arXiv cs.AI· Mingxi Zou, Yinghui Xu, Jiaxiang Chen, Langzhang Liang, Wei Zhu, Peng Ye, Runkai Zhao, Ruoxi Jiang, Sijin Chen, Song Wang, Zhuo Wang, Lizhen Qu, Zenglin Xu·· 1 小时前AI 评分43
MIGRO：面向语言模型的解耦可移植长期记忆系统
MIGRO: A Decoupled and Portable Long-Term Memory System for Language Models
AI 导读
arXiv 论文提出 MIGRO，一种将经验作为可跨模型共享持久状态的解耦长期记忆系统，遵循“精确保存历史证据、近似组织、问题已知时再决定检索”的原则。在匹配 LoCoMo 状态下，MIGRO 比最强质量基线高 7.3%，私有内存减少 99.3%；交接后续写可保留完整重建 99.9% 的质量；在百万 token BEAM 历史上平均比最紧凑基线少用 57.5% 私有存储。代码已开源。
正文 · 原文
Authors:Mingxi Zou, Yinghui Xu, Jiaxiang Chen, Langzhang Liang, Wei Zhu, Peng Ye, Runkai Zhao, Ruoxi Jiang, Sijin Chen, Song Wang, Zhuo Wang, Lizhen Qu, Zenglin Xu
Abstract:Language models evolve, but the observations, revisions, and context accumulated through user interactions should remain usable. Each new model must be able to retrieve past evidence, track revisions, and add new observations. We present MIGRO, a decoupled long-term memory system that treats experience as a persistent state shared across successive language models.
MIGRO follows a simple principle: preserve historical evidence exactly, organize it approximately, and decide what to retrieve only when the question is known. We call this asymmetric fidelity with deferred selectivity. A shared semantic index supports retrieval, while exact records retain observations and their historical relationships. Because models differ in vocabulary and tokenization, host-specific interfaces convert selected evidence into inputs the receiving model can use. Reversible encoding reduces archival storage while enabling exact recovery. The shared state and its access interfaces let successive hosts read and extend the accumulated experience.
On matched LoCoMo states, MIGRO scores 7.3% higher and uses 99.3% less private memory than the strongest quality baseline. After handoff, continued writing retains 99.9% of full-history rebuilding quality by extending the inherited state. On million-token BEAM histories, MIGRO uses 57.5% less private storage on average than the most compact baseline. MIGRO combines high answer quality, continued memory growth across hosts, and low storage cost.
Code is available at this https URL.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.15315 [cs.AI] |
| (or arXiv:2611.15315v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.15315 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Mingxi Zou [view email]
[v1]
Wed, 11 Nov 2026 10:52:21 UTC (1,833 KB)
来源：arXiv cs.AI · arxiv.org