Current Practices for Building LLM-Powered Reasoning Tools Are Ad Hoc -- and We Can Do Better

2025-07-08

AI Summary

Existing couplings between large language models (LLMs) and symbolic algorithms are largely ad hoc, undermining the formal guarantees of symbolic methods while constraining the expressive power of neural reasoning. Method: We propose the Neuro-Symbolic Transition System (NSTS)—a principled computational model that jointly and concurrently models symbolic state spaces and LLM-generated intuitive representations within a unified framework, enabling their synchronous evolution. NSTS enforces strong semantic guarantees via formally verifiable transition rules and supports scalable inference architectures. We implement a prototype in a logic programming language, achieving deep integration of LLMs with classical symbolic algorithms. Contribution/Results: This work establishes the first infrastructure framework for LLM-driven automated reasoning that simultaneously ensures theoretical rigor, formal verifiability, and practical utility. Empirical evaluation demonstrates significant promise in software verification and program synthesis tasks.

Technology Category

• Machine Learning: Neuro-Symbolic Learning
• Cognitive Modeling & Cognitive Systems: Symbolic Representations
• Knowledge Representation and Reasoning: Automated Reasoning and Theorem Proving

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

There is growing excitement about building software verifiers, synthesizers, and other Automated Reasoning (AR) tools by combining traditional symbolic algorithms and Large Language Models (LLMs). Unfortunately, the current practice for constructing such neurosymbolic AR systems is an ad hoc programming model that does not have the strong guarantees of traditional symbolic algorithms, nor a deep enough synchronization of neural networks and symbolic reasoning to unlock the full potential of LLM-powered reasoning. I propose Neurosymbolic Transition Systems as a principled computational model that can underlie infrastructure for building neurosymbolic AR tools. In this model, symbolic state is paired with intuition, and state transitions operate over symbols and intuition in parallel. I argue why this new paradigm can scale logical reasoning beyond current capabilities while retaining the strong guarantees of symbolic algorithms, and I sketch out how the computational model I propose can be reified in a logic programming language.

Problem

• Current neurosymbolic AR systems lack strong guarantees and synchronization
• Ad hoc programming limits potential of LLM-powered reasoning tools
• Need principled model to scale logical reasoning with symbolic guarantees

Innovation

• Neurosymbolic Transition Systems for principled computation
• Parallel state transitions on symbols and intuition
• Logic programming language reification of model