/- Copyright 2026 The Formal Conjectures Authors. Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at https://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. -/ module public import FormalConjecturesUtil

Simple zeros of the Riemann zeta function on the critical line

References:

    [Cl26] Claude (Anthropic), More than two thirds of the zeros of the Riemann zeta function are simple and on the critical line, August 2026. pdf, Lean formalisation.

    Wikipedia

For a nontrivial zero ρ = β + iγ of ζ (or of a Dirichlet L-function) with multiplicity m_ρ, and 0 ≤ T₁ < T₂, write

    N(T₁, T₂) = ∑_{T₁ < γ ≤ T₂} m_ρ for the zeros counted with multiplicity,

    N_d(T₁, T₂) = #{ρ : T₁ < γ ≤ T₂} for the distinct zeros, and

    Ns(T₁, T₂) = #{ρ : T₁ < γ ≤ T₂, β = 1/2, m_ρ = 1} for the simple zeros on the critical line.

[Cl26, Theorem A] proves unconditionally that Ns(T, 2T) ≥ (2/3 - o(1)) N(T, 2T) and N_d(T, 2T) ≥ (5/6 - o(1)) N(T, 2T), improving to 2 - c_MT⁻¹ = 0.6725… and (3 - c_MT⁻¹) / 2 = 0.8362… with the Montgomery–Taylor window, where c_MT⁻¹ = 1/2 + cot(1/√2)/√2; [Cl26, Theorem B] is the same for L(s, χ), χ primitive.

@[expose] public sectionopen Filternamespace ZetaZeros

The nontrivial zeros of f (those in the critical strip 0 < re s < 1) with imaginary part in (T₁, T₂].

def nontrivialZeros (f : ℂ → ℂ) (T₁ T₂ : ℝ) : Set ℂ := {s | f s = 0 ∧ 0 < s.re ∧ s.re < 1 ∧ T₁ < s.im ∧ s.im ≤ T₂}

N(T₁, T₂): the nontrivial zeros of f with imaginary part in (T₁, T₂], counted with multiplicity.

noncomputable def N (f : ℂ → ℂ) (T₁ T₂ : ℝ) : ℕ := ∑ᶠ s ∈ nontrivialZeros f T₁ T₂, (analyticOrderAt f s).toNat

N_d(T₁, T₂): the number of distinct nontrivial zeros of f with imaginary part in (T₁, T₂].

noncomputable def Nd (f : ℂ → ℂ) (T₁ T₂ : ℝ) : ℕ := (nontrivialZeros f T₁ T₂).ncard

Ns(T₁, T₂): the number of simple zeros of f on the critical line with imaginary part in (T₁, T₂].

noncomputable def Ns (f : ℂ → ℂ) (T₁ T₂ : ℝ) : ℕ := {s ∈ nontrivialZeros f T₁ T₂ | s.re = 1 / 2 ∧ analyticOrderAt f s = 1}.ncard

The Montgomery–Taylor constant c_MT = √2 tan(1/√2) / (1 + tan(1/√2)/√2), so that c_MT⁻¹ = 1/2 + cot(1/√2)/√2 = 1.3274….

noncomputable def montgomeryTaylorConst : ℝ := √2 * (1 / √2).tan / (1 + (1 / √2).tan / √2)end ZetaZerosopen ZetaZerosnamespace riemannZeta

At least two thirds of the nontrivial zeros of ζ, counted with multiplicity, are simple and on the critical line [Cl26, Theorem A (i)].

@[category research solved, AMS 11, formal_proof using lean4 at "https://github.com/anthropics/zeta-23-lean/blob/v1.0/Zeta23/FinalMult.lean#L350"] theorem simple_prop_two_thirds : ∃ o : ℝ → ℝ, o =o[atTop] (1 : ℝ → ℝ) ∧ ∀ T > 0, (2 / 3 - o T) * N riemannZeta T (2 * T) ≤ Ns riemannZeta T (2 * T) := ⊢ ∃ o, o =o[atTop] 1 ∧ ∀ T > 0, (2 / 3 - o T) * ↑(N riemannZeta T (2 * T)) ≤ ↑(Ns riemannZeta T (2 * T)) All goals completed! 🐙

At least five sixths of the nontrivial zeros of ζ, counted with multiplicity, are distinct [Cl26, Theorem A (ii)].

