Search papers, labs, and topics across Lattice.
This paper extends existing estimates for short sums of arithmetic functions at polynomial values to incorporate mixed moments with Kloosterman sums. By combining these refined estimates with improved lower bounds from Fouvry and Michel's method, the authors establish an asymptotic formula for sums involving $z^{\omega(n)} |\operatorname{Kl}(a;n)|^\nu$. The work also provides upper bounds for mixed moments of Hecke eigenvalues and Kloosterman sums, characterizes the distribution of large Kloosterman values, and confirms predicted orders of magnitude for the diagonal second moment of classical Kloosterman sums.
A precise asymptotic formula for mixed moments of arithmetic functions and Kloosterman sums reveals a deep connection between prime divisor counts and the distribution of modular exponential sums.
We extend the Nair--Tenenbaum estimates for short sums of nonnegative arithmetic functions at polynomial values to mixed moments involving Kloosterman sums. Combining these estimates with lower bounds obtained by refining the method of Fouvry and Michel, we prove that, for every fixed nonzero integer \(a\) and fixed \(z,ν>0\), \[ \sum_{n\leq x}z^{ω(n)} |\operatorname{Kl}(a;n)|^ν\asymp_{a,z,ν} x(\log x)^{z\mathfrak{s}(ν)-1}, \] where \(ω(n)\) counts distinct prime divisors and \(\mathfrak{s}(ν)\) is the \(ν\)-th absolute moment of the Sato--Tate measure. We also obtain upper bounds for mixed absolute moments of Hecke eigenvalues and Kloosterman sums, determine the frequency of large Kloosterman values on a logarithmic scale, and establish the order of magnitude predicted by Li and Sarnak for the diagonal second moment of classical Kloosterman sums. We obtain matching bounds for the smoothed spectral counting variance on the modular surface in a logarithmic range and the lower bound \(\int_T^{2T}S(t)^2\,dt\gg T^2/\log T\), where \(S(t)\) is the remainder in Weyl's law.