Harnessing Orthogonality to Train Low-Rank Neural Networks

Autor:	Coquelin, Daniel, Flügel, Katharina, Weiel, Marie, Kiefer, Nicholas, Debus, Charlotte, Streit, Achim, Götz, Markus
Rok vydání:	2024
Předmět:	Computer Science - Machine Learning Computer Science - Artificial Intelligence
Druh dokumentu:	Working Paper
Popis:	This study explores the learning dynamics of neural networks by analyzing the singular value decomposition (SVD) of their weights throughout training. Our investigation reveals that an orthogonal basis within each multidimensional weight's SVD representation stabilizes during training. Building upon this, we introduce Orthogonality-Informed Adaptive Low-Rank (OIALR) training, a novel training method exploiting the intrinsic orthogonality of neural networks. OIALR seamlessly integrates into existing training workflows with minimal accuracy loss, as demonstrated by benchmarking on various datasets and well-established network architectures. With appropriate hyperparameter tuning, OIALR can surpass conventional training setups, including those of state-of-the-art models.
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2401.08505 Zobrazit plný text záznamu View this record from Arxiv