Prehistory — From Neural Networks to the TransformerMartin Arjovsky, Soumith Chintala, Léon Bottou

Arjovsky, Chintala and Bottou publish Wasserstein GAN on arXiv

Wasserstein GAN released on arXiv

Why it matters

Redefined the instability of GAN training as a problem of the distance function

Source — Primary sources first; secondary reports are labelled “Press”

Paperarxiv.orghttps://arxiv.org/abs/1701.07875

Added in Week 35, 2026 · See that week

Same week

Sparsely-Gated Mixture-of-Experts LayerPaper

Nearby on the same axis

Finn, Abbeel and Levine publish Model-Agnostic Meta-Learning (MAML) on arXivPaperZhu et al. publish CycleGAN on arXivPaperIn-Datacenter Performance Analysis of a Tensor Processing UnitPaperGoogle Brain publishes Neural Architecture Search with Reinforcement LearningPaper