Testing out KANs' abiliy to model different low-dimensional problems and comparing their performance to traditional NNs
Using tinygrad
I plan to eventually write a blog post about my findings from this work; it will end up here: https://cprimozic.net/blog/