MemGPT: Towards LLMs as Operating Systems

TL;DR

MemGPT introduces a hierarchical memory management system inspired by operating systems to extend the effective context window of large language models, enabling better document analysis and multi-session chat capabilities.

Contribution

The paper presents MemGPT, a novel system that manages multiple memory tiers to extend LLM context windows using OS-inspired techniques, allowing for long-term interactions and large document processing.

Findings

MemGPT successfully analyzes documents larger than the LLM's context window.

It enables multi-session chat with long-term memory and dynamic reflection.

The system improves performance in tasks constrained by limited context size.

Abstract

Large language models (LLMs) have revolutionized AI, but are constrained by limited context windows, hindering their utility in tasks like extended conversations and document analysis. To enable using context beyond limited context windows, we propose virtual context management, a technique drawing inspiration from hierarchical memory systems in traditional operating systems that provide the appearance of large memory resources through data movement between fast and slow memory. Using this technique, we introduce MemGPT (Memory-GPT), a system that intelligently manages different memory tiers in order to effectively provide extended context within the LLM's limited context window, and utilizes interrupts to manage control flow between itself and the user. We evaluate our OS-inspired design in two domains where the limited context windows of modern LLMs severely handicaps their performance: document analysis, where MemGPT is able to analyze large documents that far exceed the underlying LLM's context window, and multi-session chat, where MemGPT can create conversational agents that remember, reflect, and evolve dynamically through long-term interactions with their users. We release MemGPT code and data for our experiments at https://memgpt.ai.