International Symposium On Leveraging Applications of Formal Methods, Verification and Validation • Crete, Greece
Time: Sunday, 3.11
Room: Room 2
Authors: Daniel Neider, Taylor Johnson
Abstract: This Verification for Neuro-Symbolic Artificial Intelligence (VNSAI) track explores the intersection of neural networks and symbolic reasoning in artificial intelligence, combining the strengths of both approaches to enable learning, generalization, and explainability in complex systems. This track invites researchers to share their work on developing rigorous methodologies for analyzing and reasoning about neuro-symbolic AI systems, including formal verification, automated analysis, and symbolic explanation generation. Building on last year’s track dedicated to the Safety Verification of Deep Neural Network (SVDNN), it also seeks contributions to benchmarking and standardization efforts in neural network verification, including submissions to the International Verification of Neural Networks Competition (VNN-COMP), the Applied Verification for Continuous and Hybrid Systems Competition (ARCH-COMP), and Verification of Neural Networks standard (VNN-LIB) initiatives. The track’s overall goal is to foster collaboration across disciplines and advance the state-of-the-art in neuro-symbolic AI verification and explainability.
Paper: Download