Prompting large language models has enabled significant recent progress in multi-step reasoning over text. However, when applied to text generation from semi-structured data (e.g., graphs or tables), these methods typically suffer from low semantic coverage, hallucination, and logical inconsistency. We propose MURMUR, a neuro-symbolic modular approach to text generation from semi-structured data with multi-step reasoning. MURMUR is a best-first search method that generates reasoning paths using: (1) neural and symbolic modules with specific linguistic and logical skills, (2) a grammar whose production rules define valid compositions of modules, and (3) value functions that assess the quality of each reasoning step. We conduct experiments on two diverse data-to-text generation tasks like WebNLG and LogicNLG. These tasks differ in their data representations (graphs and tables) and span multiple linguistic and logical skills. MURMUR obtains significant improvements over recent few-shot baselines like direct prompting and chain-of-thought prompting, while also achieving comparable performance to fine-tuned GPT-2 on out-of-domain data. Moreover, human evaluation shows that MURMUR generates highly faithful and correct reasoning paths that lead to 26% more logically consistent summaries on LogicNLG, compared to direct prompting.
翻译:推动使用大型语言模型使得在文本的多步推理方面取得了近期的重大进展。然而,当应用到半结构数据(如图表或表格)的文本生成时,这些方法通常会受到低语义覆盖、幻觉和逻辑不一致的影响。我们建议采用神经-同步模块法MUR,即神经-同步模块法,用多步推理从半结构数据生成文本。MURMUR是一种最佳第一搜索方法,它生成推理路径,使用的方法有:(1) 具有特定语言和逻辑技能的神经和象征性模块,(2) 其制作规则界定模块有效构成的语法,(3) 评估每个推理步骤质量的数值功能。我们实验了两种不同的数据-文字生成任务,如WebNLG和LogicNLG。这些任务在数据表述(图表和表格)方面各不相同,而且涉及多种语言和逻辑技能。MUR是一种比最近几分明的基线(如直接提示和思维链导)得到重大改进,同时取得可比的性性性性业绩,以精确的GPT-2-N-do-d-domain 直接推算出26的逻辑推算。