>Our models were designed and trained to learn concepts in order to help them solve new problems Isn’t that claim complete and utter nonsense? It is a large la

>Our models were designed and trained to learn concepts in order to help them solve new problems Isn’t that claim complete and utter nonsense? It is a large language model after all, not a large concept model.