RAG-Star: Enhancing Deliberative Reasoning with Retrieval Augmented Verification and Refinement

TL;DR

RAG-Star introduces a retrieval-augmented, tree-based reasoning framework that combines external retrieval with internal LLM knowledge, significantly improving complex reasoning performance using Monte Carlo Tree Search and verification feedback.

Contribution

It presents RAG-Star, a novel approach integrating retrieval-augmented verification with deliberative reasoning, enhancing LLMs' ability to handle complex tasks beyond simple reasoning steps.

Findings

RAG-Star outperforms previous RAG and reasoning methods.

It effectively combines external retrieval with internal LLM knowledge.

Demonstrates significant improvements on Llama-3.1-8B-Instruct and GPT-4o.

Abstract

Existing large language models (LLMs) show exceptional problem-solving capabilities but might struggle with complex reasoning tasks. Despite the successes of chain-of-thought and tree-based search methods, they mainly depend on the internal knowledge of LLMs to search over intermediate reasoning steps, limited to dealing with simple tasks involving fewer reasoning steps. In this paper, we propose \textbf{RAG-Star}, a novel RAG approach that integrates the retrieved information to guide the tree-based deliberative reasoning process that relies on the inherent knowledge of LLMs. By leveraging Monte Carlo Tree Search, RAG-Star iteratively plans intermediate sub-queries and answers for reasoning based on the LLM itself. To consolidate internal and external knowledge, we propose an retrieval-augmented verification that utilizes query- and answer-aware reward modeling to provide feedback for the inherent reasoning of LLMs. Our experiments involving Llama-3.1-8B-Instruct and GPT-4o demonstrate that RAG-Star significantly outperforms previous RAG and reasoning methods.