its not gonna do well on ARC-3 without some significant changes and effort The n... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		evilmathkid 9 days ago \| parent \| context \| favorite \| on: I trained a small transformer in 1.5hrs and it bea... its not gonna do well on ARC-3 without some significant changes and effort The new arch in that video is kinda misleading. Didn't really compare against proper baselines
		help

westurner 9 days ago [–]

Yeah, but still, how does that agent perform if paired with this inexpensively trained transformer instead of the more expensive foundation models?

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact