analiza aiqbench
This research proposes an Ising-compatible framework to formally verify the robustness of neural networks against adversarial perturbations. By utilizing logarithmic encoding for piecewise-linear activations and asymptotic step-envelope models, it achieves exact verification with reduced spin requirements compared to generic quantum Benders formulations. The work bridges machine learning safety guarantees with quantum optimization techniques, offering a pathway toward resource-efficient certification of AI systems.
Exact and Asymptotically Complete Robust Verifications of Neural Networks via Ising Solvers
arXiv:2603.00408v2 Announce Type: replace-cross Abstract: We present an Ising-compatible framework for formal neural-network robustness verification under bounded input perturbations. For piecewise-linear activations, the Exact Logarithmic PWL Model (Log-PWL) provides an exact, sound, and complete formulation with a state-optimal logarithmic encoding, reducing the binary variables per neuron from linear to information-theoretically minimal logarithmic complexity. For general bounded element-wise activations, the Asymptotic Step-Envelope Model (Step-Env) uses sound piecewise-constant envelopes whose lower and upper neuron states remain decision variables coupled to a common adversarial input. We prove that its globally optimized output bounds converge uniformly to the true network extrema as the segment width vanishes, yielding asymptotic completeness of verification. We further develop a hybrid Benders solver. In contrast to the exponential worst-case cut growth of generic quantum Benders formulations, our Helly-based conflict analysis gives a polynomial bound on infeasible iterations when infeasibility admits fixed-dimensional certificates. Interval pruning, certificate transfer for pruned networks, and layerwise classical--Ising partitioning further reduce spin requirements. Experiments show exact certification fidelity for piecewise-linear networks and near-reference accuracy for sigmoid networks with compact spin budgets.
Abstract:We present an Ising-compatible framework for formal neural-network robustness verification under bounded input perturbations. For piecewise-linear activations, the Exact Logarithmic PWL Model (Log-PWL) provides an exact, sound, and complete formulation with a state-optimal logarithmic encoding, reducing the binary variables per neuron from linear to information-theoretically minimal logarithmic complexity. For general bounded element-wise activations, the Asymptotic Step-Envelope Model (Step-Env) uses sound piecewise-constant envelopes whose lower and upper neuron states remain decision variables coupled to a common adversarial input. We prove that its globally optimized output bounds converge uniformly to the true network extrema as the segment width vanishes, yielding asymptotic completeness of verification. We further develop a hybrid Benders solver. In contrast to the exponential worst-case cut growth of generic quantum Benders formulations, our Helly-based conflict analysis gives a polynomial bound on infeasible iterations when infeasibility admits fixed-dimensional certificates. Interval pruning, certificate transfer for pruned networks, and layerwise classical--Ising partitioning further reduce spin requirements. Experiments show exact certification fidelity for piecewise-linear networks and near-reference accuracy for sigmoid networks with compact spin budgets.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Optics (physics.optics); Quantum Physics (quant-ph)
arXiv:2603.00408 [cs.LG]
(or arXiv:2603.00408v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2603.00408 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 271 cuvinte
Subiecte conexe
A brief history of quantum vs classical computational advantage
Quantum Journal · 1 sept. 2026
Security Properties of Neural Networks as Decision Problems
arXiv · 1 oct. 2026
Generalization and infinite width
LessWrong · 1 aug. 2026
News
Quantum Computing Report · 25 iul. 2026