√

The passage around this formula

Vaswani and colleagues removed recurrence entirely, proposing an architecture “based solely on attention mechanisms, dispensing with recurrence and convolutions entirely”, and reported 28.4 BLEU on WMT 2014 English-to-German and 41.8 on English-to-French with substantially less training time [ 11 ] . The core operation is a single scaled dot-product: $\mathrm{Attention}(Q, K, V) = \mathrm{softmax}\!\left(\frac{QK^{\top}}{\sqrt{d_k}}\right)V$. The analytical point is that this is primarily a hardware result dressed as an architectural one. Every position attends to every other position in one parallel matrix multiplication, which maps precisely onto accelerator hardware in a way that a sequential recurrence never can. The transformer is what made it economically…

Read this part in the article →