Nontrivial Riemann Zeros as Spectrum
Enderalp Yakaboylu
TL;DR
The paper presents a non-symmetric Riemann operator $\hat{\mathcal{R}}$ on $L^2([0,\infty))$ whose spectrum encodes the zeros of the completed eta function, including the nontrivial zeta zeros. By constructing a positive intertwiner $\hat{W}$ (via a regularized $\hat{V}_R$ and oblique projections), the authors obtain a Hilbert–Pólya-type framework in which $\hat{W}\ge0$ forces $\Re(\rho)=1/2$ for all nontrivial zeros $\rho$, thereby linking positivity to the Riemann Hypothesis. A self-adjoint Hilbert–Pólya operator $\hat{h}=\hat{W}^{1/2}\hat{\mathcal{R}}\hat{W}^{-1/2}$ is constructed, with spectrum equal to the imaginary parts of the zeros, providing a spectral realization of RH. The framework is extended to higher-order zeros and generalized to Mellin-transformable $L$-functions, suggesting a broader operator-theoretic pathway to the Generalized Riemann Hypothesis via positivity and intertwiners.
Abstract
Define $ Υ(s) := Γ(s+1)\, (1-2^{1-s}) \, ζ(s) $, and let $ Z := \left\{γ\in \mathbb{C} \;\middle|\; Υ(γ)=0 \right\} $ denote its zero set, consisting of the nontrivial zeta zeros $ρ$, together with the zeros of $ (1-2^{1-s}) $, excluding $ s = 1 $. We introduce a non-symmetric operator \[ R \colon D(R) \subset L^2([0,\infty)) \to L^2([0,\infty)) \, , \] with spectrum \[ σ(R) = \left\{ i\left(1/2- γ\right) \;\middle|\; γ\in Z \right\} \, . \] Assuming simplicity of all nontrivial zeta zeros, we construct a positive semidefinite operator $ W $ that intertwines $ R$ and its adjoint on the corresponding spectral subspace, $ R^\dagger W = W R $. The positivity of $ W $, which represents an operator-theoretic form of (Bombieri's refinement of) Weil's positivity criterion, enforces $ \Re(ρ)=1/2 $ for all $ ρ$, in accordance with the Riemann Hypothesis. The same positivity condition naturally yields a self-adjoint operator whose spectrum coincides with the imaginary parts of the nontrivial zeta zeros. We further extend the framework to accommodate higher-order zeta zeros, should they exist, and observe that it generalizes to any Mellin-transformable $L$-function satisfying a functional equation.
