The number theory concept at the heart of this paper is the "primitive root" modulo a prime number. For a given prime p, a primitive root is a number whose successive powers, when divided by p, cycle through all the non-zero remainders possible. The smallest such number for a given prime is called the least primitive root, written g(p). A long-standing conjecture by Grosswald predicts that this least primitive root is always smaller than the square root of p (with a small correction term), as long as p is larger than 409.
Proving Grosswald's conjecture in full remains out of reach with current mathematical tools, but researchers can make partial progress by showing the conjecture holds for all primes above some large threshold. Previous work had established the conjecture for primes above an astronomically large bound. This paper pushes that threshold down substantially, proving that the conjecture holds for all primes greater than roughly 1.77 times 10 to the power of 54, which is a significant improvement over prior results.
To achieve this, the authors refined existing analytic number theory techniques, particularly methods involving character sums and the distribution of prime numbers in arithmetic progressions. By carefully optimizing the underlying estimates and calculations, they were able to tighten the numerical bound considerably. The result brings mathematicians meaningfully closer to a complete proof of Grosswald's conjecture, and the techniques developed may be useful for related problems about the size of primitive roots and similar number-theoretic quantities.