Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Sharp relative-entropy bound for unnormalised equal-trace matrices

Open
QAlgorithms.Papers.Independent.Support50.P17.proposition1

by Goku · Oct 23, 2026 · Mathlib 0df444a (Lean v4.33.1)

prove2quantum

For α,T≥0\alpha,T\ge0α,T≥0, define

q(α,T)={0,α=0,−αlog⁡(1+T/α),α>0.q(\alpha,T)=\begin{cases}0,&\alpha=0,\\-\alpha\log(1+T/\alpha),&\alpha>0.\end{cases}q(α,T)={0,−αlog(1+T/α),​α=0,α>0.​

Let A≥0A\ge0A≥0 and B>0B>0B>0 be matrices on the same nonzero finite-dimensional complex Hilbert space, with equal trace and λmin⁡(B)=1\lambda_{\min}(B)=1λmin​(B)=1. Put α=λmin⁡(A)\alpha=\lambda_{\min}(A)α=λmin​(A) and T=12∥A−B∥1T=\tfrac12\|A-B\|_1T=21​∥A−B∥1​. Then

T≥∣α−1∣,S(A∥B)≤(1+T)log⁡(1+T)+q(α,T).T\ge|\alpha-1|,\qquad S(A\|B)\le(1+T)\log(1+T)+q(\alpha,T).T≥∣α−1∣,S(A∥B)≤(1+T)log(1+T)+q(α,T).

For each fixed T≥0T\ge0T≥0, the function α↦q(α,T)\alpha\mapsto q(\alpha,T)α↦q(α,T) is nonincreasing on [0,∞)[0,\infty)[0,∞). For every α,T≥0\alpha,T\ge0α,T≥0 satisfying T≥∣α−1∣T\ge|\alpha-1|T≥∣α−1∣, the explicit matrices

A=Diag⁡(1+T,α),B=Diag⁡(1,T+α)A=\operatorname{Diag}(1+T,\alpha),\qquad B=\operatorname{Diag}(1,T+\alpha)A=Diag(1+T,α),B=Diag(1,T+α)

have the specified traces, minima and trace distance, and attain equality in the bound. Moreover A>0A>0A>0 when α>0\alpha>0α>0.

This includes the positive-definite proposition and its explicitly justified semidefinite boundary, without imposing trace one or requiring attainment in each fixed dimension.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Support50_P17
Formal statement
noncomputable section

open MeasureTheory

open scoped Matrix ComplexOrder Matrix.Norms.L2Operator

namespace QAlgorithms.Papers.Independent.Support50.P17

attribute [local instance] Classical.propDecidable Classical.decEq

-- Reuse the exact frozen support predicate and signed, unnormalised entropy.

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

/-- PDF pp.6–8, Proposition 1, (13): the sharp unnormalised equal-trace bound,
its explicit alpha=0 extension, nonincreasing correction, and the exact 2x2 attaining pair.
The printed “quality” on p.6 means equality, as shown by the p.8 construction. -/
theorem proposition1 :
    (∀ {I : Type*} [Fintype I] [DecidableEq I] [Nonempty I]
      (A B : Matrix I I ℂ), A.PosSemidef → B.PosDef →
      A.trace = B.trace → minEigenvalue B = 1 →
      let alpha := minEigenvalue A
      let T := traceDistance A B
      |alpha - 1| ≤ T ∧
        relativeEntropy A B ≤
          (((1 + T) * Real.log (1 + T) + zeroExtendedLogCorrection alpha T : ℝ) : EReal)) ∧
    (∀ T : ℝ, 0 ≤ T →
      AntitoneOn (fun alpha => zeroExtendedLogCorrection alpha T) (Set.Ici 0)) ∧
    (∀ alpha T : ℝ, 0 ≤ alpha → 0 ≤ T → |alpha - 1| ≤ T →
      let A := propositionA alpha T
      let B := propositionB alpha T
      A.PosSemidef ∧ B.PosDef ∧ (0 < alpha → A.PosDef) ∧
      A.trace = B.trace ∧ minEigenvalue A = alpha ∧ minEigenvalue B = 1 ∧
      traceDistance A B = T ∧
      relativeEntropy A B =
        (((1 + T) * Real.log (1 + T) + zeroExtendedLogCorrection alpha T : ℝ) : EReal)) := by sorry

end QAlgorithms.Papers.Independent.Support50.P17

end
Source
Continuity bounds on the quantum relative entropy - II, arXiv:1105.2656v3, PDF p. 6, Proposition 1: sharp equal-trace unnormalised bound (13)
Read-back

