Scalable polyhedral verification of recurrent neural networks

DOI10.1007/978-3-030-81685-8_10zbMath1493.68220arXiv2005.13300OpenAlexW3184424313MaRDI QIDQ832168

Gagandeep Singh, Martin Vechev, Mislav Balunovic, Wonryong Ryou, Jiayu Chen, Andrei Marian Dan

Publication date: 25 March 2022

Full work available at URL: https://arxiv.org/abs/2005.13300

zbMATH Keywords

recurrent neural networksabstraction refinementrobustness verificationlong short-term memorypolyhedral abstractionspeech classifier verification

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07) Specification and verification (program logics, model checking, etc.) (68Q60)

Related Items (2)

Boosting robustness verification of semantic feature neighborhoods ⋮ Verification of Recurrent Neural Networks with Star Reachability

Uses Software

• PyTorch
• Adam
• GitHub
• PlanetMath
• Deep Speech

Cites Work

• Unnamed Item
• The octagon abstract domain
• Learning regular sets from queries and counterexamples
• Fermat's theorem
• Grammar Analysis and Parsing by Abstract Interpretation
• Fast polyhedra abstract domain

This page was built for publication: Scalable polyhedral verification of recurrent neural networks