Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Convex Petz moment, entropy slope and compact maximizers

Proved
QAlgorithms.Papers.Independent.Gap48.P41.petz_moment_properties

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

prove2quantum

Let III be a nonempty finite orthonormal basis, and let ρ,σ\rho,\sigmaρ,σ be positive semidefinite complex matrices of trace one with supp⁡ρ⊆supp⁡σ\operatorname{supp}\rho\subseteq\operatorname{supp}\sigmasuppρ⊆suppσ. For a positive integer nnn, an arbitrary effect 0≤An≤I0\le A_n\le I0≤An​≤I accepts the null state ρ⊗n\rho^{\otimes n}ρ⊗n. Write αn(An)=Tr⁡[ρ⊗n(I−An)]\alpha_n(A_n)=\operatorname{Tr}[\rho^{\otimes n}(I-A_n)]αn​(An​)=Tr[ρ⊗n(I−An​)] and βn(An)=Tr⁡[σ⊗nAn]\beta_n(A_n)=\operatorname{Tr}[\sigma^{\otimes n}A_n]βn​(An​)=Tr[σ⊗nAn​]. All logarithms are natural, with log⁡0=−∞\log0=-\inftylog0=−∞ in rates. Write D(ρ∥σ)=Tr⁡[ρ(log⁡ρ−log⁡σ)]D(\rho\Vert\sigma)=\operatorname{Tr}[\rho(\log\rho-\log\sigma)]D(ρ∥σ)=Tr[ρ(logρ−logσ)], using support-restricted spectral logarithms inside the trace.

Put ψ(s)=log⁡Tr⁡(ρ1+sσ−s)\psi(s)=\log\operatorname{Tr}(\rho^{1+s}\sigma^{-s})ψ(s)=logTr(ρ1+sσ−s) and ϕ(λ)=max⁡0≤s≤1{λs−ψ(s)}\phi(\lambda)=\max_{0\le s\le1}\{\lambda s-\psi(s)\}ϕ(λ)=max0≤s≤1​{λs−ψ(s)}. Spectral powers vanish on the kernel, including at zero and negative orders.

The function ψ\psiψ is convex on [0,1][0,1][0,1], ψ(0)=0\psi(0)=0ψ(0)=0, and ψ′(0)=D(ρ∥σ)\psi'(0)=D(\rho\Vert\sigma)ψ′(0)=D(ρ∥σ). Moreover,

λ>D(ρ∥σ) ⟹ ϕ(λ)>0.\lambda>D(\rho\Vert\sigma)\ \Longrightarrow\ \phi(\lambda)>0.λ>D(ρ∥σ) ⟹ ϕ(λ)>0.

For D(ρ∥σ)≤λ≤ψ′(1)D(\rho\Vert\sigma)\le\lambda\le\psi'(1)D(ρ∥σ)≤λ≤ψ′(1) and s∈[0,1]s\in[0,1]s∈[0,1],

s∈argmax⁡0≤t≤1{λt−ψ(t)}⟺ψ′(s)=λ.s\in\operatorname*{argmax}_{0\le t\le1}\{\lambda t-\psi(t)\} \quad\Longleftrightarrow\quad\psi'(s)=\lambda.s∈0≤t≤1argmax​{λt−ψ(t)}⟺ψ′(s)=λ.

These properties identify the entropy threshold and the compact optimizer condition, without claiming uniqueness.

Formalization Note Endpoint derivatives use the finite spectral expression on a neighborhood of the interval. The displayed strict second-derivative claim on page 3 is corrected to nonnegative curvature, as required by the equal-state case.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Gap48_P41
Formal statement
noncomputable section

namespace QAlgorithms.Papers.Independent.Gap48.P41

open scoped Matrix ComplexOrder BigOperators

open Filter QAlgorithms

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

open QAlgorithms.Papers.Independent.Information100.P71 (supportPower)

open QAlgorithms.Papers.Independent.Information100.P07 (positiveProjection)

open QAlgorithms.Papers.Independent.Dependency50.P28
  (tr SupportLE relativeEntropy scalarPower classicalD typeI accept TestSequence
    scaled logNonnegativeExtended)

open QAlgorithms.Papers.Independent.Claimed50.P15 (legendre)

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

attribute [local instance] Classical.propDecidable Classical.decEq

variable {I : Type} [Fintype I] [DecidableEq I]

