From Sufficiency to Reflection: Reinforcement-Guided Thinking Quality in Retrieval-Augmented Reasoning for LLMs
Abstract
AbstractReinforcement learning-based retrieval-augmented generation (RAG) methods enhance the reasoning abilities of large language models (LLMs). However, most rely only on final-answer rewards, overlooking intermediate reasoning quality. This paper analyzes existing RAG reasoning models and identifies three main failure patterns: (1) information insufficiency: failure to retrieve adequate support; (2) faulty reasoning: logical or content-level flaws despite sufficient information; (3) answer–reasoning inconsistency: a valid reasoning chain with a mismatched final answer. We propose TIRESRAG-R1, a novel framework using a think–retrieve–reflect process and a multi-dimensional reward system to improve reasoning and stability. TIRESRAG-R1 introduces: (1) a sufficiency reward to encourage thorough retrieval; (2) a reasoning quality reward to assess rationality and accuracy of the reasoning chain; (3) a reflection reward to detect and revise errors. It also employs a difficulty-aware reweighting strategy and training sample filtering to boost performance on complex tasks. Experiments on four multi-hop QA datasets show TIRESRAG-R1 outperforms prior RAG methods and generalizes well to single-hop tasks. Our code is available at https://github.com/probe2/TIRESRAG-R1.