Hacker News Reader

Not at the moment in our tool, though this is an area of great curiosity and research for us!

Does the discriminator model translate the images into an embedding space of it's own? Could such a space be used to generate images themselves?