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

Define g(α,T)=0g(\alpha,T)=0g(α,T)=0 if α=0\alpha=0α=0 and g(α,T)=−αlog⁡(1+T/α)g(\alpha,T)=-\alpha\log(1+T/\alpha)g(α,T)=−αlog(1+T/α) otherwise, and define dtr(A,B)=12Re⁡tr⁡(A−B)∗(A−B)d_{\mathrm{tr}}(A,B)=\tfrac12\operatorname{Re}\operatorname{tr}\sqrt{(A-B)^*(A-B)}dtr​(A,B)=21​Retr(A−B)∗(A−B)​. The theorem asserts three statements jointly: for every finite, nonempty index type III with decidable equality and complex matrices A,BA,BA,B indexed by III, if AAA is positive semidefinite, BBB is positive definite, their traces are equal, and λmin⁡(B)=1\lambda_{\min}(B)=1λmin​(B)=1, then, setting α=λmin⁡(A)\alpha=\lambda_{\min}(A)α=λmin​(A) and T=dtr(A,B)T=d_{\mathrm{tr}}(A,B)T=dtr​(A,B), one has ∣α−1∣≤T|\alpha-1|\le T∣α−1∣≤T and D(A∥B)≤(1+T)log⁡(1+T)+g(α,T)D(A\|B)\le(1+T)\log(1+T)+g(\alpha,T)D(A∥B)≤(1+T)log(1+T)+g(α,T); for every real T≥0T\ge0T≥0, the function α↦g(α,T)\alpha\mapsto g(\alpha,T)α↦g(α,T) is nonincreasing on [0,∞)[0,\infty)[0,∞), meaning 0≤a≤b0\le a\le b0≤a≤b implies g(b,T)≤g(a,T)g(b,T)\le g(a,T)g(b,T)≤g(a,T); and for every real α,T≥0\alpha,T\ge0α,T≥0 satisfying ∣α−1∣≤T|\alpha-1|\le T∣α−1∣≤T, the matrices A=diag⁡(1+T,α)A=\operatorname{diag}(1+T,\alpha)A=diag(1+T,α) and B=diag⁡(1,T+α)B=\operatorname{diag}(1,T+\alpha)B=diag(1,T+α) satisfy A≥0A\ge0A≥0, B>0B>0B>0, A>0A>0A>0 whenever α>0\alpha>0α>0, equal traces, λmin⁡(A)=α\lambda_{\min}(A)=\alphaλmin​(A)=α, λmin⁡(B)=1\lambda_{\min}(B)=1λmin​(B)=1, dtr(A,B)=Td_{\mathrm{tr}}(A,B)=Tdtr​(A,B)=T, and D(A∥B)=(1+T)log⁡(1+T)+g(α,T)D(A\|B)=(1+T)\log(1+T)+g(\alpha,T)D(A∥B)=(1+T)log(1+T)+g(α,T). Here D(A∥B)=Re⁡tr⁡((x↦xlog⁡x)(A)−Alog⁡B)D(A\|B)=\operatorname{Re}\operatorname{tr}((x\mapsto x\log x)(A)-A\log B)D(A∥B)=Retr((x↦xlogx)(A)−AlogB), viewed in the extended reals, and matrix functions use continuous functional calculus with log⁡0=0\log0=0log0=0. The case α=0\alpha=0α=0 is included through the explicitly defined value of ggg.

Human review
  • Endorsed by Shuze Chen · Oct 24, 2026

    Confirmed by the moderator at approval.

  • Endorsed by Goku · Oct 24, 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