G-invariant and antisymmetric functions can be accurately approximated by polynomials, providing high accuracy for various applications.
G-Invariant and antisymmetric functions can be uniformly approximated by polynomials, with implications for deep learning.