Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Smooth min entropy from fixed-reference alpha entropy

Proved
QAlgorithms.Papers.Independent.Support50.P23.smoothMinFromAlpha

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

prove2quantum

Let ρAB\rho_{AB}ρAB​ and σB\sigma_BσB​ be normalized finite-dimensional states. For every ε>0\varepsilon>0ε>0 and 1<α≤21<\alpha\le21<α≤2,

Hmin⁡ε(A∣B)ρ∣σ≥Hα(A∣B)ρ∣σ−log⁡2(2/ε2)α−1.H_{\min}^{\varepsilon}(A|B)_{\rho|\sigma}\ge H_\alpha(A|B)_{\rho|\sigma}-\frac{\log_2(2/\varepsilon^2)}{\alpha-1}.Hminε​(A∣B)ρ∣σ​≥Hα​(A∣B)ρ∣σ​−α−1log2​(2/ε2)​.

Smoothing uses the closed subnormalized root-fidelity ball and holds the reference fixed. No support hypothesis is imposed: support failure can give Hα=−∞H_\alpha=-\inftyHα​=−∞, and ε≥1\varepsilon\ge1ε≥1 gives smooth min entropy +∞+\infty+∞. This is the bridge from Petz entropy to one-shot smooth min entropy.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Support50_P23
Formal statement
noncomputable section

open Filter

open scoped ENNReal ComplexConjugate ComplexOrder

namespace QAlgorithms.Papers.Independent.Support50.P23

attribute [local instance] Classical.propDecidable

-- Exact accepted bodies. Other similarly named catalogue entries use different
-- smoothing, conditioning order, logarithm base, or channel-level objects.

open QAlgorithms.Papers.Independent.Gap20.P03 (fixedMin)

open QAlgorithms.Papers.Independent.Dependency50.P36 (relativeEntropy)

open QAlgorithms.Papers.Independent.Information100.P67 (IsSubnormalized)

open QAlgorithms.Papers.Independent.Information100.P69 (matrixPower)

variable {A B C D I : Type*}

variable [Fintype A] [Fintype B] [Fintype C] [Fintype D] [Fintype I]

variable [DecidableEq A] [DecidableEq B] [DecidableEq C] [DecidableEq D] [DecidableEq I]

/-- Theorem 7, PDF p.4, (22). No support restriction is added. All epsilon>0 are
retained; epsilon>=1 makes the left hand side +infinity. -/
theorem smoothMinFromAlpha [Nonempty A] [Nonempty B]
    (rho : Matrix (A × B) (A × B) ℂ) (sigma : Matrix B B ℂ)
    (hrho : IsDensityMatrix rho) (hsigma : IsDensityMatrix sigma)
    (epsilon alpha : ℝ) (hepsilon : 0 < epsilon) (halpha : 1 < alpha) (halpha2 : alpha ≤ 2) :
    finiteAlpha rho sigma alpha -
      ((Real.logb 2 (2 / epsilon ^ 2) / (alpha - 1) : ℝ) : EReal) ≤
        sineSmoothFixedMin rho sigma epsilon := by sorry

end QAlgorithms.Papers.Independent.Support50.P23

end
Source
A Fully Quantum Asymptotic Equipartition Property, arXiv:0811.1221v3, PDF p. 4, Theorem 7: smooth min entropy from alpha entropy, (22)
Read-back

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

Let A,BA,BA,B be finite, nonempty index types with decidable equality, and let ρ\rhoρ on A×BA\times BA×B and σ\sigmaσ on BBB be positive semidefinite complex matrices of trace one. Let ε,α∈R\varepsilon,\alpha\in\mathbb Rε,α∈R satisfy ε>0\varepsilon>0ε>0 and 1<α≤21<\alpha\le21<α≤2, with no upper bound assumed on ε\varepsilonε. Put Q=IA⊗σQ=I_A\otimes\sigmaQ=IA​⊗σ, define L(0)=−∞L(0)=-\inftyL(0)=−∞, L(x)=log⁡2xL(x)=\log_2xL(x)=log2​x otherwise, and define

Hα(ρ,σ)={−∞,ker⁡Q⊈ker⁡ρ,(1−α)−1L ⁣(Re⁡tr⁡[ραgα(Q)]),ker⁡Q⊆ker⁡ρ,H_\alpha(\rho,\sigma)= \begin{cases} -\infty,&\ker Q\not\subseteq\ker\rho,\\ (1-\alpha)^{-1}L\!\left(\operatorname{Re}\operatorname{tr}[\rho^\alpha g_\alpha(Q)]\right),&\ker Q\subseteq\ker\rho, \end{cases}Hα​(ρ,σ)={−∞,(1−α)−1L(Retr[ραgα​(Q)]),​kerQ⊆kerρ,kerQ⊆kerρ,​

where gα(0)=0g_\alpha(0)=0gα​(0)=0, gα(x)=x1−αg_\alpha(x)=x^{1-\alpha}gα​(x)=x1−α otherwise, and matrix functions use continuous functional calculus. For any τ\tauτ, let h(τ,σ)h(\tau,\sigma)h(τ,σ) be the extended-real supremum of real rrr satisfying 2−rQ−τ⪰02^{-r}Q-\tau\succeq02−rQ−τ⪰0, and let hε(ρ,σ)h^\varepsilon(\rho,\sigma)hε(ρ,σ) be the supremum of these values over τ⪰0\tau\succeq0τ⪰0 with Re⁡tr⁡τ≤1\operatorname{Re}\operatorname{tr}\tau\le1Retrτ≤1 and 1−F(ρ,τ)2≤ε\sqrt{1-F(\rho,\tau)^2}\le\varepsilon1−F(ρ,τ)2​≤ε, where F(X,Y)=Re⁡tr⁡X YXF(X,Y)=\operatorname{Re}\operatorname{tr}\sqrt{\sqrt X\,Y\sqrt X}F(X,Y)=RetrX​YX​​. Then

Hα(ρ,σ)−log⁡2(2/ε2)α−1≤hε(ρ,σ)H_\alpha(\rho,\sigma)-\frac{\log_2(2/\varepsilon^2)}{\alpha-1} \le h^\varepsilon(\rho,\sigma)Hα​(ρ,σ)−α−1log2​(2/ε2)​≤hε(ρ,σ)

in the extended reals. Zero τ\tauτ is permitted; when ε≥1\varepsilon\ge1ε≥1, it belongs to the smoothing set and makes the right-hand side +∞+\infty+∞.

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