i think the results shown in this paper contradict your assertion: https://openaccess.thecvf.com/content_ICCV_2019/papers/Abdal_Image2StyleGAN_How_to_Embed_Ima

i think the results shown in this paper contradict your assertion: https://openaccess.thecvf.com/content_ICCV_2019/papers/Abdal_Image2StyleGAN_How_to_Embed_Images_Into_the_StyleGAN_Latent_Space_ICCV_2019_paper.pdf https://openaccess.thecvf.com/content_ICCV_2019/papers/Abdal... given an arbitrary face, we can find its embedding in the latent space of the model. this shows that the model has the potential to generalise to real but unseen examples? on the other hand, i suspect you might be observing a bias in the structuring of the latent space. thispersondoesnotexist.com likely samples the latent space with a gaussian or uniform distribution, and while the latent space may contain the full spectrum of possibilities, the density of semantically meaningful embeddings may be structured around the distribution of the training set rather than a uniform or gaussian. i'm stretching my understanding of the topic in trying to convey this.