The highest lowest zero and other applications of positivity (Q1847907): Difference between revisions

The author investigates the \(L\)-functions associated with automorphic forms. His starting point is a technique and problem due to \textit{A. Odlyzko} [Sémin. Théor. Nombres Bordx. (2) 2, 119-141 (1990; Zbl 0722.11054)] and based on the application of the Riemann-Weil explicit formula with a subtle choice of test function. The first object was to estimate the lowest zero of the \(L\)-function, but it transpires that the same method also yields results on the gamma factors, that is on the Satake/Langlands parameters of the infinite components. The methods are applied here to forms on \(\text{SL}_n (\mathbb{R})/ \text{SO}_n (\mathbb{R})\) automorphic under \(\text{SL}_n (\mathbb{Z})\). Thus for general \(n\) and ``real'' cuspidal representations or \(n=2\) general cuspidal representations the author shows that either the generalized Riemann hypothesis is effectively false or the \(L\)-functions have a zero with imaginary part \(<14.13472\). For general \(n\) he also proves lower bounds on the eigenvalues of the Laplacian and vanishing theorems for certain cohomology groups. He also uses the theory of Eisenstein series to give an example of a noncuspidal automorphic form for \(\text{SL}_n(\mathbb{Z})\) in the complementary series of \(n=68\). The results here make use of extensive calculations, and they demonstrate vividly how effective the analytic methods of number theory combined with present-day computing possibilities can be.