AirRAG: Activating Intrinsic Reasoning for Retrieval Augmented Generation using Tree-based Search

ACL ARR 2025 February Submission867 Authors

Abstract:

Leveraging the autonomous decision-making capabilities of large language models (LLMs) has demonstrated superior performance in reasoning tasks. However, despite the success of iterative or recursive retrieval-augmented generation (RAG) techniques, these methods are often constrained to a single solution space when confronted with complex problems.
In this paper, we propose a novel thinking pattern in RAG that integrates system analysis with efficient reasoning actions, significantly \textbf{A}ctivating \textbf{i}ntrinsic \textbf{r}easoning capabilities and expanding the solution space of specific tasks via Monte Carlo Tree Search (MCTS), which we refer to as \textbf{AirRAG}. Specifically, our approach designs five fundamental reasoning actions, which are expanded to a broad tree-based reasoning space using MCTS.
The approach also incorporates self-consistency verification to explore potential reasoning paths and inference scaling law. Additionally, computationally optimal strategies are employed to allocate more inference resources to key actions, thereby enhancing overall performance. Experimental results demonstrate the effectiveness of AirRAG, showing significant performance gains on complex question-answering datasets. Furthermore, AirRAG is flexible and lightweight, making it easy to integrate with other advanced technologies.

Paper Type: Long

Research Area: Generation

Research Area Keywords: Generation, Question Answering, Language Modeling, NLP Applications

Contribution Types: Position papers

Languages Studied: English

Submission Number: 867