@[category research solved, AMS 11, formal_proof using lean4 at "https://github.com/anthropics/zeta-23-lean/blob/v1.0/Zeta23/FinalMult.lean#L358"] theorem distinct_prop_five_sixths : ∃ o : ℝ → ℝ, o =o[atTop] (1 : ℝ → ℝ) ∧ ∀ T > 0, (5 / 6 - o T) * N riemannZeta T (2 * T) ≤ Nd riemannZeta T (2 * T) := ⊢ ∃ o, o =o[atTop] 1 ∧ ∀ T > 0, (5 / 6 - o T) * ↑(N riemannZeta T (2 * T)) ≤ ↑(Nd riemannZeta T (2 * T)) All goals completed! 🐙

The proportion of simple zeros of ζ on the critical line is at least 2 - c_MT⁻¹ = 0.6725…. [Cl26, Theorem A]

@[category research solved, AMS 11, formal_proof using lean4 at "https://github.com/anthropics/zeta-23-lean/blob/v1.0/Zeta23/ThmD/Mult.lean#L435"] theorem simple_prop_montgomeryTaylor : ∃ o : ℝ → ℝ, o =o[atTop] (1 : ℝ → ℝ) ∧ ∀ T > 0, (2 - montgomeryTaylorConst⁻¹ - o T) * N riemannZeta T (2 * T) ≤ Ns riemannZeta T (2 * T) := ⊢ ∃ o, o =o[atTop] 1 ∧ ∀ T > 0, (2 - montgomeryTaylorConst⁻¹ - o T) * ↑(N riemannZeta T (2 * T)) ≤ ↑(Ns riemannZeta T (2 * T)) All goals completed! 🐙

The proportion of distinct zeros of ζ is at least (3 - c_MT⁻¹) / 2 = 0.8362…, [Cl26, Theorem A].

@[category research solved, AMS 11, formal_proof using lean4 at "https://github.com/anthropics/zeta-23-lean/blob/v1.0/Zeta23/ThmD/Mult.lean#L449"] theorem distinct_prop_montgomeryTaylor : ∃ o : ℝ → ℝ, o =o[atTop] (1 : ℝ → ℝ) ∧ ∀ T > 0, ((3 - montgomeryTaylorConst⁻¹) / 2 - o T) * N riemannZeta T (2 * T) ≤ Nd riemannZeta T (2 * T) := ⊢ ∃ o, o =o[atTop] 1 ∧ ∀ T > 0, ((3 - montgomeryTaylorConst⁻¹) / 2 - o T) * ↑(N riemannZeta T (2 * T)) ≤ ↑(Nd riemannZeta T (2 * T)) All goals completed! 🐙

Is the proportion of simple zeros of ζ on the critical line unconditionally larger than the Montgomery–Taylor bound 2 - c_MT⁻¹ = 0.6725…? [Cl26, §7.2] puts the ceiling of its method at about 0.682.

@[category research open, AMS 11] theorem simple_prop_lt : ∃ c > 2 - montgomeryTaylorConst⁻¹, ∃ o : ℝ → ℝ, o =o[atTop] (1 : ℝ → ℝ) ∧ ∀ T > 0, (c - o T) * N riemannZeta T (2 * T) ≤ Ns riemannZeta T (2 * T) := ⊢ ∃ c > 2 - montgomeryTaylorConst⁻¹, ∃ o, o =o[atTop] 1 ∧ ∀ T > 0, (c - o T) * ↑(N riemannZeta T (2 * T)) ≤ ↑(Ns riemannZeta T (2 * T)) All goals completed! 🐙

Almost all nontrivial zeros of ζ are simple and on the critical line.

@[category research open, AMS 11] theorem simple_prop_one_hundred : ∃ o : ℝ → ℝ, o =o[atTop] (1 : ℝ → ℝ) ∧ ∀ T > 0, (1 - o T) * N riemannZeta T (2 * T) ≤ Ns riemannZeta T (2 * T) := ⊢ ∃ o, o =o[atTop] 1 ∧ ∀ T > 0, (1 - o T) * ↑(N riemannZeta T (2 * T)) ≤ ↑(Ns riemannZeta T (2 * T)) All goals completed! 🐙

Is the proportion of distinct zeros of ζ unconditionally larger than the Montgomery–Taylor bound (3 - c_MT⁻¹) / 2 = 0.8362…?

@[category research open, AMS 11] theorem distinct_prop_lt : ∃ c > (3 - montgomeryTaylorConst⁻¹) / 2, ∃ o : ℝ → ℝ, o =o[atTop] (1 : ℝ → ℝ) ∧ ∀ T > 0, (c - o T) * N riemannZeta T (2 * T) ≤ Nd riemannZeta T (2 * T) := ⊢ ∃ c > (3 - montgomeryTaylorConst⁻¹) / 2, ∃ o, o =o[atTop] 1 ∧ ∀ T > 0, (c - o T) * ↑(N riemannZeta T (2 * T)) ≤ ↑(Nd riemannZeta T (2 * T)) All goals completed! 🐙

Almost all nontrivial zeros of ζ are simple.

