Problem
Research questions and friction points this paper is trying to address.
Neural Network Verification
Formal Verification
Deep Learning
Cyber-Physical Systems
Fairness
NNV3: Expanding Neural Network Verification to New Architectures and Domains
🤖 AI Summary
This work addresses the limitations of existing formal verification tools in covering novel architectures, such as graph neural networks (GNNs), and complex scenarios where deterministic verification is often infeasible. To this end, we develop NNV3, a MATLAB-based framework that introduces new families of star sets, including ModelStar. By integrating set-based reachability analysis, conformal inference-driven probabilistic reachability, and continuous-region fairness certification, the proposed framework enables efficient verification of models processing GNNs and video inputs. Furthermore, this project constructs multi-domain benchmarks that significantly enhance the robustness and usability of the verification tool. Ultimately, this work provides a scalable solution for the comprehensive verification of artificial intelligence systems.
📝 Abstract
We present NNV3, the latest version of the Neural Network Verification (NNV) tool, a MATLAB framework for formal verification of deep learning models and learning-enabled cyber-physical systems. Building on the set-based reachability foundation of NNV 1.0 (FFNNs, CNNs, NNCS) and NNV 2.0 (RNNs, SSNNs, neural ODEs), NNV3 introduces new members of the Star-set family: ModelStar for verifying networks under weight perturbation, VolumeStar for video and 3D volumetric inputs, and GraphStar for graph neural networks. A conformal-inference-based probabilistic reachability mode complements sound analysis for problems where deterministic verification is intractable, while FairNNV certifies counterfactual and individual fairness properties over continuous input regions. NNV3 introduces new benchmarks for malware detection, graph-based power-system models, medical imaging, variable-length time series data, and action recognition. NNV3 also incorporates tutorials and developer guides through a unified documentation site. This paper details these major updates, demonstrating NNV's maturation into a comprehensive, robust, and accessible verification tool for a diverse range of AI systems.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Neural Network Verification
Star-set
Conformal Inference
Graph Neural Networks
Fairness Certification
🔎 Similar Papers
No similar papers found.
