The paper deals with a classical question in number theory: when does a certain sequence of numbers first become negative? The sequences in question are called Hecke eigenvalues, which are special numbers attached to mathematical objects known as Hecke-Maass cusp forms. These forms are highly symmetric functions that encode deep arithmetic information. A natural question is how long these eigenvalue sequences can stay positive before turning negative for the first time, and the paper works on finding sharp lower bounds for this "first sign change."
Earlier work by Lamzouri established that for a related class of objects called holomorphic Hecke cusp forms, the first negative eigenvalue cannot appear too early, specifically not before a position that grows roughly as a power of the logarithm of the weight parameter. The present paper extends this result to Hecke-Maass forms, which are a different but related type of cusp form, and does so without needing to assume a major unproven conjecture called the Generalized Ramanujan Conjecture. This makes the result more broadly applicable and robust.
Beyond just reproducing the earlier result in a new setting, the authors also prove analogous lower bounds for the coefficients of objects called symmetric power L-functions, which are built from Hecke-Maass forms through a standard algebraic construction. These L-functions generalize the Riemann zeta function and carry rich information about primes and arithmetic. The sign change results here are described as "omega results," meaning they establish that the first sign change cannot happen too late either, giving a two-sided picture of where the transition from positive to negative values must occur.