Under MSE training of deep linear networks, data augmentation and hard-wired invariance share global optima and critical points; regularization adds saddles and its solution path limits to the hard-wired optimum.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Understanding Learning Invariance in Deep Linear Networks
Under MSE training of deep linear networks, data augmentation and hard-wired invariance share global optima and critical points; regularization adds saddles and its solution path limits to the hard-wired optimum.