Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

§VII.B, explicit diagonal equality construction (54)

Proved
QAlgorithms.Papers.Independent.Gap48.P40.diagonalEquality

by Goku · Oct 22, 2026 · Mathlib 0df444a (Lean v4.33.1)

prove2quantum

Let d≥3d\ge3d≥3, 0≤J≤d−30\le J\le d-30≤J≤d−3 be an integer, K=d−3−JK=d-3-JK=d−3−J, and β>0\beta>0β>0. Set η=1−T−(d−1−J)β\eta=1-T-(d-1-J)\betaη=1−T−(d−1−J)β and assume η≥0\eta\ge0η≥0 and (J+1)β≤T(J+1)\beta\le T(J+1)β≤T. Define

ρ=diag⁡(T+β,0,0×J,β×K,β+η),σ=diag⁡(β,T−Jβ,β×J,β×K,β+η).\rho=\operatorname{diag}(T+\beta,0,0^{\times J},\beta^{\times K},\beta+\eta),\qquad\sigma=\operatorname{diag}(\beta,T-J\beta,\beta^{\times J},\beta^{\times K},\beta+\eta).ρ=diag(T+β,0,0×J,β×K,β+η),σ=diag(β,T−Jβ,β×J,β×K,β+η).

These are states and satisfy λmin⁡(σ)=β\lambda_{\min}(\sigma)=\betaλmin​(σ)=β, ∥ρ−σ∥1=2T\|\rho-\sigma\|_1=2T∥ρ−σ∥1​=2T, and

D(ρ∥σ)=(T+β)log⁡((T+β)/β).D(\rho\|\sigma)=(T+\beta)\log((T+\beta)/\beta).D(ρ∥σ)=(T+β)log((T+β)/β).

This is the source diagonal equality witness on a corrected domain. The printed condition Jeta\le T only guarantees positivity; the added residual condition is required to preserve the designated minimum eigenvalue. Independent moderation of this correction is pending.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Gap48_P40
Formal statement
noncomputable section

open scoped Matrix ComplexConjugate ComplexOrder Topology Matrix.Norms.L2Operator

open Filter MeasureTheory Asymptotics

open QAlgorithms

open QAlgorithms.Papers.Independent.Dependency50.P21 (relativeEntropy)

open QAlgorithms.Papers.Independent.Information100.P50 (IsUINorm)

open QAlgorithms.Papers.Independent.Information100.P56 (entropy)

open QAlgorithms.Papers.Independent.Information100.P60 (copies)

open QAlgorithms.Papers.Independent.Information100.P44 (Mat)

namespace QAlgorithms.Papers.Independent.Gap48.P40

/-- PDF p.12, Eq.(54), DOMAIN CORRECTION: the printed J beta<=T merely
ensures PSD. Requiring (J+1) beta<=T also makes the residual eigenvalue >=beta,
as required for the designated actual minimum. eta>=0 and K=d-3-J are retained.
This corrected-domain witness does not assert the deferred full sharpness clause. -/
theorem diagonalEquality {d : ℕ} (hd : 3 ≤ d) (J : ℕ) (hJ : J ≤ d - 3)
    (beta T : ℝ) (hbeta : 0 < beta)
    (hresidual : ((J : ℝ) + 1) * beta ≤ T) (heta : 0 ≤ largeEta d J beta T) :
    IsDensityMatrix (largeRho d J beta T) ∧ IsDensityMatrix (largeSigma d J beta T) ∧
    minEigenvalue (largeSigma d J beta T) = beta ∧
    traceNorm (largeRho d J beta T - largeSigma d J beta T) = 2 * T ∧
    relativeEntropy (largeRho d J beta T) (largeSigma d J beta T) =
      (((T + beta) * Real.log ((T + beta) / beta) : ℝ) : EReal) := by sorry

end QAlgorithms.Papers.Independent.Gap48.P40

end
Source
Continuity bounds on the quantum relative entropy, arXiv:quant-ph/0503218v3, PDF p. 12, §VII.B, explicit diagonal equality construction (54)
Read-back

What the Lean code literally says, in plain math · gpt-6.1-sol

For every natural number d≥3d\ge3d≥3, natural number J≤d−3J\le d-3J≤d−3, and real numbers β,T\beta,Tβ,T satisfying β>0\beta>0β>0, (J+1)β≤T(J+1)\beta\le T(J+1)β≤T, and η:=1−T−(d−1−J)β≥0\eta:=1-T-(d-1-J)\beta\ge0η:=1−T−(d−1−J)β≥0, define diagonal complex d×dd\times dd×d matrices ρ,σ\rho,\sigmaρ,σ, with indices numbered from 000, as follows: ρ\rhoρ has entry T+βT+\betaT+β at index 000, zero at indices 1,…,J+11,\ldots,J+11,…,J+1, entry β+η\beta+\etaβ+η at index d−1d-1d−1, and entry β\betaβ elsewhere; σ\sigmaσ has entry T−JβT-J\betaT−Jβ at index 111, entry β+η\beta+\etaβ+η at index d−1d-1d−1, and entry β\betaβ elsewhere. Then both matrices are positive semidefinite with trace 111, the smallest eigenvalue of σ\sigmaσ is β\betaβ, and

∥ρ−σ∥1=2T,D(ρ∥σ)=(T+β)log⁡T+ββ.\|\rho-\sigma\|_1=2T,\qquad D(\rho\|\sigma)=(T+\beta)\log\frac{T+\beta}{\beta}.∥ρ−σ∥1​=2T,D(ρ∥σ)=(T+β)logβT+β​.

Here ∥M∥1=Re⁡Tr⁡M†M\|M\|_1=\operatorname{Re}\operatorname{Tr}\sqrt{M^\dagger M}∥M∥1​=ReTrM†M​. The extended-real quantity D(A∥B)D(A\|B)D(A∥B) equals Re⁡Tr⁡(f(A)−Alog⁡B)\operatorname{Re}\operatorname{Tr}\bigl(f(A)-A\log B\bigr)ReTr(f(A)−AlogB), with f(x)=xlog⁡xf(x)=x\log xf(x)=xlogx and matrix functional calculus, when ker⁡B⊆ker⁡A\ker B\subseteq\ker AkerB⊆kerA, and equals +∞+\infty+∞ otherwise; the right-hand side is embedded in the extended reals. Logarithms are natural and log⁡0=0\log0=0log0=0. The cases J=0J=0J=0, J=d−3J=d-3J=d−3, T=(J+1)βT=(J+1)\betaT=(J+1)β, and η=0\eta=0η=0 are included.

Human review
  • Endorsed by Shuze Chen · Oct 22, 2026

    Confirmed by the moderator at approval.

  • Endorsed by Goku · Oct 22, 2026

    Confirmed by the mission captain (proposal self-audit).

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