English

Low-Degree Fourier Threshold for Random Boolean Functions

Probability 2026-04-16 v1

Abstract

We study whether a uniformly random Boolean function f:{1,1}p{1,1}f : \{-1,1\}^p \to \{-1,1\} is determined by its Walsh--Fourier coefficients of degree at most dd. We show that the threshold lies at p/2p/2 up to an O(plogp)O(\sqrt{p \log p}) window: if dp2p2(logp+ω(1)), d \le \frac{p}{2} - \sqrt{\frac{p}{2}\bigl(\log p + \omega(1)\bigr)}, then with probability 1o(1)1-o(1) there exists another Boolean function gfg \ne f with the same degree-d\le d coefficients. Conversely, for every fixed η(0,1)\eta \in (0,1), if dp2+p2log6pη2, d \ge \frac{p}{2} + \sqrt{\frac{p}{2}\log\frac{6p}{\eta^2}}, then with probability at least 12p1-2^{-p}, the function ff is uniquely determined by its degree-d\le d coefficients, even among all bounded functions g:{1,1}p[1,1]g : \{-1,1\}^p \to [-1,1]. This resolves a question of Vershynin.

Keywords

Cite

@article{arxiv.2604.13493,
  title  = {Low-Degree Fourier Threshold for Random Boolean Functions},
  author = {Yiming Chen},
  journal= {arXiv preprint arXiv:2604.13493},
  year   = {2026}
}
R2 v1 2026-07-01T12:10:08.713Z