Linear Recursive Feature Machines provably recover low-rank matrices
Sign Up to like & getrecommendations! Published in 2024 at "Proceedings of the National Academy of Sciences of the United States of America"
DOI: 10.1073/pnas.2411325122
Abstract: Significance Understanding the mechanism of how neural networks learn features from data is a fundamental problem in machine learning. Our work explicitly connects the mechanism of neural feature learning to a widely used class of… read more here.
Keywords: machines provably; linear recursive; provably recover; feature ... See more keywords