News for hackers. Stuff that matters.
Re:GPT‑6.2 Astra and Sol by arfelious on Tuesday November 17, 2026 @01:26PM (#50535913)
That may not be the main subject of the plot in this case but I found results of Gemini 4 Carbon in FrontierCode 2.0 interesting. Besides the cost staying practically same between different reasoning levels, the performance decrease in high reasoning seems a bit weird. That usually happened in models with xhigh/max reasoning where they overthinked the solution

There are no comments on this comment yet.