eess.SYNov 16, 2026

Hyperlyve: Parallelized Enumeration and Exact Lyapunov Stability Certification for Deep ReLU Networks

Authors: Brian Yarbrough, Jesse Wayment, Jingbo Wang, Philip E. Paré, Shreyas Sundaram

Organizations: The authors are with the Elmore Family School of Electrical Engineering at Purdue University, West Lafayette, IN 47906, USA.

Abstract

We introduce Hyperlyve (Hyperplane Partitioned Lyapunov Verification and Enumeration), a framework for proving the asymptotic stability of nonlinear dynamical systems, including those with neural network controllers. Hyperlyve learns and formally certifies continuous piecewise-affine (CPA) neural Lyapunov functions within a counterexample-guided inductive synthesis (CEGIS) loop. Hyperlyve's key innovation is a parallel algorithm that enumerates the hyperplane arrangement formed by the ReLU activations of deep networks and certifies that Lyapunov conditions hold within each polyhedral region. We derive sound and complete regional tests to verify the asymptotic stability of the exact smooth dynamics (rather than a piecewise-linear approximation) within a compact domain of interest. Furthermore, we formally characterize the computational complexity of the CPA neural Lyapunov verification problem as co-NP-complete, filling a key gap in the literature. Empirical results demonstrate that Hyperlyve enumerates ReLU networks over 150-times faster than competing tools, which contributes to Hyperlyve proving the asymptotic stability of systems that are currently intractable for state-of-the-art verifiers.

Figures & tables

Explore similar work

CardsList
  1. (DNN)2(\text{DNN})^2: Doubly Non-Negative Relaxations for Deep Neural Networks

    Aug 25, 2026Hanna Jiamei Zhang, Alan Papalia, Michael Everett +1Neural Network VerificationReLU Neural Networks

  2. Learning and Verifying Stabilizing Controllers with Structured Neural Control Lyapunov Functions

    Nov 16, 2026Jose Luis Ledesma Martin, Patricia PauliControl Lyapunov FunctionsNeural Network Verification