SEER: Facilitating Structured Reasoning and Explanation via Reinforcement Learning

arXiv:2401.13246v2 Announce Type: replace
Abstract: Elucidating the reasoning process with structured explanations from question to answer is crucial, as it significantly enhances the interpretability, traceability, and trustworthiness of question-answering (QA) systems. However, structured explanations demand models to perform intricately structured reasoning, which poses great challenges. Most existing methods focus on single-step reasoning through supervised learning, ignoring logical dependencies between steps. Moreover, existing reinforcement learning (RL) based methods overlook the structured relationships, underutilizing the potential of RL in structured reasoning. In …

abstract arxiv challenges cs.cl demand focus interpretability process question reasoning reinforcement reinforcement learning seer systems traceability type via