A norm penalty constraining activations to a sqrt(d)-radius hypersphere accelerates grokking by up to 6x on modular arithmetic via radial suppression in activation dynamics.
arXiv preprint arXiv:2504.17243 , year=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Radial Suppression Accelerates Algorithmic Generalization: A Geometric Analysis of Delayed Generalization
A norm penalty constraining activations to a sqrt(d)-radius hypersphere accelerates grokking by up to 6x on modular arithmetic via radial suppression in activation dynamics.