Attack algorithms [19, 7, 76, 40, 2] aim to maliciously generate adversarial examples to fool a victim model, while adversarial defense aims at building robustness against them.

We study three notions of uncertainty quantification--calibration, confidence intervals and prediction sets--for binary classification in the distribution-free setting, that is without making any distributional assumptions on the data. With a focus towards calibration, we establish a'tripod' of theorems that connect these three notions for score-based classifiers. A direct implication is that distributionfree calibration is only possible, even asymptotically, using a scoring function whose level sets partition the feature space into at most countably many sets. Parametric calibration schemes such as variants of Platt scaling do not satisfy this requirement, while nonparametric schemes based on binning do.

Since utilities can serveas target values to learn the scoring function through square loss regression, the optimality ofsorting byexpected utilities isequivalent tothe consistencyofregression.

Onthe (non-)existenceofconvex, calibrated surrogatelossesforranking. InAdvancesin Neural Information Processing Systems 25, pages 197-205.

Textual backdoor attacks are a kind of practical threat to NLP systems. By injecting a backdoor in the training phase, the adversary could control model predictions via predefined triggers.

For challenging state estimation problems arising in domains like vision and robotics, particle-based representations attractively enable temporal reasoning aboutmultipleposteriormodes.