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.
Continuous Optimization for Satisfiability Modulo Theories on Linear Real Arithmetic
Department of Electrical and Computer Engineering, National University of Singapore, Singapore(电子与计算机工程系,新加坡国立大学);School of Computing and Data Science, The University of Hong Kong, Hong Kong, China(计算与数据科学学院,香港大学)
Lazy Modeling of Variants of Token Swapping Problem and Multi-agent Path Finding through Combination of Satisfiability Modulo Theories and Conflict-based Search
Fondazione Bruno Kessler(布鲁诺·凯斯勒基金会);Università degli Studi di Trento(特伦托大学)
包含布尔和数值变量的混合问题的建模是人工智能的长期兴趣。然而,在混合域中执行推理和学习是一项特别艰巨的任务。对这种域建模的能力在“学习设计”任务中是至关重要的,也就是说,学习应用程序的目标是从示例中学习如何执行新对象的自动{ \em de novo}设计。在本文中,我们提出了结构化学习模理论,一个最大限度的学习方法在混合域的基础上可满足性模理论,它允许结合布尔推理和优化连续线性算术约束。其主要思想是利用最先进的广义可满足性模理论求解器来实现结构化输出支持向量机的推理和分离预言。我们验证我们的方法在人工和现实世界的情况下。
Incremental Satisfiability Modulo Theory for Verification of Deep Neural Networks
Institute of Software, Chinese Academy of Sciences(中国科学院软件研究所);University of Chinese Academy of Sciences(中国科学院大学);Nanjing Institute of Software Technology, ISCAS(中国科学院软件研究所南京软件技术研究院)
Learning Chordal Markov Networks by Constraint Satisfaction
University of Helsinki(赫尔辛基大学);Aalto University(阿尔托大学);Helsinki Institute of Information Technology(赫尔辛基信息技术研究所);Griffith University(格里菲斯大学);Åbo Akademi University(奥布学术大学)
Neuro Symbolic Reasoning for Planning: Counterexample Guided Inductive Synthesis using Large Language Models and Satisfiability Solving
Florida International University(佛罗里达国际大学);SRI International(SRI国际(斯坦福国际研究院));United States Military Academy(美国军事学院(西点军校));University of Colorado Boulder(科罗拉多大学博尔德分校);University of Central Florida(中佛罗里达大学);Vanderbilt University(范德堡大学)
Autoformalizing Natural Language to First-Order Logic: A Case Study in Logical Fallacy Detection
Stanford University(斯坦福大学);X, the moonshot factory(X moonshot工厂);Max Planck Institute for Intelligent Systems(马克斯·普朗克智能系统研究所);ETH Zürich(苏黎世联邦理工学院);University of Toronto(多伦多大学)