Comment on OpenAI o1 Results on ARC-AGI-PubparentComments−wslh1ySo basically it's a kind of overfitting with pattern matching features? This doesn't undermine the power of LLMs but it is great to study their limitations.
Comments
So basically it's a kind of overfitting with pattern matching features? This doesn't undermine the power of LLMs but it is great to study their limitations.