Yes, in the end it just makes it produce better answers most of the time, it doesn't eliminate hallucinations. Not sure how you expect it to objectively tell th

Yes, in the end it just makes it produce better answers most of the time, it doesn't eliminate hallucinations. Not sure how you expect it to objectively tell the correctness, short of formulating the prompt as axioms and performing formal verification. Which is probably not what you want from it.