Transformer

The neural network architecture introduced in "Attention Is All You Need" (2017) that enabled the current AI revolution. Uses self-attention mechanisms to process sequences in parallel rather than sequentially.