost of LLMs from a memory hierarchy perspective, showing that introducing an explicit memory layer between parameter memory and external retrieval can substantially reduce these costs by externalizing specific knowledge. Beyond computational efficiency, LLMs face broader challenges arising from how information is distributed over time and context, requiring systems capable of managing heterogeneous knowledge spanning different temporal scales and sources. To address this challenge, we propose MemOS, a memory operating system that treats memory as a manageable system resource. It unifies the representation, scheduling, and evolution of plaintext, activation-based, and parameter-level memories, enabling cost-efficient storage and retrieval. As the basic unit, a MemCube encapsulates both memory content and metadata such as provenance and versioning. MemCubes can be composed, migrated, and fused over time, enabling flexible transitions between memory types and bridging retrieval with parameter-based learning. MemOS establishes a memory-centric system framework that brings controllability, plasticity, and evolvability to LLMs, laying the foundation for continual learning and personalized modeling.
format Preprint
id arxiv_https___arxiv_org_abs_2507_03724
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MemOS: A Memory OS for AI System
Li, Zhiyu
Xi, Chenyang
Li, Chunyu
Chen, Ding
Chen, Boyu
Song, Shichao
Niu, Simin
Wang, Hanyu
Yang, Jiawei
Tang, Chen
Yu, Qingchen
Zhao, Jihao
Wang, Yezhaohui
Liu, Peng
Lin, Zehao
Wang, Pengyuan
Huo, Jiahao
Chen, Tianyi
Chen, Kai
Li, Kehang
Tao, Zhen
Lai, Huayi
Wu, Hao
Tang, Bo
Wang, Zhengren
Fan, Zhaoxin
Zhang, Ningyu
Zhang, Linfeng
Yan, Junchi
Yang, Mingchuan
Xu, Tong
Xu, Wei
Chen, Huajun
Wang, Haofen
Yang, Hongkang
Zhang, Wentao
Xu, Zhi-Qin John
Chen, Siheng
Xiong, Feiyu
Computation and Language
Large Language Models (LLMs) have become an essential infrastructure for Artificial General Intelligence (AGI), yet their lack of well-defined memory management systems hinders the development of long-context reasoning, continual personalization, and knowledge consistency.Existing models mainly rely on static parameters and short-lived contextual states, limiting their ability to track user preferences or update knowledge over extended periods.While Retrieval-Augmented Generation (RAG) introduces external knowledge in plain text, it remains a stateless workaround without lifecycle control or integration with persistent representations.Recent work has modeled the training and inference cost of LLMs from a memory hierarchy perspective, showing that introducing an explicit memory layer between parameter memory and external retrieval can substantially reduce these costs by externalizing specific knowledge. Beyond computational efficiency, LLMs face broader challenges arising from how information is distributed over time and context, requiring systems capable of managing heterogeneous knowledge spanning different temporal scales and sources. To address this challenge, we propose MemOS, a memory operating system that treats memory as a manageable system resource. It unifies the representation, scheduling, and evolution of plaintext, activation-based, and parameter-level memories, enabling cost-efficient storage and retrieval. As the basic unit, a MemCube encapsulates both memory content and metadata such as provenance and versioning. MemCubes can be composed, migrated, and fused over time, enabling flexible transitions between memory types and bridging retrieval with parameter-based learning. MemOS establishes a memory-centric system framework that brings controllability, plasticity, and evolvability to LLMs, laying the foundation for continual learning and personalized modeling.
title MemOS: A Memory OS for AI System
topic Computation and Language
url https://arxiv.org/abs/2507.03724