Larimar: Large Language Models with Episodic Memory Control
Payel Das Subhajit Chaudhury Elliot Nelson Igor Melnyk Sarathkrishna Swaminathan Sihui Dai Aurélie Lozano Georgios Kollias Vijil Chenthamarakshan Jiří Navrátil Soham Dan Pin-Yu Chen Payel Das*, Subhajit Chaudhury*, Elliot Nelson, Igor Melnyk, Sarath Swaminathan, Sihui Dai, Aurélie Lozano, Georgios Kollias, Vijil Chenthamarakshan, Jiří Navrátil, Soham Dan, Pin-Yu Chen
Abstract
Efficient and accurate updating of knowledge stored in Large Language Models (LLMs) is one of the most pressing research challenges today. This paper presents Larimar - a novel, brain-inspired architecture for enhancing LLMs with a distributed episodic memory. Larimar’s memory allows for dynamic, one-shot updates of knowledge without the need for computationally expensive re-training or fine-tuning. Experimental results on multiple fact editing benchmarks demonstrate that Larimar attains accuracy comparable to most competitive baselines, even in the challenging sequential editing setup, but also excels in speed—yielding speed-ups of 4-10x depending on the base LLM —as well as flexibility due to the proposed architecture being simple, LLM-agnostic, and hence general. We further provide mechanisms for selective fact forgetting and input context length generalization with Larimar and show their effectiveness.
中文速览
大语言模型（LLM）部署后，内部存储的知识会随时间过时或出错，但用传统微调方式纠正既慢又容易损坏其他已有知识。受大脑海马体"快速单次记忆"机制的启发，研究者提出了 Larimar——一种为 LLM 外挂分布式情节记忆（episodic memory）模块的架构，新知识只需一次性写入记忆矩阵（通过求解线性方程组的伪逆），无需对 LLM 本身做任何梯度更新，随后该记忆作为条件信号引导 LLM 解码器输出正确答案。在多个事实编辑基准测试中，Larimar 的准确率与当前最强基线相当，但速度快 4–10 倍，同时还天然支持"选择性遗忘"特定事实和长上下文泛化等功能。这项工作表明，借鉴脑科学的互补学习机制可以让知识更新变得又快又灵活，为在真实工业场景中持续维护 LLM 的准确性提供了一条切实可行的路径。
原文 arXiv:2403.11901;中英对照 + 大白话阅读 https://aha.fim.ai/paper/2403.11901v4