Work in progress

This repository contains a very fast implementation of Kolmogorov-Arnold Network (KAN). The forward time of FaskKAN is 3.33x faster than efficient KAN, and the implementation is a LOT easier.

The original implementation of KAN is pykan.

One can install fast-kan via pip:

git clone https://github.com/ZiyaoLi/fast-kan
cd fast-kan
pip install .

Run an example training of FastKAN network on MNIST:

python examples/train_mnist.py

1. Used Gaussian Radial Basis Functions to approximate the B-spline basis, which is the bottleneck of KAN and efficient KAN:

$$b_{i}(u)=\exp\left(-\left(\frac{u-u_i}{h}\right)^2\right)$$

The rationale of doing so is that these RBF functions well approximate the B-spline basis (up to a linear transformation) and are very easy to calculate (as long as the grids are uniform). Results are shown in the figure below (code in notebook).

2. Used LayerNorm to scale inputs to the range of spline grids, so there is no need to adjust the grids.

3. FastKAN is 3.33x compared with efficient_kan in forward speed. (see notebook, 742us -> 223us on V100)

4. Accuracy on MNIST is equivalent / slightly improved.

5. More importantly, the approximation made in FastKAN suggests that KAN is equivalent to a certain RBF Network. This finding bridges between RBF Networks and KANs.

FastKANLayer supports users to plot the learned curves dim-by-dim. See notebook for an example usage.

Copyright 2024 Li, Ziyao. Licensed under the Apache License, Version 2.0 (the "License");

@article{li2024kolmogorovarnold,
      title={Kolmogorov-Arnold Networks are Radial Basis Function Networks}, 
      author={Ziyao Li},
      year={2024},
      eprint={2405.06721},
      archivePrefix={arXiv},
      primaryClass={cs.LG}
}