We present a simple picture of the training process of self-supervised learning methods with
dual deep networks. In our picture, these methods learn their high-dimensional embeddings …

Self-training and contrastive learning have emerged as leading techniques for incorporating
unlabeled data, both under distribution shift (unsupervised domain adaptation) and when it …

As datasets in neuroscience increase in size and complexity, interpreting these high-
dimensional data is becoming more critical. However, develo** an intuition for patterns or …

Abstract Self-Supervised Learning (SSL) has emerged as the solution of choice to learn
transferable representations from unlabeled data. However, SSL requires to build samples …

Abstract Machine learning for differential equations paves the way for computationally
efficient alternatives to numerical solvers, with potentially broad impacts in science and …

In this paper, we provide a comprehensive toolbox for understanding and enhancing self-
supervised learning (SSL) methods through the lens of matrix information theory …

With the emergence of rich classification models and high computing power, recognition
systems are widely used in various fields. Unfortunately, as the scale of open systems …

Abstract Domain Generalization (DG), designed to enhance out-of-distribution (OOD)
generalization, is all about learning invariance against domain shifts utilizing sufficient …

The maximum entropy encoding framework provides a unified perspective for many non-
contrastive learning methods like SimSiam, Barlow Twins, and MEC. Inspired by this …

Self-supervised learning (SSL) has recently received significant attention due to its ability to
train high-performance encoders purely on unlabeled data-often scraped from the internet …