Comment

Comment by Legend2440

>It took a while, but the original ARC ultimately fell to exactly the approach it was supposed to be protected from, i.e. big data memorisation

No it didn't. People tried big data memorization, and it didn't work. Base LLMs (even with millions of synthetic examples) never solved ARC-AGI-1.

It took a real algorithmic advancement - reasoning models - to solve it.

parent Legend2440 3 months ago

Discussion

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton