Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

TℓT_\ellTℓ​ (and UℓU_\ellUℓ​ for ℓ∣N\ell\mid Nℓ∣N) preserves the cusp forms of nebentypus ε\varepsilonε on Γ1(N)\Gamma_1(N)Γ1​(N)

Proved
MTT.exists_cuspForm_heckePrime_pos

by cbirkbeck · Sep 6, 2026 · Mathlib 0df444a (Lean v4.33.1)

congruence-subgroupshecke-operatorsmodular-forms

Let N≥1N\ge1N≥1, k≥1k\ge1k≥1, let ε\varepsilonε be a Dirichlet character modulo NNN with values in C\mathbf CC, and let g∈Sk(Γ1(N))g\in S_k(\Gamma_1(N))g∈Sk​(Γ1​(N)) have nebentypus ε\varepsilonε:

g(γz)=ε(d) (cz+d)k g(z)for all γ=(abcd)∈Γ0(N), z∈H.g(\gamma z)=\varepsilon(d)\,(cz+d)^k\,g(z)\qquad\text{for all }\gamma=\begin{pmatrix}a&b\\ c&d\end{pmatrix}\in\Gamma_0(N),\ z\in\mathfrak H .g(γz)=ε(d)(cz+d)kg(z)for all γ=(ac​bd​)∈Γ0​(N), z∈H.

For a prime ℓ\ellℓ define, as in the mission,

(Tℓg)(z)=1ℓ∑b=0ℓ−1g(z+bℓ)+ε(ℓ) ℓk−1 g(ℓz).(T_\ell g)(z)=\frac1\ell\sum_{b=0}^{\ell-1}g\Big(\frac{z+b}{\ell}\Big)+\varepsilon(\ell)\,\ell^{k-1}\,g(\ell z).(Tℓ​g)(z)=ℓ1​b=0∑ℓ−1​g(ℓz+b​)+ε(ℓ)ℓk−1g(ℓz).

When ℓ∤N\ell\nmid Nℓ∤N this is the classical Hecke operator TℓT_\ellTℓ​ on Sk(N,ε)S_k(N,\varepsilon)Sk​(N,ε); when ℓ∣N\ell\mid Nℓ∣N one has ε(ℓ)=0\varepsilon(\ell)=0ε(ℓ)=0 and it is the operator UℓU_\ellUℓ​.

Claim. TℓgT_\ell gTℓ​g is again a cusp form of weight kkk on Γ1(N)\Gamma_1(N)Γ1​(N) with nebentypus ε\varepsilonε. Formally: there is g′∈Sk(Γ1(N))g'\in S_k(\Gamma_1(N))g′∈Sk​(Γ1​(N)) with g′(z)=(Tℓg)(z)g'(z)=(T_\ell g)(z)g′(z)=(Tℓ​g)(z) for every zzz, and g′(γz)=ε(d)(cz+d)kg′(z)g'(\gamma z)=\varepsilon(d)(cz+d)^kg'(z)g′(γz)=ε(d)(cz+d)kg′(z) for all γ∈Γ0(N)\gamma\in\Gamma_0(N)γ∈Γ0​(N).

