Scientific Machine Learning with Kolmogorov-Arnold Networks

TL;DR

This paper reviews the adoption of Kolmogorov-Arnold Networks in scientific machine learning, highlighting their advantages over traditional MLPs in interpretability, flexibility, and modeling complex dynamics, supported by benchmarking and analysis.

Contribution

It provides a comprehensive review of recent KAN-based models, benchmarking their performance against MLPs, and discusses future challenges and research directions in the field.

Findings

KANs outperform MLPs in accuracy and convergence

KANs better capture complex nonlinear and high-frequency features

Benchmarking shows consistent improvements over traditional neural networks

Abstract

The field of scientific machine learning, which originally utilized multilayer perceptrons (MLPs), is increasingly adopting Kolmogorov-Arnold Networks (KANs) for data encoding. This shift is driven by the limitations of MLPs, including poor interpretability, fixed activation functions, and difficulty capturing localized or high-frequency features. KANs address these issues with enhanced interpretability and flexibility, enabling more efficient modeling of complex nonlinear interactions and effectively overcoming the constraints associated with conventional MLP architectures. This review categorizes recent progress in KAN-based models across three distinct perspectives: (i) data-driven learning, (ii) physics-informed modeling, and (iii) deep-operator learning. Each perspective is examined through the lens of architectural design, training strategies, application efficacy, and comparative evaluation against MLP-based counterparts. By benchmarking KANs against MLPs, we highlight consistent improvements in accuracy, convergence, and spectral representation, clarifying KANs' advantages in capturing complex dynamics while learning more effectively. In addition to reviewing recent literature, this work also presents several comparative evaluations that clarify central characteristics of KAN modeling and hint at their potential implications for real-world applications. Finally, this review identifies critical challenges and open research questions in KAN development, particularly regarding computational efficiency, theoretical guarantees, hyperparameter tuning, and algorithm complexity. We also outline future research directions aimed at improving the robustness, scalability, and physical consistency of KAN-based frameworks.