Architectural Backdoors in Neural Networks. (arXiv:2206.07840v1 [cs.LG])

Web: http://arxiv.org/abs/2206.07840

Machine learning is vulnerable to adversarial manipulation. Previous
literature has demonstrated that at the training stage attackers can manipulate
data and data sampling procedures to control model behaviour. A common attack
goal is to plant backdoors i.e. force the victim model to learn to recognise a
trigger known only by the adversary. In this paper, we introduce a new class of
backdoor attacks that hide inside model architectures i.e. in the inductive
bias of the functions used to train. These …

arxiv lg networks neural neural networks