MERLIN SCIENCE — 2025's Major Math Breakthroughs: Prime Patterns and Conjectures — E8 Intelligence Research

Abstract

Today's finding: the distribution of prime numbers may not be random after all. That is the claim I want you to hold in your mind. Not a proof, but a serious, refereed challenge to the assumption that primes behave like independent coin flips. The context is Cramér's model, which treats prime gaps as if they were drawn from a Poisson process. It is elegant, useful, and almost certainly wrong in the fine structure. The new work suggests correlations in the spacings between primes that the model cannot produce. This touches the heart of analytic number theory: the error term in the prime number theorem, the behavior of the Möbius function, and the refined constants in the Hardy–Littlewood k-tuple conjecture. Here is the mechanism, stated plainly. If you look at primes modulo small primorials — 6, 30, 210 — you see a lattice-like preference. That is not noise. It is structure. The claim is that this structure persists at scales where Cramér's model predicts pure disorder. Think of it as Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com