Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper proves the Hardy-Littlewood-Chowla conjecture, which posits that the Möbius function exhibits cancellation on shifted primes, specifically $\sum_{p\le x}μ(p+2)=o(π(x))$. This result, a key model case for the parity problem in sieve theory, also implies that among primes $p$ where $p+2$ is square-free, the sets where $\mu(p+2)=1$ and $\mu(p+2)=-1$ have equal asymptotic density of $1/2$. The proof leverages advanced sieve methods and analytic number theory techniques.
The long-standing Hardy-Littlewood-Chowla conjecture, a benchmark for sieve theory, is finally proven, resolving the cancellation of the Möbius function on shifted primes.
An old conjecture dating back at least to Hildebrand's work in 1989, asserts that the Möbius function exhibits cancellation on shifted primes; that is, $$\sum_{p\le x}μ(p+2)=o(π(x)),\quad \text{or equivalently}\quad \sum_{n\le x}Λ(n)μ(n+2)=o(x).$$ This is also a well-known model case for the parity problem in sieve theory. In this paper, we prove this conjecture. As an application, we show that each of the two sets $\{p:\, μ(p+2)=1\}$ and $\{p:\, μ(p+2)=-1\}$ has asymptotic density $1/2$ among the primes $p$ with $μ^2(p+2)=1$: \begin{equation*} \lim_{x\to\infty} \frac{\#\{p\le x:\ μ(p+2)=κ\}} {\#\{p\le x:\ μ^2(p+2)=1\}} =\frac12, \quad κ\in\{-1,\,1\}. \end{equation*}