Eventually subpower prime energies at doubly exponential arithmetic cutoffs
OpenMasulloRH.sparse_subpower_square_prime_energyexplicit-formulaprime-countingriemann-hypothesis
For all sufficiently large j, the square-sampled finite prime energy E(D_j) at D_j=2^(2^j) is at most D_j^(1/(j+1)). This single growing allowance is checked at sparse finite cutoffs. E(N) is the sum over k=4,...,N+3 of G(k²)²/[k³(1+log k)²] for the corrected von Mangoldt ramp.
Preamble
import Definitions.Def_MasulloRH_RegularizedTrace import Mathlib set_option autoImplicit false
Formal statement
open Finset ArithmeticFunction MeasureTheory Set
namespace MasulloRH
theorem sparse_subpower_square_prime_energy :
let G (x : ℝ) := (∑ n ∈ range (⌊x⌋₊+1),vonMangoldt n*max (x/(n:ℝ)-1) 0) -
x*(Real.log x-1-Real.eulerMascheroniConstant)
let E (N : ℕ) := ∑ n ∈ range N, let k := (n:ℝ)+4; (G (k^2))^2/(k^3*(1+Real.log k)^2)
∃ J : ℕ,∀ j ≥ J, let D := (2:ℕ)^(2^j)
E D ≤ (D:ℝ)^(1/((j:ℝ)+1)) := by sorry
end MasulloRHSource
Open finite-arithmetic input for b41359f9. It relaxes the all-cutoff constant 81 to an unbounded allowance at doubly exponential cutoffs. The converse uses the actual continuous prime-ramp energy and the Dirichlet series of the nonnegative sampled coefficients.