News for hackers. Stuff that matters.
Re:Astra can read and write weights to improve Qwen by famouswaffles on Monday November 02, 2026 @01:59PM (#50341809)
This is pretty massive, and a result I (and I think most?) would have never expected. Holy Shit. How?? Did OpenAI figure out a way to train this? Are they even aware?
Re:Astra can read and write weights to improve Qwen by captainbland on Monday November 02, 2026 @05:09PM (#50344607)
It's probably emergent from training on other ML tasks and general ML/mathematical related reasoning rather than specifically trained for to be honest. Otherwise it would probably have been in an official marketing communication or something.