179 questions
No questions match those filters.
← All questions
EvaluationMedium· Calibration· Long-form Generation
The model reports 80% confidence on a six-sentence answ...
This is one of the questions in the full AI/ML interview bank. Pro unlocks all 1789 questions; Premium includes the same bank plus the highest daily Practice limit.
See plansA scalar confidence score on a multi-sentence answer is mathematically ill-defined without context. It could represent the probability that the entire document is correct (a point mass at 0.80) or the expected fraction of correct sentences. The interpretation depends on whether a single incorrect sentence voids the entire document or merely represents a partial error.