KANs with learnable univariate spline activations on edges achieve better accuracy than MLPs with fewer parameters, faster scaling, and direct visualization for scientific discovery.
Physics-informed machine learning
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
method 1
citation-polarity summary
roles
method 1polarities
use method 1representative citing papers
Physics-informed neural networks and autoencoders trained on synthetic data locate phase boundaries in the diluted Ising model by exploiting symmetry-breaking biases.
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.
-
Uncovering Magnetic Phases with Synthetic Data and Physics-Informed Training
Physics-informed neural networks and autoencoders trained on synthetic data locate phase boundaries in the diluted Ising model by exploiting symmetry-breaking biases.