Hi there! I'm guessing you've probably notice this as well. I've noticed that gpt-3.5 turbo now seems to apply CoT reasoning automatically, perhaps rendering some of this guidance less relevant. I believe this is a recent change, perhaps with the updates they introduced in the past few days.
Hi there! I'm guessing you've probably notice this as well. I've noticed that gpt-3.5 turbo now seems to apply CoT reasoning automatically, perhaps rendering some of this guidance less relevant. I believe this is a recent change, perhaps with the updates they introduced in the past few days.