标题： KnowRL：探索事实性的知识强化学习 显示英文标题

标题： KnowRL: Exploring Knowledgeable Reinforcement Learning for Factuality

摘要： 大型语言模型（LLMs），特别是慢思考模型，经常表现出严重的幻觉，由于在推理过程中无法准确识别知识边界，导致输出错误内容。 虽然 强化学习（RL）可以增强复杂的推理能力，但其结果导向的奖励机制通常缺乏对思考过程的事实监督，进一步加剧了幻觉问题。 为了解决慢思考模型中的高幻觉问题，我们提出了知识增强的强化学习，KnowRL。 KnowRL通过将基于知识验证的事实奖励整合到RL训练过程中，引导模型进行基于事实的慢思考，帮助它们识别自己的知识边界。 KnowRL通过将基于知识验证的事实奖励整合到RL训练过程中，引导模型进行基于事实的慢思考，帮助它们识别自己的知识边界。 在 RL训练期间的这种针对性事实输入使模型能够学习并内化基于事实的推理策略。 通过直接奖励推理步骤中遵守事实的行为，KnowRL促进了更可靠的思考过程。 在三个幻觉评估数据集和两个推理评估数据集上的实验结果表明，KnowRL有效缓解了慢思考模型中的幻觉问题，同时保持了其原有的强大推理能力。 我们的代码可在 https://github.com/zjunlp/KnowRL 获取。 显示英文摘要

摘要： Large Language Models (LLMs), particularly slow-thinking models, often exhibit severe hallucination, outputting incorrect content due to an inability to accurately recognize knowledge boundaries during reasoning. While Reinforcement Learning (RL) can enhance complex reasoning abilities, its outcome-oriented reward mechanism often lacks factual supervision over the thinking process, further exacerbating the hallucination problem. To address the high hallucination in slow-thinking models, we propose Knowledge-enhanced RL, KnowRL. KnowRL guides models to perform fact-based slow thinking by integrating a factuality reward, based on knowledge verification, into the RL training process, helping them recognize their knowledge boundaries. KnowRL guides models to perform fact-based slow thinking by integrating a factuality reward, based on knowledge verification, into the RL training process, helping them recognize their knowledge boundaries. This targeted factual input during RL training enables the model to learn and internalize fact-based reasoning strategies. By directly rewarding adherence to facts within the reasoning steps, KnowRL fosters a more reliable thinking process. Experimental results on three hallucination evaluation datasets and two reasoning evaluation datasets demonstrate that KnowRL effectively mitigates hallucinations in slow-thinking models while maintaining their original strong reasoning capabilities. Our code is available at https://github.com/zjunlp/KnowRL.