Skip to content
Better HN
Top
New
Best
Ask
Show
Jobs
Search
⌘K
New deepseek paper: Natively Trainable Sparse Attention mechanism | Better HN
New deepseek paper: Natively Trainable Sparse Attention mechanism
(opens in new tab)
(twitter.com)
5 points
redlock
1y ago
1 comments
Share
1 comments
default
newest
oldest
eunos
1y ago
Authored and Uploaded by none others than Liang Wenfeng himself
j
/
k
navigate · click thread line to collapse
