Oceania
Structured Semidefinite Programming for Recovering Structured Preconditioners Arun Jambulapati Simons Institute
We develop a general framework for finding approximately-optimal preconditioners for solving linear systems.
Enhancing Adversarial Contrastive Learning via Adversarial Invariant Regularization
However, it is unclear how the style-independence property benefits ACL-learned robust representations.
Getting ViT in Shape: Scaling Laws for Compute-Optimal Model Design Ibrahim Alabdulmohsin null, Xiaohua Zhai null, Alexander Kolesnikov, Lucas Beyer null
However, the simple power-law relation becomes more complicated when compute is considered.
Neural (Tangent Kernel) Collapse
We adopt the natural assumption that the empirical NTK develops a block structure aligned with the class labels, i.e., samples within the same class have stronger
One-Pass Distribution Sketch for Measuring Data Heterogeneity in Federated Learning
Federated learning (FL) is a machine learning paradigm where multiple client devices train models collaboratively without data exchange.