Santiago Pourteau (@santipourteau) 's Twitter Profile
Santiago Pourteau

@santipourteau

ID: 1787661176320507904

calendar_today07-05-2024 01:50:21

17 Tweet

26 Followers

39 Following

Alex Bodner (@alexbodner_) 's Twitter Profile Photo

I’m thrilled to present the KAN Convolutional Layer, a promising neural architecture for image processing. Last week KANs came out, as an alternative to the MLP. We extended this idea to Convolutional Layers, creating the KAN Convolution. Join me in this thread to more about it🧵

Alex Bodner (@alexbodner_) 's Twitter Profile Photo

All the code is available in the following repository: github.com/AntonioTepsich… Co Authored with Jack spolski Santiago Pourteau @antotepsich Lets dive in what we've created 👇

Valeriy M., PhD, MBA, CQF (@predict_addict) 's Twitter Profile Photo

With KAN superpowers you can! KAN Convolution by Alex Bodner , Jack spolski, Santiago Pourteau, @antotepsich shows amazing results with tight run between KAN vs Classic convolutions with a MLP at the end. 🔥🔥🔥🔥🔥 KAN Convs gets only 0.04 less accuracy with almost half the

Ethan (@torchcompiled) 's Twitter Profile Photo

So this is one of their many you true colors. Is this someone you can trust to suppress all other scientists and let this be the sole lab leading humanity to AGI? You had the Mandate of Heaven and you fumbled the bag sir

Alex Bodner (@alexbodner_) 's Twitter Profile Photo

💾🧠Want to know how much VRAM you will need for training your AI model? 💾🧠 Now you can use our app in which you can input a torch/tensorflow summary or the parameter count and get an estimate of the required memory! Use it in: howmuchvram.com Everything is open source!

Alex Bodner (@alexbodner_) 's Twitter Profile Photo

I’m excited to share an update on the Convolutional Kolmogorov-Arnold Networks paper, where we extended the KANs idea to Convolutional Layers. 🧵Join me in this thread to more about it🧵 (I share the code and paper below)

Alex Bodner (@alexbodner_) 's Twitter Profile Photo

This work was co-authored with: Antonio Tepsich Jack spolski and Santiago Pourteau You can find the paper at: arxiv.org/abs/2406.13155 and all the code in our GitHub repo: github.com/AntonioTepsich… Special thanks to the professors martin palazzo and Gastón Castro for the help in this