To compress or not to compress—self-supervised learning and information theory: A review

R Shwartz Ziv, Y LeCun - Entropy, 2024 - mdpi.com
Deep neural networks excel in supervised learning tasks but are constrained by the need for
extensive labeled data. Self-supervised learning emerges as a promising alternative …

Firerisk: A remote sensing dataset for fire risk assessment with benchmarks using supervised and self-supervised learning

S Shen, S Seneviratne, X Wanyan… - … Conference on Digital …, 2023 - ieeexplore.ieee.org
In recent decades, wildfires have caused tremendous property losses, fatalities, and
extensive damage to forest ecosystems. Inspired by the abundance of publicly available …

Reverse engineering self-supervised learning

I Ben-Shaul, R Shwartz-Ziv, T Galanti… - Advances in …, 2023 - proceedings.neurips.cc
Understanding the learned representation and underlying mechanisms of Self-Supervised
Learning (SSL) often poses a challenge. In this paper, we 'reverse engineer'SSL, conducting …

Transfer learning for chemically accurate interatomic neural network potentials

V Zaverkin, D Holzmüller, L Bonfirraro… - Physical Chemistry …, 2023 - pubs.rsc.org
Develo** machine learning-based interatomic potentials from ab initio electronic structure
methods remains a challenging task for computational chemistry and materials science. This …

An information theory perspective on variance-invariance-covariance regularization

R Shwartz-Ziv, R Balestriero… - Advances in …, 2023 - proceedings.neurips.cc
Abstract Variance-Invariance-Covariance Regularization (VICReg) is a self-supervised
learning (SSL) method that has shown promising results on a variety of tasks. However, the …

Uncertainty Quantification in Machine Learning for Biosignal Applications--A Review

IP de Jong, AI Sburlea, M Valdenegro-Toro - arxiv preprint arxiv …, 2023 - arxiv.org
Uncertainty Quantification (UQ) has gained traction in an attempt to fix the black-box nature
of Deep Learning. Specifically (medical) biosignals such as electroencephalography (EEG) …

Beyond deep ensembles: A large-scale evaluation of bayesian deep learning under distribution shift

F Seligmann, P Becker, M Volpp… - Advances in Neural …, 2024 - proceedings.neurips.cc
Bayesian deep learning (BDL) is a promising approach to achieve well-calibrated
predictions on distribution-shifted data. Nevertheless, there exists no large-scale survey that …