English

Holographic functions and neural networks

Combinatorics 2026-05-22 v1 Machine Learning Probability

Abstract

A fuzzy Boolean function is a map f:\cuben[0,1]f:\cube^n\to [0,1], where nNn\in\mathbb N. We introduce and compare three ways of saying that such a function has bounded complexity. The first is a sampling property: the value f(x)f(x) can be recovered, up to small error and with high probability, from the values of a bounded number of randomly chosen coordinates of xx. We call this the holographic property. The second is a structural property: ff is uniformly close to a bounded-degree polynomial in boundedly many bounded linear coordinate forms. The third is computational: ff is uniformly close to the output of a neural network with a bounded number of non-input neurons, bounded Lipschitz activation functions and bounded incoming weights. We prove that these three properties are equivalent up to quantitative changes of the parameters. The implication from holography to polynomial structure uses a variant of a weak version of hypergraph regularity.

Keywords

Cite

@article{arxiv.2605.22666,
  title  = {Holographic functions and neural networks},
  author = {Balazs Szegedy},
  journal= {arXiv preprint arXiv:2605.22666},
  year   = {2026}
}
R2 v1 2026-07-22T07:26:37.782Z