Quantum-Inspired Analysis of Neural Network Vulnerabilities: The Role of Conjugate Variables in System Attacks

arXiv:2402.10983v1 Announce Type: cross
Abstract: Neural networks demonstrate inherent vulnerability to small, non-random perturbations, emerging as adversarial attacks. Such attacks, born from the gradient of the loss function relative to the input, are discerned as input conjugates, revealing a systemic fragility within the network structure. Intriguingly, a mathematical congruence manifests between this mechanism and the quantum physics' uncertainty principle, casting light on a hitherto unanticipated interdisciplinarity. This inherent susceptibility within neural network systems is generally intrinsic, highlighting not only the …

adversarial adversarial attacks analysis arxiv attacks born cs.cr cs.lg emerging function input loss network networks neural network neural networks non quant-ph quantum random relative role structure system vulnerabilities vulnerability