@[category research open, AMS 11] theorem distinct_prop_one_hundred : ∃ o : ℝ → ℝ, o =o[atTop] (1 : ℝ → ℝ) ∧ ∀ T > 0, (1 - o T) * N riemannZeta T (2 * T) ≤ Nd riemannZeta T (2 * T) := ⊢ ∃ o, o =o[atTop] 1 ∧ ∀ T > 0, (1 - o T) * ↑(N riemannZeta T (2 * T)) ≤ ↑(Nd riemannZeta T (2 * T)) All goals completed! 🐙end riemannZetanamespace DirichletCharactervariable {M : ℕ} [NeZero M] {χ : DirichletCharacter ℂ M}

At least two thirds of the nontrivial zeros of L(s, χ), χ primitive, counted with multiplicity, are simple and on the critical line, [Cl26, Theorem B].

@[category research solved, AMS 11, formal_proof using lean4 at "https://github.com/anthropics/zeta-23-lean/blob/v1.0/Zeta23/ThmE/Mult.lean#L324"] theorem simple_prop_two_thirds (hχ : χ.IsPrimitive) : ∃ o : ℝ → ℝ, o =o[atTop] (1 : ℝ → ℝ) ∧ ∀ T > 0, (2 / 3 - o T) * N (LFunction χ) T (2 * T) ≤ Ns (LFunction χ) T (2 * T) := M:ℕinst✝:NeZero Mχ:DirichletCharacter ℂ Mhχ:χ.IsPrimitive⊢ ∃ o, o =o[atTop] 1 ∧ ∀ T > 0, (2 / 3 - o T) * ↑(N (LFunction χ) T (2 * T)) ≤ ↑(Ns (LFunction χ) T (2 * T)) All goals completed! 🐙

At least five sixths of the nontrivial zeros of L(s, χ), χ primitive, counted with multiplicity, are distinct, [Cl26, Theorem B].

@[category research solved, AMS 11, formal_proof using lean4 at "https://github.com/anthropics/zeta-23-lean/blob/v1.0/Zeta23/ThmE/Mult.lean#L333"] theorem distinct_prop_five_sixths (hχ : χ.IsPrimitive) : ∃ o : ℝ → ℝ, o =o[atTop] (1 : ℝ → ℝ) ∧ ∀ T > 0, (5 / 6 - o T) * N (LFunction χ) T (2 * T) ≤ Nd (LFunction χ) T (2 * T) := M:ℕinst✝:NeZero Mχ:DirichletCharacter ℂ Mhχ:χ.IsPrimitive⊢ ∃ o, o =o[atTop] 1 ∧ ∀ T > 0, (5 / 6 - o T) * ↑(N (LFunction χ) T (2 * T)) ≤ ↑(Nd (LFunction χ) T (2 * T)) All goals completed! 🐙

The proportion of simple zeros of L(s, χ), χ primitive, on the critical line is at least 2 - c_MT⁻¹ = 0.6725…, [Cl26, Theorem B].

@[category research solved, AMS 11, formal_proof using lean4 at "https://github.com/anthropics/zeta-23-lean/blob/v1.0/Zeta23/ThmDE/Mult.lean#L429"] theorem simple_prop_montgomeryTaylor (hχ : χ.IsPrimitive) : ∃ o : ℝ → ℝ, o =o[atTop] (1 : ℝ → ℝ) ∧ ∀ T > 0, (2 - montgomeryTaylorConst⁻¹ - o T) * N (LFunction χ) T (2 * T) ≤ Ns (LFunction χ) T (2 * T) := M:ℕinst✝:NeZero Mχ:DirichletCharacter ℂ Mhχ:χ.IsPrimitive⊢ ∃ o, o =o[atTop] 1 ∧ ∀ T > 0, (2 - montgomeryTaylorConst⁻¹ - o T) * ↑(N (LFunction χ) T (2 * T)) ≤ ↑(Ns (LFunction χ) T (2 * T)) All goals completed! 🐙

The proportion of distinct zeros of L(s, χ), χ primitive, is at least (3 - c_MT⁻¹) / 2 = 0.8362…, [Cl26, Theorem B].

