Paper
in
Workshop: Security and Safety in Machine Learning Systems
Detecting Adversarial Attacks through Neural Activations
Graham Annett
Abstract: