Attention Is All You Need on “Papers I keep rereading”, a list by Andre Coleman on TheLysts.
Details
Photo
—
Name
—
Comment
The transformer paper. Changed the whole field, feels like the 1992 Dream Team of ML papers. I've read it four times and STILL forget how positional encoding works.
Title
Attention Is All You Need
Institution
Google Brain / University of Toronto collab, widely taught at MIT