pykan
pykan copied to clipboard
Kolmogorov Arnold Networks
Congratulations on this great piece of work. I have tried to do simple tests like functional approximations and it turns out for a variety of models KANs are poor performing...
I had a great time reading the paper, but understanding the code was a bit difficult. I think its really important to make the basic unit as simple and clean...
Similar to [here](https://github.com/KindXiaoming/pykan/issues/144) I get an attribute error when calling some methods (namely `plot`) after I've trained a model if I haven't first run a forward inference pass on the...
Hi! I made a test of how KAN could assist denoising diffusion and adapted the very initial spiral-diffusion model utilizing an MLP to KAN. You can see that a two...
https://github.com/Boris-73-TA/awesome-kan https://github.com/Boris-73-TA/OrthogPolyKANs https://github.com/AntonioTepsich/Convolutional-KANs 早期基本结论:参数少很多,效果接近大卷积 https://arxiv.org/abs/2405.07344 TKAN: Temporal Kolmogorov-Arnold Networks

When I tried to train various binary classification problems with KAN layer, I got the result that it seems to be difficult to beat MLP in general in most of...
Can you tell me how fix_symbolic is implemented? Can we convert our own symbolic function library into B-spline parameters, compare them, and choose the closest one?
**1. Fix description** ```python width: list number of neurons in each layer. e.g., [2,5,5,3] means 2D inputs, 5D outputs, with 2 layers of 5 hidden neurons. ``` should be ```python...