Neural Approaches to SAT Solving: Design Choices and Interpretability

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Mojžíšek, David, Hůla, Jan, Li, Ziwei, Zhou, Ziyu, Janota, Mikoláš
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916671055200256
author Mojžíšek, David
Hůla, Jan
Li, Ziwei
Zhou, Ziyu
Janota, Mikoláš
author_facet Mojžíšek, David
Hůla, Jan
Li, Ziwei
Zhou, Ziyu
Janota, Mikoláš
contents In this contribution, we provide a comprehensive evaluation of graph neural networks applied to Boolean satisfiability problems, accompanied by an intuitive explanation of the mechanisms enabling the model to generalize to different instances. We introduce several training improvements, particularly a novel closest assignment supervision method that dynamically adapts to the model's current state, significantly enhancing performance on problems with larger solution spaces. Our experiments demonstrate the suitability of variable-clause graph representations with recurrent neural network updates, which achieve good accuracy on SAT assignment prediction while reducing computational demands. We extend the base graph neural network into a diffusion model that facilitates incremental sampling and can be effectively combined with classical techniques like unit propagation. Through analysis of embedding space patterns and optimization trajectories, we show how these networks implicitly perform a process very similar to continuous relaxations of MaxSAT, offering an interpretable view of their reasoning process. This understanding guides our design choices and explains the ability of recurrent architectures to scale effectively at inference time beyond their training distribution, which we demonstrate with test-time scaling experiments.
format Preprint
id arxiv_https___arxiv_org_abs_2504_01173
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Neural Approaches to SAT Solving: Design Choices and Interpretability
Mojžíšek, David
Hůla, Jan
Li, Ziwei
Zhou, Ziyu
Janota, Mikoláš
Machine Learning
Artificial Intelligence
In this contribution, we provide a comprehensive evaluation of graph neural networks applied to Boolean satisfiability problems, accompanied by an intuitive explanation of the mechanisms enabling the model to generalize to different instances. We introduce several training improvements, particularly a novel closest assignment supervision method that dynamically adapts to the model's current state, significantly enhancing performance on problems with larger solution spaces. Our experiments demonstrate the suitability of variable-clause graph representations with recurrent neural network updates, which achieve good accuracy on SAT assignment prediction while reducing computational demands. We extend the base graph neural network into a diffusion model that facilitates incremental sampling and can be effectively combined with classical techniques like unit propagation. Through analysis of embedding space patterns and optimization trajectories, we show how these networks implicitly perform a process very similar to continuous relaxations of MaxSAT, offering an interpretable view of their reasoning process. This understanding guides our design choices and explains the ability of recurrent architectures to scale effectively at inference time beyond their training distribution, which we demonstrate with test-time scaling experiments.
title Neural Approaches to SAT Solving: Design Choices and Interpretability
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2504.01173