Neural Network Verification in Control
This work tackles the critical problem of ensuring safety in AI-controlled systems, though it appears to be a tutorial/synthesis of existing methods rather than presenting novel research.
This tutorial addresses the challenge of verifying safety properties in neural network-based control systems by introducing and unifying recent verification techniques, then extending them to provide formal guarantees for neural feedback loops.
Learning-based methods could provide solutions to many of the long-standing challenges in control. However, the neural networks (NNs) commonly used in modern learning approaches present substantial challenges for analyzing the resulting control systems' safety properties. Fortunately, a new body of literature could provide tractable methods for analysis and verification of these high dimensional, highly nonlinear representations. This tutorial first introduces and unifies recent techniques (many of which originated in the computer vision and machine learning communities) for verifying robustness properties of NNs. The techniques are then extended to provide formal guarantees of neural feedback loops (e.g., closed-loop system with NN control policy). The provided tools are shown to enable closed-loop reachability analysis and robust deep reinforcement learning.