An unconditional \(\mathrm{GL}_n\) large sieve (Q2219334)

From MaRDI portal
scientific article
Language Label Description Also known as
English
An unconditional \(\mathrm{GL}_n\) large sieve
scientific article

    Statements

    An unconditional \(\mathrm{GL}_n\) large sieve (English)
    0 references
    0 references
    0 references
    20 January 2021
    0 references
    The authors prove a large sieve inequality for \(\operatorname{GL}_n\), with applications to zeroes of \(L\)-functions. The main result (Theorem 1.1 in the present paper under review) is as follows: Let \(F\) be a number field with ring of integers \(\mathcal{O}_F\), and fix a real number \(\varepsilon>0\), as well as an integer \(n\geq 1\). Then, for all \(N,Q\geq 1\) and for all complex-valued functions \(a(\mathfrak{n})\) on the set of ideals \(\mathfrak{n}\) of \(\mathcal{O}_F\), we have \[ \sum_{\pi \in \mathfrak{F}_n(Q)}\Big\vert \sum_{\substack{\operatorname{N}\mathfrak n \leq N \\ (\mathfrak{n},\mathfrak{q}_\pi)=\mathcal{O}_F}}\lambda_{\pi}(\mathfrak{n})a(\mathfrak{n}) \Big\vert^2 <\!\!\!<_{n,[F:\mathbb Q],\varepsilon} (NQ)^\varepsilon(N+Q^{n^2+n}|\mathfrak{F}_n(Q)|)\sum_{\operatorname{N}\mathfrak n \leq N}|a(\mathfrak{n})|^2 \] where the notation `\(f<\!\!\!<_\nu g\)' signifies that there exists an effectively computable constant \(c_\nu>0\), depending at most on \(\nu\), such that \(|f|\leq c_\nu|g|\). Here, \(\mathfrak{F}_n(Q)\) denotes the universal family of all cuspidal automorphic representations \(\pi\) of \(\operatorname{GL}_n(\mathbb A_F)\) with unitary central character, normalized to be trivial on the diagonally embedded copy of the positive reals, whose analytic conductor \(C(\pi)\) is bounded from above by \(Q\); \(\mathfrak{q}_\pi\) denotes the arithmetic conductor of \(\pi\); and \(\lambda_\pi(\mathfrak{n})\) is the Hecke eigenvalue of \(\pi\) at the nonzero integral ideal \(\mathfrak{n}\) of \(\mathcal{O}_F\). Assuming both the generalized Ramanujan conjecture and the generalized Lindelöf hypothesis, the term \(Q^{n^2+n}\) can be improved to \(\min\{Q^n,\sqrt{N}\}\). On the other hand, their result is sharp for \(N>Q^{n^2+3n}\). As for the applications to \(L\)-functions, given real numbers \(0\leq \sigma\leq 1\) and \(T \geq 1\), let \(N_\pi(\sigma,T)\) be the cardinality of the set of zeroes of \(L(s,\pi)\) whose real part is strictly greater than \(\sigma\) and whose imaginary part is bounded, in absolute value, by \(T\) from above. Then (Theorem 1.2): \[ \sum_{\pi \in \mathfrak{F}_n(Q)}N_\pi(\sigma,T) <\!\!\!<_{n,[F:\mathbb Q]}(QT^{[F:\mathbb Q]})^{10^7n^4(1-\sigma)}. \] This uses Theorem 1.1 together with ideas of \textit{P. X. Gallagher} [Invent. Math. 11, 329--339 (1970; Zbl 0219.10048)], and \textit{K. Soundararajan} and the first author [Duke Math. J. 168, No. 7, 1231--1268 (2019; Zbl 1426.11053)]. Finally, the authors also give a bound on \(|L(\frac 12,\pi)|\) in terms of the analytic conductor of \(\pi\); see Theorem 1.3 for the precise statement.
    0 references
    0 references
    automorphic representations
    0 references
    \(L\)-functions
    0 references
    large sieve
    0 references
    Selberg sieve
    0 references
    log-free zero density estimate
    0 references
    subconvexity
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references