HN
Hacker News
Top
New
Best
Ask
Show
Jobs
Comment by _hl_ | Hacker News Reader
Full thread
_hl_
·
Attention is, after all, not what you need.
View on HN
jeremycochoy
·
Well, RWKV is using some linear (non-quadratic) form of attention so... strictly speaking... you still need a bit of attention :D
Reply on news.ycombinator.com