News for hackers. Stuff that matters.
Re:Claude 4 by comova on Thursday May 22, 2025 @01:07PM (#44064144)
I believe this is to improve performance by shortening the context window for long thinking processes. I don't think this is referring to real-time summarizing for the users' sake.
Re:Claude 4 by usaar333 on Thursday May 22, 2025 @01:53PM (#44064711)
When you do a chat are reasoning traces for prior model outputs in the LLM context?
Re:Claude 4 by int_19h on Thursday May 22, 2025 @04:21PM (#44066443)
No, they are normally stripped out.
Re:Claude 4 by j_maffe on Thursday May 22, 2025 @01:56PM (#44064751)
> I don't think this is referring to real-time summarizing for the users' sake.

That's exactly what it's referring to.