Re:Highlights from the Claude 4 system prompt
by antupis on Tuesday May 27, 2025 @06:53AM
(#44105677)
I think we already hit somekind of performance wall begin of this year. It feels that models are now balancing between rule following and agentic case and general stuff. eg Claude 4 sonet just feels better in Cursor and follows rules very well, and same time it gets equal or worse scores in benchmark against 3.7 Sonet.
[ Parent
| on HN ]
There are no comments on this comment yet.