Lex Fridman Podcast
#452 – Dario Amodei: Anthropic CEO on Claude, AGI & the Future of AI & Humanity
Like, you know, maybe that's just something, like, maybe the linear representation hypothesis and supersession hypothesis is the right way to understand a one-layer model, but it's not the right way to understand larger models. And so I think, I mean, first of all, the Cunningham et al paper sort of cut through that a little bit and sort of suggested that this wasn't the case.
0
💬
0
Comments
Log in to comment.
There are no comments yet.