News for hackers. Stuff that matters.

A comment on Towards understanding multiple attention sinks in LLMs.

Re:Towards understanding multiple attention sinks in LLMs by thw20 on Tuesday March 17, 2026 @05:36AM (#47410424)
The up to date paper documenting and analysing the observation is now available on arxiv!
[ Parent | on HN ]

There are no comments on this comment yet.