Proceedings of the 34th International Conference on Machine Learning, PMLR 70, 1321–1330 (2017). Primary paper PDF proceedings.mlr.press/v70/guo17a/guo17a.pdf . Section 2 defines top-label confidence calibration and reliability diagrams; Figure 1 compares error and confidence, section 4 presents temperature scaling. First three authors share equal-contribution credit, not solely first author.