Conference paper

Algorithmic Polynomial Freiman-Ruzsa Theorems

Abstract

We prove algorithmic versions of the polynomial Freiman-Ruzsa theorem of Gowers, Green, Manners, and Tao (Annals of Mathematics, 2025) in additive combinatorics. In particular, we give classical and quantum polynomial-time algorithms that, for A⊆F2nA \subseteq \mathbb{F}_2^n with doubling constant~KK, learn an explicit description of a subspace V⊆F2nV \subseteq \mathbb{F}_2^n of size ∣V∣≤∣A∣|V| \leq |A| such that AA can be covered by KCK^C translates of VV, for a universal constant C>1C>1.