rusty_kan 0.1.1

An implementation of Kolmogorov-Arnold Networks in Rust
Documentation
  • Coverage
  • 69.88%
    58 out of 83 items documented30 out of 58 items with examples
  • Size
  • Source code size: 107.61 kB This is the summed size of all the files inside the crates.io package for this release.
  • Documentation size: 6.2 MB This is the summed size of all files generated by rustdoc for all configured targets
  • Ø build duration
  • this release: 20s Average build duration of successful builds.
  • all releases: 19s Average build duration of successful builds in releases after 2024-10-23.
  • Links
  • LordSaumya/DeepIron
    3 0 0
  • crates.io
  • Dependencies
  • Versions
  • Owners
  • LordSaumya

Rusty-KAN

Kolmogorov-Arnold Networks are a type of neural network that is based on the Kolmogorov-Arnold representation theorem. Essentially, it states that any continuous function can be represented as a composition of a finite number of univariate functions.

Therefore, a KAN is a neural network where the activation functions are placed on the edges and act as learnable weights for the network. The nodes simply sum the incoming activations.

For example, see the following image, which shows how functions are composed in a KAN to approximate continuous functions: alt text

This is the basis of the Rusty-KAN library, which is a Rust implementation of the KAN.

Installation

To install Rusty-KAN, simply add the following line to your Cargo.toml file:

[dependencies]

rusty-kan = "0.1.0"

References