Residual Galois representations of eigenforms at a p-adic place
DefinitionKN_EigenformResidualGaloisRepresentationgalois-representationsmodular-formsnumber-theoryp-adic-l-functions
Let be a normalized cuspidal eigenform of level and weight , and fix an embedding of its algebraic coefficient field into . The data structure records a coefficient ring integral at this place, a surjective reduction map to a finite field of characteristic , and a faithful two-dimensional representation
for a finite Galois extension . It requires to be unramified outside and, for every prime , requires
The kernel of the reduction map is characterized by strict -adic norm less than one. This makes the reduction congruence available without choosing an ad hoc total map from to a finite field. Faithfulness says that is the field cut out by the residual representation, so the representation has finite image for the intended reason.
Definition code
import Definitions.Def_KN_SeededPrimeGaloisData
set_option autoImplicit false
noncomputable section
namespace HorizontalPadicL
open NumberField Ideal FrobeniusDensity
/-- A textbook residual Galois representation attached to a normalized
eigenform at a chosen `p`-adic place.
The subring `A` is a ring of coefficients integral at the chosen place and
`reduction` is its map to a finite residue field of characteristic `p`. The
finite Galois extension `L / ℚ` is the extension cut out by the representation:
faithfulness says that no redundant extension has been included. Thus the
matrix representation has finite image. The last three fields express that
it is unramified outside `N * p` and has the usual Frobenius trace and
determinant. -/
inductive EigenformResidualGaloisRepresentationData
{N k p : ℕ} {ι : MTT.Qbar →+* ℂ} [Fact p.Prime]
(f : MTT.Eigenform N k ι) (ιp : MTT.Qbar →+* ℂ_[p]) : Type 2
| ofRepresentation
(A : Subring MTT.Qbar)
(𝕜 L : Type)
[Field 𝕜] [Fintype 𝕜]
[Field L] [NumberField L] [IsGalois ℚ L]
(reduction : A →+* 𝕜)
(ρ : TaylorWiles.ResidualRep L 𝕜)
(residueCharacteristic : ringChar 𝕜 = p)
(reduction_surjective : Function.Surjective reduction)
(coeff_mem : ∀ n : ℕ, f.coeff n ∈ A)
(nebentype_mem : ∀ n : ℕ, f.epsilon n ∈ A)
(integral_at_place : ∀ x : A, ‖ιp (x : MTT.Qbar)‖ ≤ 1)
(reduction_eq_zero_iff : ∀ x : A,
reduction x = 0 ↔ ‖ιp (x : MTT.Qbar)‖ < 1)
(faithful : Function.Injective ρ)
(unramified_outside : ∀ ⦃l : ℕ⦄, l.Prime → Nat.Coprime l (N * p) →
∀ (Q : Ideal (𝓞 L)), Q.IsPrime → Q.LiesOver (ratPrimeIdeal l) →
Q.inertia (L ≃ₐ[ℚ] L) = ⊥)
(trace_frobenius : ∀ ⦃l : ℕ⦄, (hl : l.Prime) → Nat.Coprime l (N * p) →
∀ (Q : Ideal (𝓞 L)) (_ : Q.IsPrime)
(_ : Q.LiesOver (ratPrimeIdeal l)),
haveI : Finite ((𝓞 L) ⧸ Q) :=
finite_quotient_of_ne_bot (ne_bot_of_liesOver_ratPrimeIdeal hl)
Matrix.trace (ρ (arithFrobAt ℤ (L ≃ₐ[ℚ] L) Q)) =
reduction ⟨f.coeff l, coeff_mem l⟩)
(det_frobenius : ∀ ⦃l : ℕ⦄, (hl : l.Prime) → Nat.Coprime l (N * p) →
∀ (Q : Ideal (𝓞 L)) (_ : Q.IsPrime)
(_ : Q.LiesOver (ratPrimeIdeal l)),
haveI : Finite ((𝓞 L) ⧸ Q) :=
finite_quotient_of_ne_bot (ne_bot_of_liesOver_ratPrimeIdeal hl)
Matrix.det (ρ (arithFrobAt ℤ (L ≃ₐ[ℚ] L) Q)) =
reduction ⟨f.epsilon l * (l : MTT.Qbar) ^ (k - 1),
A.mul_mem (nebentype_mem l)
(A.pow_mem (natCast_mem A l) (k - 1))⟩) :
EigenformResidualGaloisRepresentationData f ιp
end HorizontalPadicL
Source
Kriz--Nordentoft, Horizontal p-adic L-functions, arXiv:2310.20678v3, Section 4, p. 27, equation (4.2), https://arxiv.org/pdf/2310.20678