X · @dwarkesh_sp
· X / Twitter
RT Eric Neyman: It's kinda crazy that Paul wrote this, specifically naming attention (in the colloquial sense) as a possibly-game-changing computation…
RT Eric NeymanIt's kinda crazy that Paul wrote this, specifically naming attention (in the colloquial sense) as a possibly-game-changing computational mechanism, before Attention Is All You Need.Dwarkesh Patel: In 2016 (before transformers) Paul wrote,"It’s plausible that a large neural network can replicate “fast” hum