AssistRAG: Boosting the Potential of Large Language Models with an Intelligent Information Assistant

TL;DR

AssistRAG introduces an intelligent assistant within LLMs that manages knowledge and reasoning through a two-phase training process, significantly improving factual accuracy and complex reasoning over previous retrieval-augmented methods.

Contribution

The paper presents AssistRAG, a novel framework that integrates an intelligent assistant into LLMs, enhancing retrieval and reasoning without extensive retraining.

Findings

Outperforms existing benchmarks in accuracy and reasoning.

Benefits less advanced LLMs significantly.

Enhances factual correctness and decision-making.

Abstract

The emergence of Large Language Models (LLMs) has significantly advanced natural language processing, but these models often generate factually incorrect information, known as "hallucination". Initial retrieval-augmented generation (RAG) methods like the "Retrieve-Read" framework was inadequate for complex reasoning tasks. Subsequent prompt-based RAG strategies and Supervised Fine-Tuning (SFT) methods improved performance but required frequent retraining and risked altering foundational LLM capabilities. To cope with these challenges, we propose Assistant-based Retrieval-Augmented Generation (AssistRAG), integrating an intelligent information assistant within LLMs. This assistant manages memory and knowledge through tool usage, action execution, memory building, and plan specification. Using a two-phase training approach, Curriculum Assistant Learning and Reinforced Preference Optimization. AssistRAG enhances information retrieval and decision-making. Experiments show AssistRAG significantly outperforms benchmarks, especially benefiting less advanced LLMs, by providing superior reasoning capabilities and accurate responses.