TagLast edit: 5 Feb 2021 18:17 UTC by plex

Generative Adversarial Network is a machine learning architecture containing two modules. The generator attempts to create an output that is similar to the network’s training data, while the discriminator attempts to tell the generator’s outputs apart from the training data. The generator is reinforced based on how well its outputs fool the discriminator, so the two modules are adversaries.

GANs are best known for working well with images; for example, generating pictures of human faces.

Google’s new text-to-image model—Parti, a demon­stra­tion of scal­ing benefits

Kayden22 Jun 2022 20:00 UTC
32 points
4 comments1 min readLW link

[Question] GPT-3 + GAN

stick10917 Oct 2020 7:58 UTC
4 points
4 comments1 min readLW link

GAN Discrim­i­na­tors Don’t Gen­er­al­ize?

tryactions8 Jun 2020 20:36 UTC
18 points
7 comments2 min readLW link

Gen­er­a­tive ad­ver­sar­ial mod­els, in­formed by arguments

jessicata27 Jun 2016 19:28 UTC
0 points
0 comments2 min readLW link

Steganog­ra­phy and the Cy­cleGAN—al­ign­ment failure case study

Jan Czechowski11 Jun 2022 9:41 UTC
28 points
0 comments4 min readLW link

Can you force a neu­ral net­work to keep gen­er­al­iz­ing?

Q Home12 Sep 2022 10:14 UTC
2 points
10 comments5 min readLW link

Digi­tal Molec­u­lar Assem­blers: What syn­thetic me­dia/​gen­er­a­tive AI ac­tu­ally rep­re­sents, and where I think it’s going

Yuli_Ban28 Feb 2023 14:03 UTC
26 points
4 comments15 min readLW link
No comments.