The subspace approximation conjecture for functions on F2n\mathbb{F}_2^n

At least 19 years old · documented by

Let G=F2nG=\mathbb{F}_2^n, let f∈A(G)f\in A(G), and let ϵ∈(0,1]\epsilon\in(0,1]. Write

Af:=∥f∥A(G)∥f∥L∞(G)−1.A_f:=\|f\|_{A(G)}\|f\|_{L^\infty(G)}^{-1}.

Here codim⁡V\operatorname{codim}V denotes the codimension of a subspace V≤GV\leq G, and ∥⋅∥L2(x′+V)\|\cdot\|_{L^2(x'+V)} is the normalized L2L^2-norm on the coset x′+Vx'+V. Subspace approximation conjecture. There is a subspace VV of GG such that

codim⁡V≪ϵ−2Af,\operatorname{codim}V\ll\epsilon^{-2}A_f,

and

sup⁡x′∈G∥f−f(x′)∥L2(x′+V)≤ϵ∥f∥L∞(G).\sup_{x'\in G}\|f-f(x')\|_{L^2(x'+V)}\leq\epsilon\|f\|_{L^\infty(G)}.

The statement is proposed as a likely optimal bound for the model analogue of the paper’s main theorem. The source does not report a proof or disproof, so its status remains open.

References

Primary source

Tom Sanders, “The Littlewood-Gowers problem”, arXiv:math/0605522 (2010).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.