@[category research solved, AMS 11, formal_proof using lean4 at "https://github.com/anthropics/zeta-23-lean/blob/v1.0/Zeta23/ThmDE/Mult.lean#L437"] theorem distinct_prop_montgomeryTaylor (hχ : χ.IsPrimitive) : ∃ o : ℝ → ℝ, o =o[atTop] (1 : ℝ → ℝ) ∧ ∀ T > 0, ((3 - montgomeryTaylorConst⁻¹) / 2 - o T) * N (LFunction χ) T (2 * T) ≤ Nd (LFunction χ) T (2 * T) := M:ℕinst✝:NeZero Mχ:DirichletCharacter ℂ Mhχ:χ.IsPrimitive⊢ ∃ o, o =o[atTop] 1 ∧ ∀ T > 0, ((3 - montgomeryTaylorConst⁻¹) / 2 - o T) * ↑(N (LFunction χ) T (2 * T)) ≤ ↑(Nd (LFunction χ) T (2 * T)) All goals completed! 🐙

Is the proportion of simple zeros of L(s, χ), χ primitive, on the critical line unconditionally larger than the Montgomery–Taylor bound 2 - c_MT⁻¹ = 0.6725…?

@[category research open, AMS 11] theorem simple_prop_lt (hχ : χ.IsPrimitive) : ∃ c > 2 - montgomeryTaylorConst⁻¹, ∃ o : ℝ → ℝ, o =o[atTop] (1 : ℝ → ℝ) ∧ ∀ T > 0, (c - o T) * N (LFunction χ) T (2 * T) ≤ Ns (LFunction χ) T (2 * T) := M:ℕinst✝:NeZero Mχ:DirichletCharacter ℂ Mhχ:χ.IsPrimitive⊢ ∃ c > 2 - montgomeryTaylorConst⁻¹, ∃ o, o =o[atTop] 1 ∧ ∀ T > 0, (c - o T) * ↑(N (LFunction χ) T (2 * T)) ≤ ↑(Ns (LFunction χ) T (2 * T)) All goals completed! 🐙

Almost all nontrivial zeros of L(s, χ), χ primitive, are simple and on the critical line.

@[category research open, AMS 11] theorem simple_prop_one_hundred (hχ : χ.IsPrimitive) : ∃ o : ℝ → ℝ, o =o[atTop] (1 : ℝ → ℝ) ∧ ∀ T > 0, (1 - o T) * N (LFunction χ) T (2 * T) ≤ Ns (LFunction χ) T (2 * T) := M:ℕinst✝:NeZero Mχ:DirichletCharacter ℂ Mhχ:χ.IsPrimitive⊢ ∃ o, o =o[atTop] 1 ∧ ∀ T > 0, (1 - o T) * ↑(N (LFunction χ) T (2 * T)) ≤ ↑(Ns (LFunction χ) T (2 * T)) All goals completed! 🐙

Is the proportion of distinct zeros of L(s, χ), χ primitive, unconditionally larger than the Montgomery–Taylor bound (3 - c_MT⁻¹) / 2 = 0.8362…?

@[category research open, AMS 11] theorem distinct_prop_lt (hχ : χ.IsPrimitive) : ∃ c > (3 - montgomeryTaylorConst⁻¹) / 2, ∃ o : ℝ → ℝ, o =o[atTop] (1 : ℝ → ℝ) ∧ ∀ T > 0, (c - o T) * N (LFunction χ) T (2 * T) ≤ Nd (LFunction χ) T (2 * T) := M:ℕinst✝:NeZero Mχ:DirichletCharacter ℂ Mhχ:χ.IsPrimitive⊢ ∃ c > (3 - montgomeryTaylorConst⁻¹) / 2, ∃ o, o =o[atTop] 1 ∧ ∀ T > 0, (c - o T) * ↑(N (LFunction χ) T (2 * T)) ≤ ↑(Nd (LFunction χ) T (2 * T)) All goals completed! 🐙

Almost all nontrivial zeros of L(s, χ), χ primitive, are simple.

@[category research open, AMS 11] theorem distinct_prop_one_hundred (hχ : χ.IsPrimitive) : ∃ o : ℝ → ℝ, o =o[atTop] (1 : ℝ → ℝ) ∧ ∀ T > 0, (1 - o T) * N (LFunction χ) T (2 * T) ≤ Nd (LFunction χ) T (2 * T) := M:ℕinst✝:NeZero Mχ:DirichletCharacter ℂ Mhχ:χ.IsPrimitive⊢ ∃ o, o =o[atTop] 1 ∧ ∀ T > 0, (1 - o T) * ↑(N (LFunction χ) T (2 * T)) ≤ ↑(Nd (LFunction χ) T (2 * T)) All goals completed! 🐙end DirichletCharacter