Score: 1

Fast and Scalable Score-Based Kernel Calibration Tests

Published: October 16, 2025 | arXiv ID: 2510.14711v1

By: Pierre Glaser , David Widmann , Fredrik Lindsten and more

Potential Business Impact:

Checks if computer predictions are trustworthy.

Business Areas:
Skill Assessment Education

We introduce the Kernel Calibration Conditional Stein Discrepancy test (KCCSD test), a non-parametric, kernel-based test for assessing the calibration of probabilistic models with well-defined scores. In contrast to previous methods, our test avoids the need for possibly expensive expectation approximations while providing control over its type-I error. We achieve these improvements by using a new family of kernels for score-based probabilities that can be estimated without probability density samples, and by using a conditional goodness-of-fit criterion for the KCCSD test's U-statistic. We demonstrate the properties of our test on various synthetic settings.

Repos / Data Links

Page Count
26 pages

Category
Statistics:
Machine Learning (Stat)