“If entropix is indeed allowing the LLM to not simply replicate reasoning steps, then maybe this really is the key to deeper reasoning”

Posted on by

Link. I just wrote about feeling reassured by a paper showing absence of reasoning. I hope entropix doesn’t change that.