I don't think some people are aware that "large" has always referred to the training inputs not resulting the size of the model. LLM is not "we made a language

I don't think some people are aware that "large" has always referred to the training inputs not resulting the size of the model. LLM is not "we made a language model and it is big" -- it's "we trained this model on a lot of language"