News for hackers. Stuff that matters.
Re:GPT‑6.2 Astra and Sol by jotato on Tuesday November 17, 2026 @01:10PM (#50535638)
which benchmark are you looking at?
Re:GPT‑6.2 Astra and Sol by Isaackoz on Tuesday November 17, 2026 @01:23PM (#50535867)
I was loosely looking at the average of all of them. Regarding your edit, it does look like the difference for frontiercode is negligible; the difference of 3% in score is meaningless to me. Guess we'll see how it performs today on my project.
Re:GPT‑6.2 Astra and Sol by derac on Tuesday November 17, 2026 @01:35PM (#50536045)
I think 1. the problems on FrontierCode are incredibly hard. 2. it's 3% in absolute terms, but it's a 10% improvement.