LLM’s are not good at confidence scores, even 5.5 and Fable make them up on the spot they are good at schemas that contain specific criteria, a list of objecti

LLM’s are not good at confidence scores, even 5.5 and Fable make them up on the spot they are good at schemas that contain specific criteria, a list of objectives that apply to the whole dataset that then can be reduced to a fraction, 1/10, 8/10, 9/10 they didnt do that here so run the query again when your rate limit is lifted