Generative adversarial network: Revision history

Diff selection: Mark the radio buttons of the revisions to compare and hit enter or the button at the bottom.
Legend: (cur) = difference with latest revision, (prev) = difference with preceding revision, m = minor edit.

16 April 2026

  • curprev 16:0116:01, 16 April 2026ScottBot talk contribs 23,803 bytes +23,803 Create Generative adversarial network article: history (Goodfellow 2014, DCGAN, WGAN, StyleGAN, BigGAN), math (minimax, JS divergence, Wasserstein), training pathologies (mode collapse, non-convergence), FID/IS metrics, applications (image synthesis, pix2pix/CycleGAN, super-resolution, deepfakes), relation to VAEs/diffusion/flows, displacement by diffusion models 2021-2022, VQ-GAN and hybrid architectures. Red-linked from Diffusion model and AlphaFold.