Why this holds. For ℓ∤N\ell\nmid Nℓ∤N the double coset Γ1(N)(100ℓ)Γ1(N)\Gamma_1(N)\begin{pmatrix}1&0\\0&\ell\end{pmatrix}\Gamma_1(N)Γ1​(N)(10​0ℓ​)Γ1​(N) is the disjoint union of the cosets Γ1(N)(1b0ℓ)\Gamma_1(N)\begin{pmatrix}1&b\\0&\ell\end{pmatrix}Γ1​(N)(10​bℓ​), 0≤b<ℓ0\le b<\ell0≤b<ℓ, and Γ1(N) σℓ(ℓ001)\Gamma_1(N)\,\sigma_\ell\begin{pmatrix}\ell&0\\0&1\end{pmatrix}Γ1​(N)σℓ​(ℓ0​01​) with σℓ∈SL2(Z)\sigma_\ell\in SL_2(\mathbf Z)σℓ​∈SL2​(Z), σℓ≡diag⁡(ℓ−1,ℓ)(modN)\sigma_\ell\equiv\operatorname{diag}(\ell^{-1},\ell)\pmod Nσℓ​≡diag(ℓ−1,ℓ)(modN) (Diamond–Shurman, Proposition 5.2.1); for ℓ∣N\ell\mid Nℓ∣N the last coset is absent. On forms of nebentypus ε\varepsilonε the diamond operator ⟨ℓ⟩\langle\ell\rangle⟨ℓ⟩ acts as ε(ℓ)\varepsilon(\ell)ε(ℓ), which turns the coset sum into the displayed formula (Diamond–Shurman, Proposition 5.3.1 and (5.3)). Directly: right multiplication by γ∈Γ0(N)\gamma\in\Gamma_0(N)γ∈Γ0​(N) permutes the ℓ+1\ell+1ℓ+1 matrices (1b0ℓ)\begin{pmatrix}1&b\\0&\ell\end{pmatrix}(10​bℓ​), (ℓ001)\begin{pmatrix}\ell&0\\0&1\end{pmatrix}(ℓ0​01​) up to left factors in Γ0(N)\Gamma_0(N)Γ0​(N) whose lower-right entries are congruent to ddd, ℓd\ell dℓd or ℓ−1d\ell^{-1}dℓ−1d modulo NNN, and the resulting factors ε(ℓ)±1\varepsilon(\ell)^{\pm1}ε(ℓ)±1 are absorbed exactly by the coefficient ε(ℓ)\varepsilon(\ell)ε(ℓ) in the definition. Each summand is holomorphic, and vanishes at every cusp since it is a slash of ggg by an element of GL2+(Q)GL_2^+(\mathbf Q)GL2+​(Q), so TℓgT_\ell gTℓ​g is a cusp form.

Remark. The hypothesis k≥1k\ge1k≥1 is needed only because the mission's heckePrime writes the normalisation ℓk−1\ell^{k-1}ℓk−1 with truncated natural-number subtraction, which for k=0k=0k=0 no longer matches the weight-kkk slash operator; all weights used in the mission satisfy k≥2k\ge2k≥2. This corrects the earlier node MTT.exists_cuspForm_heckePrime, stated without this hypothesis.

Preamble
import Definitions.Def_MTT_Arithmetic
set_option autoImplicit false
noncomputable section
open scoped BigOperators
Formal statement
theorem MTT.exists_cuspForm_heckePrime_pos
    {N k : ℕ} (hN : 0 < N) (hk : 1 ≤ k) (e : DirichletCharacter ℂ N)
    (g : CuspForm (MTT.GammaOne N) (k : ℤ))
    (hg : ∀ γ : CongruenceSubgroup.Gamma0 N, ∀ z : UpperHalfPlane,
      g ((Matrix.SpecialLinearGroup.mapGL ℝ γ.val) • z) =
        e (γ.val 1 1 : ZMod N) *
          (((γ.val 1 0 : ℤ) : ℂ) * z + ((γ.val 1 1 : ℤ) : ℂ)) ^ k * g z)
    (l : ℕ) (hl : l.Prime) :
    ∃ g' : CuspForm (MTT.GammaOne N) (k : ℤ),
      (∀ z, g' z = MTT.heckePrime k (e (l : ZMod N)) l g z) ∧
      ∀ γ : CongruenceSubgroup.Gamma0 N, ∀ z : UpperHalfPlane,
        g' ((Matrix.SpecialLinearGroup.mapGL ℝ γ.val) • z) =
          e (γ.val 1 1 : ZMod N) *
            (((γ.val 1 0 : ℤ) : ℂ) * z + ((γ.val 1 1 : ℤ) : ℂ)) ^ k * g' z := by sorry
Source
Diamond–Shurman, A First Course in Modular Forms (GTM 228), §5.2, Proposition 5.2.1 and Proposition 5.2.2 (pp. 169–172), and §5.3, Proposition 5.3.1 with (5.3) (pp. 179–180). Formal proof ported from: AINTLIB, projects/LeanModularForms/LeanModularForms/HeckeRIngs/GL2/HeckeT_p.lean (heckeT_p) and HeckeT_p_Gamma1.lean (heckeT_p_fun_eq_coset_sum).

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