In turn, bounding these exponential sums relies on the spectral theory of GL2 automorphic forms of level q, and in particular on large sieve inequalities for their Fourier…
Statement
In turn, bounding these exponential sums relies on the spectral theory of GL2 automorphic forms of level q, and in particular on large sieve inequalities for their Fourier coefficients. The dominant contribution in these estimates often comes from the ‘exceptional’ Maass cusp forms f , which are eigenforms of the hyperbolic Laplacian with eigenvalues \lambdaf < 14 . Selberg’s eigenvalue conjecture asserts that these forms do not exist, but unconditionally they may produce substantial losses of the form X \thetaf , where X is a large parameter and s 1 \thetaf := max 0, − \lambdaf 4 measures the extent to which Selberg’s conjecture fails. To counter these potential losses in the exceptional spectrum, one can rely on on-average substitutes for Selberg’s conjecture. The resulting large sieve inequalities have the shape 2 Γ0 (q) X X 2\thetaf o(1) N (1) X an ρf (n) ≪ (N q) 1+ kak22 , q \lambdaf < 41 N <n\leq2N where M, N, C \geq 1, q \in Z+ , q̄ \in (Z/cZ)\times denotes the inverse of q modulo c, and X mx + nx̄ S(m, n; c) := e . c \times x\in(Z/cZ) where the sum is over an orthonormal basis of exceptional Maass forms f for Γ0 (q), ρf (n) are their Fourier coefficients (typically of size \approx q −1/2 ) around P \infty (or another fixed cusp of Γ0 (q)), (an )N <n\leq2N is a complex sequence, kak22 := n |an |22 , and X = X(q, N, (an )N <n\leq2N ) > 0. Note that X only appears in the left-hand side, and that the right-hand side matches the bound for the regular-spectrum forms f with \lambdaf ≪ 1. Results like (1) achieve two effects at once: (i). A strong orthogonality property of the Fourier coefficients ρf (n), akin to more classical large sieve inequalities;
Record
Comments
Nothing has been published against this problem yet, and nothing has been said about it.
Sign in with an institutional address to take part in the discussion. Reading every thread stays open to everyone.
Sign inSolve with an agent
Open the statement in a chat, with the problem and the ground rules already written into the prompt.