Crafting Personalized Agents through Retrieval-Augmented Generation on Editable Memory Graphs

TL;DR

This paper presents EMG-RAG, a novel method combining retrieval-augmented generation with editable memory graphs and reinforcement learning to create personalized AI agents that utilize smartphone memories, improving application performance and usability.

Contribution

Introduction of EMG-RAG, a new framework that integrates retrieval-augmented generation with editable memory graphs and reinforcement learning for personalized agent creation.

Findings

Achieved about 10% improvement over existing methods.

Validated effectiveness on real-world smartphone data.

Enhanced usability of AI assistants in practical deployment.

Abstract

In the age of mobile internet, user data, often referred to as memories, is continuously generated on personal devices. Effectively managing and utilizing this data to deliver services to users is a compelling research topic. In this paper, we introduce a novel task of crafting personalized agents powered by large language models (LLMs), which utilize a user's smartphone memories to enhance downstream applications with advanced LLM capabilities. To achieve this goal, we introduce EMG-RAG, a solution that combines Retrieval-Augmented Generation (RAG) techniques with an Editable Memory Graph (EMG). This approach is further optimized using Reinforcement Learning to address three distinct challenges: data collection, editability, and selectability. Extensive experiments on a real-world dataset validate the effectiveness of EMG-RAG, achieving an improvement of approximately 10% over the best existing approach. Additionally, the personalized agents have been transferred into a real smartphone AI assistant, which leads to enhanced usability.