On a Goldbach-type problem for the Liouville function

Alexander P. Mangerel
2024-05-02
Abstract:Let $\lambda$ denote the Liouville function. We show that for all sufficiently large integers $N$, the (non-trivial) convolution sum bound $$ \left|\sum_{1 \leq n < N} \lambda(n) \lambda(N-n)\right| < N-1 $$ holds. This (essentially) answers a question posed at the 2018 AIM workshop on Sarnak's conjecture.
Number Theory
What problem does this paper attempt to address?
### What problem does this paper attempt to solve? This paper attempts to solve a problem similar to the Goldbach conjecture, but for the Liouville function. Specifically, the author attempts to prove that for a sufficiently large integer \(N\), the following bound for the convolution sum holds: \[ \left| \sum_{n < N} \lambda(n) \lambda(N - n) \right| < N - 1 \] where \(\lambda\) represents the Liouville function. This problem was proposed at the 2018 AIM workshop regarding the Sarnak conjecture. #### Problem background - **Liouville function**: The Liouville function \(\lambda(n)\) is defined as: \[ \lambda(n) = (-1)^{\Omega(n)} \] where \(\Omega(n)\) is the number of prime factors of \(n\) (including multiplicities). - **Convolution sum**: Consider the convolution sum in the following form: \[ L_\lambda(N) := \sum_{1 \leq n < N} \lambda(n) \lambda(N - n) \] - **Objective**: Prove that for all sufficiently large integers \(N\), the above convolution sum satisfies: \[ |L_\lambda(N)| < N - 1 \] #### Analogy with the Goldbach conjecture This problem can be regarded as an analogy to the binary Goldbach conjecture. In the binary Goldbach conjecture, if the von Mangoldt function \(\Lambda_1(n)\) is used to replace the Liouville function \(\lambda(n)\), then proving the existence of prime numbers \(p\) and \(q\) such that \(p + q = N\) is equivalent to proving: \[ L_{\Lambda_1}(N) := \sum_{n < N} \Lambda_1(n) \Lambda_1(N - n) > 0 \] #### Existing results and challenges Although there are some known results, such as the conjecture of Corrádi and Kátai and the theorem of De Koninck, Germán and Kátai, estimating the convolution sum \(L_\lambda(N)\) without relying on special assumptions remains a nontrivial problem. Therefore, the main contribution of this paper is to provide an effective proof method to solve this problem. #### Main conclusion Through a series of complex analyses, the author finally proves that for a sufficiently large \(N\), indeed: \[ |L_\lambda(N)| < N - 1 \] This result not only answers the question raised at the 2018 AIM workshop but also provides new tools and methods for further research on the Liouville function and other related problems.