Convolutional Kolmogorov-Arnold Networks

Alexander Dylan Bodner, Antonio Santiago Tepsich, Jack Natan Spolski, Santiago Pourteau
6/19/2024
cs.CVcs.AI

No official code found. Best implementation: AntonioTepsich/Convolutional-KANs (921 stars, MIT, last commit 2025-04-08).

Abstract

In this paper, we present Convolutional Kolmogorov-Arnold Networks, a novel architecture that integrates the learnable spline-based activation functions of Kolmogorov-Arnold Networks (KANs) into convolutional layers. By replacing traditional fixed-weight kernels with learnable non-linear functions, Convolutional KANs offer a significant improvement in parameter efficiency and expressive power over standard Convolutional Neural Networks (CNNs). We empirically evaluate Convolutional KANs on the Fashion-MNIST dataset, demonstrating competitive accuracy with up to 50% fewer parameters compared to baseline classic convolutions. This suggests that the KAN Convolution can effectively capture complex spatial relationships with fewer resources, offering a promising alternative for parameter-efficient deep learning models.

View on arXivView PDF

Code Implementations(1)

AntonioTepsich/Convolutional-KANsCommunity100%
MIT

This project extends the idea of the innovative architecture of Kolmogorov-Arnold Networks (KAN) to the Convolutional Layers, changing the classic linear transformation of the convolution to learnable non linear activations in each pixel.

92194Python, Jupyter NotebookMay 7, 20241 years agoMIT
cnncomputer-visiondeep-learning

Similar Papers

Convolutional Neural Networks Do Work with Pre-Defined Filters

83% match2024

KANICE: Kolmogorov-Arnold Networks with Interactive Convolutional Elements

83% match2 repos32024

KANs for Computer Vision: An Experimental Study

82% match2024

SoftReMish: A Novel Activation Function for Enhanced Convolutional Neural Networks for Visual Recognition Performance

82% match2025

ASU-CNN: An Efficient Deep Architecture for Image Classification and Feature Visualizations

81% match2023

SpiralMLP: A Lightweight Vision MLP Architecture

81% match1 repo42024

Cite this paper

@article{bodner2024convolutional,
  title  = {Convolutional Kolmogorov-Arnold Networks},
  author = {Alexander Dylan Bodner and Antonio Santiago Tepsich and Jack Natan Spolski and Santiago Pourteau},
  year   = {2024},
  eprint = {2406.13155},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2406.13155v3}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2406.13155.svg)](https://codeofpaper.com/paper/2406.13155)

Discussion