/-- PDF pp.3--4, unnumbered properties and (9): convexity, slope, positivity,
and membership in the set of maximizers. SOURCE CORRECTION: the displayed
psi''>0 on p.3 must be psi''>=0; identical states give psi=0. The prose convexity
claim is retained without an invented nondegeneracy or unique-maximizer premise. -/
theorem petz_moment_properties [Nonempty I]
    (rho sigma : Mat I) (hrho : IsDensityMatrix rho) (hsigma : IsDensityMatrix sigma)
    (hsupp : SupportLE rho sigma) :
    ConvexOn ℝ (Set.Icc 0 1) (petzPsi rho sigma) ∧
    petzPsi rho sigma 0 = 0 ∧
    HasDerivAt (petzPsi rho sigma) (relativeEntropy rho sigma).toReal 0 ∧
    (∀ lambda : ℝ, (relativeEntropy rho sigma).toReal < lambda →
      0 < petzPhi rho sigma lambda) ∧
    (∀ lambda : ℝ, (relativeEntropy rho sigma).toReal ≤ lambda →
      lambda ≤ deriv (petzPsi rho sigma) 1 → ∀ s ∈ Set.Icc (0 : ℝ) 1,
      (IsGreatest ((fun t : ℝ => lambda * t - petzPsi rho sigma t) '' Set.Icc 0 1)
        (lambda * s - petzPsi rho sigma s) ↔ deriv (petzPsi rho sigma) s = lambda)) := by sorry

end QAlgorithms.Papers.Independent.Gap48.P41

end
Source
Strong Converse and Stein's Lemma in the Quantum Hypothesis Testing, arXiv:quant-ph/9906090v1, PDF p. 3, §2, unnumbered properties and (9): convex moment, slope, positive transform
Read-back

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

Let III be a nonempty finite type with decidable equality, and let ρ,σ\rho,\sigmaρ,σ be positive semidefinite complex matrices indexed by III, each of trace 111, with ker⁡σ⊆ker⁡ρ\ker\sigma\subseteq\ker\rhokerσ⊆kerρ. Define X[t]X^{[t]}X[t] by spectral functional calculus using the function that is 000 at zero and xtx^txt otherwise, and set

ψ(s)=log⁡Re⁡Tr⁡ ⁣(ρ[1+s]σ[−s]),ϕ(λ)=sup⁡0≤t≤1{λt−ψ(t)},D=Re⁡Tr⁡ ⁣(ρ(log⁡ρ−log⁡σ))Re⁡Tr⁡ρ.\psi(s)=\log\operatorname{Re}\operatorname{Tr}\!\left(\rho^{[1+s]}\sigma^{[-s]}\right), \qquad \phi(\lambda)=\sup_{0\le t\le1}\{\lambda t-\psi(t)\}, \qquad D=\frac{\operatorname{Re}\operatorname{Tr}\!\left(\rho(\log\rho-\log\sigma)\right)} {\operatorname{Re}\operatorname{Tr}\rho}.ψ(s)=logReTr(ρ[1+s]σ[−s]),ϕ(λ)=0≤t≤1sup​{λt−ψ(t)},D=ReTrρReTr(ρ(logρ−logσ))​.

The matrix logarithms use spectral functional calculus with the total real logarithm, for which log⁡0=0\log0=0log0=0; the displayed denominator equals 111. Then ψ\psiψ is convex on [0,1][0,1][0,1], ψ(0)=0\psi(0)=0ψ(0)=0, and its ordinary derivative at 000 exists and equals DDD. Every real λ>D\lambda>Dλ>D satisfies ϕ(λ)>0\phi(\lambda)>0ϕ(λ)>0. Moreover, for every real λ\lambdaλ satisfying D≤λ≤ψ′(1)D\le\lambda\le\psi'(1)D≤λ≤ψ′(1) and every s∈[0,1]s\in[0,1]s∈[0,1], including either endpoint, the value λs−ψ(s)\lambda s-\psi(s)λs−ψ(s) is the greatest member of {λt−ψ(t):t∈[0,1]}\{\lambda t-\psi(t):t\in[0,1]\}{λt−ψ(t):t∈[0,1]} if and only if ψ′(s)=λ\psi'(s)=\lambdaψ′(s)=λ. No assumption that ρ≠σ\rho\ne\sigmaρ=σ is imposed.

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