Teaching Temporal Logics to Neural Networks
This addresses the challenge of neuro-symbolic computing for verification tasks, but is incremental as it builds on existing methods with new data.
The authors tackled the problem of whether neural networks can learn to solve linear-time temporal logic (LTL) formulas end-to-end, and demonstrated that Transformers trained on data from classical solvers can predict correct solutions, even for formulas where solvers timed out, achieving correct solutions for most formulas.
We study two fundamental questions in neuro-symbolic computing: can deep learning tackle challenging problems in logics end-to-end, and can neural networks learn the semantics of logics. In this work we focus on linear-time temporal logic (LTL), as it is widely used in verification. We train a Transformer on the problem to directly predict a solution, i.e. a trace, to a given LTL formula. The training data is generated with classical solvers, which, however, only provide one of many possible solutions to each formula. We demonstrate that it is sufficient to train on those particular solutions to formulas, and that Transformers can predict solutions even to formulas from benchmarks from the literature on which the classical solver timed out. Transformers also generalize to the semantics of the logics: while they often deviate from the solutions found by the classical solvers, they still predict correct solutions to most formulas.