Oliver Daniels-Koch comments on Latent Adversarial Training

[Oliver Daniels-Koch]

Is anyone aware of preliminary empirical work here? (Not including standard adversarial training)