Kolmogorov Neural Networks can represent discontinuous functions

Abstract: In this paper, we show that the Kolmogorov two hidden layer neural network model with a continuous, discontinuous bounded or unbounded activation function in the second hidden layer can precisely represent continuous, discontinuous bounded and all unbounded multivariate functions, respectively.

It is a pretty cool result. Basically the lay summary (from my understanding) is that a multivariate function with n-arguments can be perfectly represented by a network with 2 hidden layers (one having width n and another width 2n) and two activation functions.

The abstract says it can represent "all unbounded multivariate functions." But there are uncomputable functions, so there must be some restriction.

What is a Kolmogorov Neural Network?

How would this ability show up if we built gpt out of these? More out of the box thinking?