>> There's no way to verify you're getting the original prompt. > I do suggest a verification method for readers to pursue … you should be able to come to exac

>> There's no way to verify you're getting the original prompt. > I do suggest a verification method for readers to pursue … you should be able to come to exactly equal output given the same inputs for obviously low-temperature inputs 90ish% of the time. This sounds like “correct, there’s no way to verify,” but with more words.