LLM = Large Language Model. Asking it to show its work will cause it to generate more tokens (you could think of it as the logical leaps between each step being smaller the more detail you go into)