Deep Learning
Quantifying Learnability and Describability of Visual Concepts Emerging in Representation Learning
The increasing impact of black box models, and particularly of unsupervised ones, comes with an increasing interest in tools to understand and interpret them.
Improving Calibration through the Relationship with Adversarial Robustness
Neural networks lack adversarial robustness, i.e., they are vulnerable to adversarial
Deep Energy-Based Modeling of Discrete-Time Physics
However, this may not be possible in discrete time, which is often the case in practical learning and computation.
Learning from Positive and Unlabeled Data with Arbitrary Positive Shift Supplemental Materials A Nomenclature
Boldface denotes the best performing algorithm according to mean misclassification rate.
Drawing out of Distribution with Neuro-Symbolic Generative Models
Learning general-purpose representations from perceptual inputs is a hallmark of human intelligence.
