PRESTO: A Processing-in-Memory-Based k-SAT Solver Using Recurrent Stochastic Neural Network With Unsupervised Learning.
Daehyun KimNael Mizanur RahmanSaibal MukhopadhyayPublished in: IEEE J. Solid State Circuits (2024)
Keyphrases
- sat solvers
- unsupervised learning
- neural network
- recurrent neural networks
- sat problem
- orders of magnitude
- sat solving
- feed forward
- propositional satisfiability
- back propagation
- search tree
- max sat
- constraint satisfaction
- boolean satisfiability
- supervised learning
- search space
- constraint solver
- boolean formula
- search strategies
- sat instances
- semi supervised
- reinforcement learning
- feature selection