Litcius/Paper detail

Verisig 2.0: Verification of Neural Network Controllers Using Taylor Model Preconditioning

Radoslav Ivanov, Taylor J. Carpenter, James Weimer, Rajeev Alur, George J. Pappas, Insup Lee

2021Lecture notes in computer science57 citationsDOIOpen Access PDF

Abstract

Abstract This paper presents Verisig 2.0, a verification tool for closed-loop systems with neural network (NN) controllers. We focus on NNs with tanh/sigmoid activations and develop a Taylor-model-based reachability algorithm through Taylor model preconditioning and shrink wrapping. Furthermore, we provide a parallelized implementation that allows Verisig 2.0 to efficiently handle larger NNs than existing tools can. We provide an extensive evaluation over 10 benchmarks and compare Verisig 2.0 against three state-of-the-art verification tools. We show that Verisig 2.0 is both more accurate and faster, achieving speed-ups of up to 21x and 268x against different tools, respectively.

Topics & Concepts

Computer scienceReachabilitySigmoid functionArtificial neural networkFocus (optics)Taylor seriesState (computer science)Artificial intelligenceAlgorithmMathematicsMathematical analysisPhysicsOpticsAdversarial Robustness in Machine LearningModel Reduction and Neural NetworksAdvanced Neural Network Applications
Verisig 2.0: Verification of Neural Network Controllers Using Taylor Model Preconditioning | Litcius