KANs with learnable univariate spline activations on edges achieve better accuracy than MLPs with fewer parameters, faster scaling, and direct visualization for scientific discovery.
A resource model for neural scaling law
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
background 2
citation-polarity summary
fields
cs.LG 3roles
background 2polarities
background 2representative citing papers
Strong superposition causes neural loss to scale as the inverse of model dimension due to geometric feature overlaps, explaining scaling laws for broad frequency distributions.
Tuning the depth-width ratio positions models in an efficient neural interaction interval that correlates with better generalization under fixed budgets and remains stable with scale.
citing papers explorer
-
KAN: Kolmogorov-Arnold Networks
KANs with learnable univariate spline activations on edges achieve better accuracy than MLPs with fewer parameters, faster scaling, and direct visualization for scientific discovery.
-
Superposition Yields Robust Neural Scaling
Strong superposition causes neural loss to scale as the inverse of model dimension due to geometric feature overlaps, explaining scaling laws for broad frequency distributions.
-
Law of Neural Interaction: Depth-Width Shape, Interaction Efficiency, and Generalization
Tuning the depth-width ratio positions models in an efficient neural interaction interval that correlates with better generalization under fixed budgets and remains stable with scale.