So, I've had my H100s grind for you all, and have some interesting new results AND fresh models! So, what did I find? Well because my blog article are too damn long (I know some of you are not reading the whole thing...), here is a TL;DR: I found that LLMs seem to think in a universal language. Duri
