arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.18095cs.AI
收藏

神经符号语言推理作为可满足性模理论

Neurosymbolic Language Reasoning as Satisfiability Modulo Theory

Seoul National University(首尔国立大学);U. Mass. Amherst(马萨诸塞大学阿姆赫斯特分校);Microsoft(微软公司)

Hyunseok Oh, Sam Stern, Youngki Lee, Matthai Philipose

AI总结:

Logitext通过结合LLM和SMT求解,实现自然语言与逻辑推理的联合处理,提升内容审核和法律任务的准确性和覆盖性。

摘要:

自然语言理解需要交织文本和逻辑推理,但大语言模型往往难以可靠地进行此类推理。现有的神经符号系统将LLM与求解器结合,但仍然局限于完全形式化的任务,如数学或程序合成,无法处理自然文档中仅有部分逻辑结构的情况。我们引入了Logitext,一种神经符号语言,将文档表示为自然语言文本约束(NLTCs),使部分逻辑结构显式化。我们开发了一种算法,将基于LLM的约束评估与可满足性模理论(SMT)求解相结合,实现文本-逻辑推理的联合推理。在新的内容审核基准测试、LegalBench和Super-Natural Instructions上的实验表明,Logitext在准确性和覆盖性上均有所提升。本工作是首次将基于LLM的推理视为SMT理论,将神经符号方法扩展到完全形式化领域之外。

英文摘要:

Natural language understanding requires interleaving textual and logical reasoning, yet large language models often fail to perform such reasoning reliably. Existing neurosymbolic systems combine LLMs with solvers but remain limited to fully formalizable tasks such as math or program synthesis, leaving natural documents with only partial logical structure unaddressed. We introduce Logitext, a neurosymbolic language that represents documents as natural language text constraints (NLTCs), making partial logical structure explicit. We develop an algorithm that integrates LLM-based constraint evaluation with satisfiability modulo theory (SMT) solving, enabling joint textual-logical reasoning. Experiments on a new content moderation benchmark, together with LegalBench and Super-Natural Instructions, show that Logitext improves both accuracy and coverage. This work is the first that treats LLM-based reasoning as an SMT theory, extending neurosymbolic methods beyond fully formalizable domains.

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。

↑