Prehistory — From Neural Networks to the TransformerKaiming He et al. (Microsoft Research Asia)Microsoft Research
He et al. publish Deep Residual Learning for Image Recognition (ResNet) on arXiv
Deep Residual Learning for Image Recognition (ResNet) released on arXiv
Why it matters
Made training of very deep networks possible through residual connections. A basic component of most later architectures, including the Transformer
Source — Primary sources first; secondary reports are labelled “Press”
Paperarxiv.orghttps://arxiv.org/abs/1512.03385Added in Week 35, 2026 · See that week