Predicting Propositional Satisfiability Based on Graph Attention Networks

Abstract Boolean satisfiability problems (SAT) have very rich generic and domain-specific structures. How to capture these structural features in the embedding space and feed them to deep learning models is an important factor influencing the use of neural networks to solve SAT problems. Graph neura...

Full description

Bibliographic Details

Published in:	International Journal of Computational Intelligence Systems
Main Authors:	Wenjing Chang, Hengkai Zhang, Junwei Luo
Format:	Article
Language:	English
Published:	Springer 2022-09-01
Subjects:	Random SAT problem Propositional satisfiability Satisfiability features Deep learning Graph attention networks
Online Access:	https://doi.org/10.1007/s44196-022-00139-9