Any STS work on how the term "attention" is used in ML (esp transformers)?
2 reposts18 likes
Ooh, you were ahead of the game! I know @nickseaver.website taught a class on attention a few years ago. Maybe he knows!
The conclusion I came to eventually is that I am very confused about what attention is, it doesn't seem very important, and I would rather think about other things. But I lived two houses down from @nickseaver.website for a semester in 2017! I mean not the website, the person.