News for hackers. Stuff that matters.
Re:Qwen4-Max by expectfun on Monday November 16, 2026 @11:41AM (#50519611)
1. It's fun.

2. I think they still are _somewhat_ informative. Like, the same model may produce very different pelicans at different reasoning efforts, even though the training data is the same. So looking at pelicans may give you _some_ information (like "what reasoning effort I want to use that model with?".

There are no comments on this comment yet.