Check out my latest Medium post wherein I argue that Big Tech’s dominance of AI research will prove to be strategically disastrous for the West.

While previous work has developed algorithmic approaches to attacking and defending Variational Autoencoders (VAEs), there remains a lack of formalization for what it even means for a VAE to be robust to adversarial attack.

My recent Neurips paper, written with my friends and colleagues at the University of Oxford, studies the often used and yet poorly understood method of adding Gaussian noise to data and neural network activations.