logoalt Hacker News

5555watchyesterday at 10:04 PM0 repliesview on HN

> model except to limit its effectiveness in developing frontier LLMs

Does this imply that they're actively using it for their frontier development and that it's very effective?