No, it's not necessary for it to have had the optimization somewhere in its training set given how it works. LLMs can solve problems they were not trained on, w

No, it's not necessary for it to have had the optimization somewhere in its training set given how it works. LLMs can solve problems they were not trained on, with varying degrees of success.