"Keep in mind that that GPT4.5 was only trained with pretraining, supervised finetuning, and RLHF, so this is not yet a reasoning model. Therefore, this model release does not push forward model capability in cases where reasoning is critical (math, code, etc.)."
- Andrej Karpathy - OpenAI Researcher
@ElhamPazhakh@rustybrick They are probably using context, not brand. Still I hope it wont stick with final version, looks like a fail to me and it will sure anger a lot of people.
@pedrodias Great analogy, I don't think ratio should be the same, I would still put more bucks on the ads basket as they are way easier to predict and to be accountable for, but most ratios (maybe 90%+ in Brazil) are insanely far from ideal.
Hello @Similarweb i cannot access any panel on #DigitalEdge2021 because they all show me the message: This session is only available for specific ticket types. Help?