1 code implementation • 17 Dec 2023 • Yuhao Zhou, Stavros Tripakis
Verifying the inductiveness of a candidate inductive invariant in the context of NNCS is hard because of the scale and nonlinearity of neural networks.
1 code implementation • 1 Jun 2023 • Rômulo Meira-Góes, Ian Dardik, Eunsuk Kang, Stéphane Lafortune, Stavros Tripakis
This paper proposes a notion of \emph{robustness} as an explicit, first-class property of a transition system that captures how robust it is against possible \emph{deviations} in the environment.
no code implementations • 22 Mar 2022 • Charis Eleftheriadis, Nikolaos Kekatos, Panagiotis Katsaros, Stavros Tripakis
Two pretrained neural networks are deemed equivalent if they yield similar outputs for the same inputs.
no code implementations • 8 Oct 2021 • Rômulo Meira-Góes, Eunsuk Kang, Stéphane Lafortune, Stavros Tripakis
In this paper, we propose an approach for analyzing control systems with respect to their tolerance against environmental perturbations.
1 code implementation • 5 Oct 2021 • Lisa Oakley, Alina Oprea, Stavros Tripakis
We outline a class of threat models under which adversaries can perturb system transitions, constrained by an $\varepsilon$ ball around the original transition probabilities.
3 code implementations • 2 Apr 2020 • Max von Hippel, Cole Vick, Stavros Tripakis, Cristina Nita-Rotaru
Distributed protocols should be robust to both benign malfunction (e. g. packet loss or delay) and attacks (e. g. message replay) from internal or external adversaries.
Cryptography and Security Formal Languages and Automata Theory
1 code implementation • 4 Mar 2020 • Igor Buzhinsky, Arseny Nerinovsky, Stavros Tripakis
In this paper, we propose several metrics to measure robustness of classifiers to natural adversarial examples, and methods to evaluate them.
1 code implementation • 10 May 2017 • Eunsuk Kang, Stephane Lafortune, Stavros Tripakis
In this paper, we introduce the problem of synthesizing a property-preserving platform mapping: A set of implementation decisions ensuring that a desired property is preserved from a high-level design into a low-level platform implementation.
Software Engineering
no code implementations • 25 May 2016 • Georgios Giantamidis, Stavros Tripakis
We develop three algorithms for solving this problem: (1) the PTAP algorithm, which transforms a set of input-output traces into an incomplete Moore machine and then completes the machine with self-loops; (2) the PRPNI algorithm, which uses the well-known RPNI algorithm for automata learning to learn a product of automata encoding a Moore machine; and (3) the MooreMI algorithm, which directly learns a Moore machine using PTAP extended with state merging.