CheckRLM: Effective Knowledge-Thought Coherence Checking in Retrieval-Augmented Reasoning
For users of reasoning language models, CheckRLM reduces factual errors in long-horizon reasoning, but the improvement is incremental over existing RAG methods.
CheckRLM improves reasoning reliability in RLMs by detecting and correcting factual errors in reasoning chains using RAG, outperforming baselines with lower costs.
Reasoning Language Models (RLMs) have significantly improved performance on complex tasks by extending the reasoning chain. However, these chains are prone to containing factual errors, particularly in knowledge-intensive tasks. To address this issue, we propose CheckRLM, a framework that improves the reliability of the reasoning process through Retrieval-Augmented Generation (RAG) by timely checking and correcting factual errors. Specifically, CheckRLM extracts factual claims from the reasoning chain to identify and localize subtle knowledge inconsistencies during inference. Upon detection of errors, a refinement mechanism performs minimal-cost yet precise corrections by leveraging external knowledge, ensuring coherence between the reasoning chain and correct knowledge. Extensive experiments demonstrate that CheckRLM substantially outperforms existing baselines, exhibiting a strong capability to mitigate error accumulation in long-horizon reasoning with lower costs. The code and data are available at https://github.com/AI9Stars/CheckRLM.