Communication Dans Un Congrès Année : 2025

Verifying neural networks with PyRAT

Résumé

We present PyRAT, a tool based on abstract interpretation to verify the safety and robustness of neural networks. PyRAT uses multiple abstractions to find the reachable states of a neural network starting from its input. Its analysis is fast and accurate. PyRAT has already been used in several industrial and academic collaborations, to ensure safety guarantees, with its second place at the VNN-Comp 2024 showcasing its performance.

Fichier principal
Vignette du fichier
_SAS_tool_paper_PyRAT.pdf (4.31 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

cea-05394082 , version 1 (02-12-2025)

Licence

Identifiants

Citer

Augustin Lemesle, Julien Lehmann, Tristan Le Gall, Hichem-Zakaria Chihani. Verifying neural networks with PyRAT. SAS 2025 - 32nd Static Analysis Symposium, Oct 2025, Singapour, Singapore. pp.1 1-33, ⟨10.1007/978-3-032-07106-4_2⟩. ⟨cea-05394082⟩
199 Consultations
243 Téléchargements

Altmetric

Partager

  • More