They say this is their last fully non-reasoning model, but that research on both types will continue. 

No, they said that GPT4.5 and GPT5 will be their last non-reasoning models.

They say it's currently limited to Pro users,

Meh, it's coming to plus users in ~a week.

It claims to be more accurate at standard questions and with a lower hallucination rate than any previous OAI model (and presumably any others). 

I think this is a big point and a better world knowledge is going to prove tremendously useful when it comes to applying RL to base models and a lower hallucination rate leads to effective exploration of the reasoning space + a better dataset after rejection sampling. Which should lead to lots of gains over models trained with RL over 4o. 

Not to speak about alignment - but looks like a big W for OpenAI - especially if they're going to raise in the near future (<6 months).

Spot on

Is this a massive exfohazard? 

Very Unlikely

Should this have been published?

Yes

I know this sounds fantastic but can someone please dumb down what KANs are for me, why they're so revolutionary (in practice, not in theory) that all the big labs would wanna switch to them?

Or is it the case that having MLPs is still a better thing for GPUs and in practice that will not change?

And how are KANs different from what SAEs attempt to do

• ^[4]
• AI life coaches

not excited about this - such a coach is either going to give very politically correct opinions, or target audiences with glaring insecurities, like young or low confidence men.. just like human coaches. 

I dont know if you are aware, but this post was covered by Yannic Kilcher in his video "No, Anthropic's Claude 3 is NOT sentient" (link to timestamp) 

If I understand this right, you train a transformer on data generated from a hidden markov process, of the form {0,1,R} and find that there is a mechanism for tracking when R occurs in the residual stream, as well as that the transformer learns the hidden markov process. is that correct?