MRSNorm pairs channels into 2D phasors, normalizes by the mean of per-pair L2 norms, and ties each pair's affine weight, reporting stable ResNet training on CIFAR-100 under high learning rates.
Understanding gradient clipping in private sgd: A geometric perspective.Advances in neural information processing systems, 33:13773– 13782, 2020
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2026 1verdicts
REJECT 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
Phasor Attention: Mean Root Square Normalization for Phase Manifold Preservation
MRSNorm pairs channels into 2D phasors, normalizes by the mean of per-pair L2 norms, and ties each pair's affine weight, reporting stable ResNet training on CIFAR-100 under high learning rates.