Why Machine Learning Models Should Be Allowed to Say “I Don’t Know”

Confidence, calibration, abstention, and the case for human review in real-world ML systems Many machine-learning systems behave as if every input deserves an answer. We encounter a version of this pr