← InfoGAN: Interpretable Representation Learning ... - arxiv.org