A curvature penalty for KANs, derived to respect compositional effects and equipped with a proven upper bound on full-model curvature, produces smoother activations while preserving accuracy.
Opening the black-box: symbolic regression with Kolmogorov- Arnold networks for energy applications
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
-
KANs need curvature: penalties for compositional smoothness
A curvature penalty for KANs, derived to respect compositional effects and equipped with a proven upper bound on full-model curvature, produces smoother activations while preserving accuracy.