论文信息 - Neural Guidance for SAT Solving

Neural Guidance for SAT Solving

We use neural guidance to direct search of the DPLL algorithm. We compare SATsolving performance of various heuristics and two neural architectures: LSTM and a message-passing architecture. By a large margin the best one is the message passing architecture, which has more desirable theoretical properties and which is capable of solving complicated instances of SAT problems even when used with a naive implementation of the DPLL algorithm.

[1] David L. Dill,et al. Learning a SAT Solver from Single-Bit Supervision , 2018, ICLR.

[2] Joao Marques-Silva,et al. PySAT: A Python Toolkit for Prototyping with SAT Oracles , 2018, SAT.

[3] Taro Sekiyama,et al. Automated proof synthesis for propositional logic with deep neural networks , 2018, ArXiv.

[4] Cezary Kaliszyk,et al. Reinforcement Learning of Theorem Proving , 2018, NeurIPS.

[5] Richard Evans,et al. Can Neural Networks Understand Logical Entailment? , 2018, ICLR.

[6] Josef Urban,et al. ATPboost: Learning Premise Selection in Binary Setting with ATP Feedback , 2018, IJCAR.

[7] Cezary Kaliszyk,et al. Deep Network Guided Proof Search , 2017, LPAR.

[8] Pushmeet Kohli,et al. Learning Continuous Semantic Representations of Symbolic Expressions , 2016, ICML.

[9] Andrei Voronkov,et al. First-Order Theorem Proving and Vampire , 2013, CAV.

[10] Gilles Audemard,et al. Predicting Learnt Clauses Quality in Modern SAT Solvers , 2009, IJCAI.

[11] Jens Otten,et al. leanCoP 2.0and ileanCoP 1.2: High Performance Lean Theorem Proving in Classical and Intuitionistic Logic (System Descriptions) , 2008, IJCAR.

[12] Niklas Sörensson,et al. An Extensible SAT-solver , 2003, SAT.

[13] J. A. Robinson,et al. Review: Martin Davis, George Logemann, Donald Loveland, A Machine Program for Theorem-Proving , 1967 .

[14] Donald W. Loveland,et al. A machine program for theorem-proving , 2011, CACM.