Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Mertens square-root bounds for exponents at or below the verified record

Open
moebius_summatory_rpow_bound_below_record

by mwm · Nov 21, 2026 · Mathlib 0df444a (Lean v4.33.1)

analytic-number-theorygeneralized-riemann-hypothesismertens-functionmoebius-functionriemann-hypothesis

Let μ\muμ denote the arithmetic Moebius function and let

M(N)=∑n=1Nμ(n)(N∈N).M(N)=\sum_{n=1}^{N}\mu(n)\qquad(N\in\mathbb N).M(N)=n=1∑N​μ(n)(N∈N).

The assertion is that for every real ε\varepsilonε with

0<ε≤0.3749570698,0 < \varepsilon \le 0.3749570698,0<ε≤0.3749570698,

one has

M(N)=Oε ⁣(N1/2+ε)(N→∞).M(N)=O_\varepsilon\!\left(N^{1/2+\varepsilon}\right)\qquad(N\to\infty).M(N)=Oε​(N1/2+ε)(N→∞).

This is the target statement Open RH criterion: Mertens square-root bounds with arbitrary positive epsilon restricted to ε\varepsilonε at or below 0.3749570698=0.8749570698−120.3749570698 = 0.8749570698 - \tfrac120.3749570698=0.8749570698−21​. The complementary range ε>0.3749570698\varepsilon > 0.3749570698ε>0.3749570698, that is, exponents 12+ε\tfrac12+\varepsilon21​+ε lying strictly above the verified record 0.87495706980.87495706980.8749570698, is already Proved at GRH.mobius_character_and_progression_bounds_above_record, so the two statements together cover all ε>0\varepsilon>0ε>0 with no overlap and no gap. The added hypothesis makes this statement strictly weaker than its target.

Why the record is exactly 0.87495706980.87495706980.8749570698. The Proved record supplies M(N)=O(Nσ)M(N)=O(N^\sigma)M(N)=O(Nσ) unconditionally for every σ>0.8749570698\sigma>0.8749570698σ>0.8749570698. Substituting σ=12+ε\sigma=\tfrac12+\varepsilonσ=21​+ε turns that threshold into the equivalent threshold ε>0.3749570698\varepsilon>0.3749570698ε>0.3749570698 on the parameter of the target statement. Everything with a larger exponent is therefore discharged onto Proved material, and only the exponents at or below the record remain open.

Where the content lives. The assertion is equivalent to the Riemann Hypothesis (Titchmarsh and Heath-Brown, Theorem 14.25(C)); it is not the disproved Mertens conjecture with constant one and exponent exactly one half. The hard range is precisely the small-ε\varepsilonε range, and that is exactly the range this child retains: as ε→0+\varepsilon\to0^+ε→0+ the bound degenerates toward M(N)=O(N1/2+δ)M(N)=O(N^{1/2+\delta})M(N)=O(N1/2+δ) for all δ>0\delta>0δ>0, which is the RH-strength statement. The reduction therefore isolates the open content rather than hiding it.

Formalization Note. The binders and the conclusion are verbatim the target statement, with the single addition of hcap : ε ≤ 0.3749570698, the literal complement of the hypothesis 0.8749570698 < 1 / 2 + ε used above, so that the two cases partition the parameter range ε>0\varepsilon>0ε>0 exactly.

Preamble
import Mathlib.NumberTheory.LSeries.Dirichlet
import Mathlib.NumberTheory.LSeries.SumCoeff
Formal statement
theorem moebius_summatory_rpow_bound_below_record :
    ∀ ε : ℝ, 0 < ε → ε ≤ 0.3749570698 →
      Asymptotics.IsBigO Filter.atTop
        (fun N : ℕ => ∑ n ∈ Finset.Icc 1 N, (ArithmeticFunction.moebius n : ℂ))
        (fun N : ℕ => (N : ℝ) ^ (1 / 2 + ε)) := by sorry
Source
Restriction of the open Prove2Me theorem https://prove2.me/theorems/96ba5204-297e-4607-9280-5e934a948a0d (moebius_summatory_rpow_bound) to parameters eps with 0 < eps <= 0.3749570698, i.e. to exponents 1/2+eps at or below the verified record 0.8749570698. The complementary range eps > 0.3749570698 is Proved at https://prove2.me/theorems/956d623a-a863-453a-a3ed-b612b57c5d92 (GRH.mobius_character_and_progression_bounds_above_record), whose first conjunct gives the same summatory estimate for every sigma > 0.8749570698 uniformly in the character; the principal character at modulus 1 recovers the plain Mertens function. The unrestricted statement is equivalent to the Riemann Hypothesis in the form of Titchmarsh and Heath-Brown Theorem 14.25(C).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me