FactAlign：大语言模型的长文本事实性对齐
FactAlign: Long-form Factuality Alignment of Large Language Models
AI总结：
本文提出 FactAlign 框架及句子级 fKTO 算法，利用细粒度事实性评估对齐大语言模型，在开放域与信息寻求任务中同时提升长回答的事实准确性、有用性和事实性 F1。
摘要：
大语言模型已展现出作为下一代信息访问引擎的巨大潜力。然而，幻觉以及生成非事实性内容的问题阻碍了其可靠性。这一问题在长文本回答中尤为突出，因为在这类回答中评估并确保事实准确性十分复杂。本文中，我们通过提出 FactAlign 来弥补这一空白，这是一种新颖的对齐框架，旨在提升大语言模型长文本回答的事实性，同时保持其有用性。我们引入了 fKTO，这是一种细粒度、句子级的对齐算法，扩展了 Kahneman-Tversky Optimization（KTO）对齐方法。FactAlign 利用自动事实性评估的最新进展，采用细粒度事实性评估来指导对齐过程。我们在开放域提示和信息寻求问题上的实验表明，FactAlign 显著提高了大语言模型回答的事实准确性，同时也提升了其有用性。进一步分析发现，FactAlign 能够训练大语言模型在不损失事实精确性的情况下提供更多信息，从而提高事实性 F1 分数。我们的源代码、数据集和训练后的模型已公开发布于 https://github.com/MiuLab/FactAlign。
英文摘要：
Large language models have demonstrated significant potential as the next-generation information access engines. However, their reliability is hindered by issues of hallucination and generating non-factual content. This is particularly problematic in long-form responses, where assessing and ensuring factual accuracy is complex. In this paper, we address this gap by proposing FactAlign, a novel alignment framework designed to enhance the factuality of LLMs' long-form responses while maintaining their helpfulness. We introduce fKTO, a fine-grained, sentence-level alignment algorithm that extends the Kahneman-Tversky Optimization (KTO) alignment method. Leveraging recent advances in automatic factuality evaluation, FactAlign utilizes fine-grained factuality assessments to guide the alignment process. Our experiments on open-domain prompts and information-seeking questions demonstrate that FactAlign significantly improves the factual accuracy of LLM responses while also improving their helpfulness. Further analyses identify that FactAlign is capable of training LLMs to provide more information without losing factual precision, thus improving the factual F1 score. Our source code, datasets, and trained models are publicly available at https://github.com/MiuLab/FactAlign
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
