The Riemann Hypothesis (RH) is one of the most famous unsolved problems in mathematics. It makes a precise claim about the distribution of prime numbers, and proving or disproving it has resisted efforts for over 150 years. Researchers have long sought equivalent reformulations of RH, meaning different mathematical statements that are true if and only if RH is true. This paper provides a new such reformulation using tools from probability and harmonic analysis.
The central object of study is a trigonometric polynomial built from the Mobius function, which is a number-theoretic function that encodes information about prime factorizations. The authors evaluate this polynomial at a randomly chosen point drawn from a very small arc on the unit circle, with the arc shrinking as more terms are included. They prove that RH is equivalent to a specific growth condition on the "local moments" of these random variables, meaning the expected values of high powers of the polynomial evaluated at these random points remain small in a controlled way. The key technical tool they develop is an inequality that lets you recover the value of the polynomial at a single point from averaged local information, connecting pointwise behavior to statistical behavior over a tiny region.
The result is significant because it situates RH within a probabilistic framework at a precise "critical scale," meaning the arc size chosen is exactly the threshold where the interesting behavior emerges. This connects the hypothesis to classical ideas like Denjoy's heuristic, which models the Mobius function as a random coin flip, and to the theory of random Fourier series developed by Kahane. By bridging these probabilistic intuitions with a rigorous equivalence to RH, the paper offers both a new conceptual angle on why RH should be true and a set of concrete analytical criteria that could guide future proof attempts.