For fixed q, the optimal constants BHint_{d,q} grow like exp(c_q sqrt(d log d)), which is subexponential, answering the question of Becker, Klein, Slote, Volberg and Zhang.
Asymptotic contractivity of the Bohnenblust--Hille inequality for polynomials with few interacting variables
1 Pith paper cite this work. Polarity classification is still indexing.
abstract
Let $K_{m,M}$ denote the optimal Bohnenblust--Hille constant on the class of $m$-homogeneous polynomials all of whose monomials involve at most $M$ different variables. We prove that, for every fixed $M$, these constants are asymptotically contractive: \[ \lim_{m\to\infty}K_{m,M}=1. \] More precisely, \[ 1\le K_{m,M}\le A_M^{M/m}m^{(M^2-1)/(2m)}, \] where $A_M$ depends only on $M$. The argument combines bounded projections onto exact support levels, a random colouring of the active variables, the classical multilinear Bohnenblust--Hille inequality and interpolation with Parseval's identity. We also point out that the qualitative conclusion follows from a recent, more general support-sensitive Bohnenblust--Hille inequality, although its direct application gives a slightly larger power of the homogeneous degree.
citation-role summary
citation-polarity summary
fields
math.FA 1years
2026 1verdicts
ACCEPT 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Subexponential Bohnenblust-Hille Inequalities on Finite Cyclic Groups
For fixed q, the optimal constants BHint_{d,q} grow like exp(c_q sqrt(d log d)), which is subexponential, answering the question of Becker, Klein, Slote, Volberg and Zhang.