Provable concept learning for interpretable predictions using variational autoencoders

Open in new window