Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

§III/§VI: smooth entropy and mutual-information specializations

Proved
QAlgorithms.Papers.Independent.Support50.P21.smooth_specializations

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

prove2quantum

All systems are nonzero finite-dimensional complex Hilbert spaces, and logarithms have base two.

Let ε≥0\varepsilon\ge0ε≥0, let ρ\rhoρ and ρAB\rho_{AB}ρAB​ be density operators, and let σB\sigma_BσB​ be a density operator. The smooth entropy substitutions are

Hmin⁡ε(ρ)=−Dmax⁡ε(ρ∥I),Hmax⁡ε(ρ)=−Dmin⁡ε(ρ∥I),H_{\min}^{\varepsilon}(\rho)=-D_{\max}^{\varepsilon}(\rho\Vert I),\qquad H_{\max}^{\varepsilon}(\rho)=-D_{\min}^{\varepsilon}(\rho\Vert I),Hminε​(ρ)=−Dmaxε​(ρ∥I),Hmaxε​(ρ)=−Dminε​(ρ∥I), Hmin⁡ε(ρAB∣σB)=−Dmax⁡ε(ρAB∥IA⊗σB),Hmax⁡ε(ρAB∣σB)=−Dmin⁡ε(ρAB∥IA⊗σB).H_{\min}^{\varepsilon}(\rho_{AB}\mid\sigma_B)=-D_{\max}^{\varepsilon}(\rho_{AB}\Vert I_A\otimes\sigma_B),\quad H_{\max}^{\varepsilon}(\rho_{AB}\mid\sigma_B)=-D_{\min}^{\varepsilon}(\rho_{AB}\Vert I_A\otimes\sigma_B).Hminε​(ρAB​∣σB​)=−Dmaxε​(ρAB​∥IA​⊗σB​),Hmaxε​(ρAB​∣σB​)=−Dminε​(ρAB​∥IA​⊗σB​).

For the original marginals ρA,ρB\rho_A,\rho_BρA​,ρB​, define Imin⁡/max⁡ε(A:B)=Dmin⁡/max⁡ε(ρAB∥ρA⊗ρB)I_{\min/\max}^{\varepsilon}(A:B)=D_{\min/\max}^{\varepsilon}(\rho_{AB}\Vert\rho_A\otimes\rho_B)Imin/maxε​(A:B)=Dmin/maxε​(ρAB​∥ρA​⊗ρB​). Only the first argument is smoothed; the comparison operator is held fixed. These are definitions with the paper's trace-norm smoothing and rank-based max-entropy convention.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Support50_P21
Formal statement
noncomputable section

namespace QAlgorithms.Papers.Independent.Support50.P21

open QAlgorithms QAlgorithms.Entanglement Filter

open scoped Matrix ComplexConjugate ComplexOrder

attribute [local instance] Classical.propDecidable Classical.decEq

-- Use the accepted declarations directly; no paper-local aliases hide dimension binders.

open QAlgorithms.Papers.Independent.Claimed50.P33
  (FiniteState IsAverageLocalMonotone IsAdaptiveLOCCMonotone)

open QAlgorithms.Papers.Independent.Information100.P63 (Mat)

open QAlgorithms.Papers.Independent.Dependency50.P35 (tr elog maxRelative)

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

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

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

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

section Statements

variable [Nonempty I] [Nonempty J]

/-- PDF pp.4,7: literal smooth substitutions, retaining the original marginals
and the paper's rank-based max convention. These are definitional identities. -/
theorem smooth_specializations (eps : ℝ) (heps : 0 ≤ eps)
    (rho : Mat I) (hrho : IsDensityMatrix rho)
    (rhoAB : Mat (I × J)) (hAB : IsDensityMatrix rhoAB)
    (sigmaB : Mat J) (hB : IsDensityMatrix sigmaB) :
    smoothHMin eps rho = -traceSmoothMax eps rho 1 ∧
    smoothHMax eps rho = -traceSmoothMin eps rho 1 ∧
    smoothConditionalMin eps rhoAB sigmaB =
      -traceSmoothMax eps rhoAB (Matrix.kronecker (1 : Mat I) sigmaB) ∧
    smoothConditionalMax eps rhoAB sigmaB =
      -traceSmoothMin eps rhoAB (Matrix.kronecker (1 : Mat I) sigmaB) ∧
    smoothMutualMin eps rhoAB =
      traceSmoothMin eps rhoAB (Matrix.kronecker (traceRight rhoAB) (traceLeft rhoAB)) ∧
    smoothMutualMax eps rhoAB =
      traceSmoothMax eps rhoAB (Matrix.kronecker (traceRight rhoAB) (traceLeft rhoAB)) := by sorry

end Statements

end QAlgorithms.Papers.Independent.Support50.P21

end
Source
Min- and Max- Relative Entropies and a New Entanglement Monotone, arXiv:0803.2770v3, PDF p. 4, §III/§VI: smooth entropy and mutual-information specializations
Read-back

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

Let I,JI,JI,J be finite index types with decidable equality, let ε≥0\varepsilon\ge0ε≥0, and let ρ,ρAB,σB\rho,\rho_{AB},\sigma_Bρ,ρAB​,σB​ be positive semidefinite complex matrices of trace 111 on III, I×JI\times JI×J, and JJJ, respectively. For matrices r,Yr,Yr,Y on the same index type, define Dmax⁡(r∥Y)=inf⁡{ℓ∈R:2ℓY−r is positive semidefinite}D_{\max}(r\Vert Y)=\inf\{\ell\in\mathbb R:2^\ell Y-r\text{ is positive semidefinite}\}Dmax​(r∥Y)=inf{ℓ∈R:2ℓY−r is positive semidefinite} and Dmin⁡(r∥Y)=−L(Re⁡Tr⁡(PrY))D_{\min}(r\Vert Y)=-L(\operatorname{Re}\operatorname{Tr}(P_rY))Dmin​(r∥Y)=−L(ReTr(Pr​Y)), where PrP_rPr​ is the orthogonal projection onto the range of rrr, and L(x)=log⁡2xL(x)=\log_2xL(x)=log2​x for x>0x>0x>0 and −∞-\infty−∞ otherwise. Define Smax⁡ε(X∥Y)S_{\max}^{\varepsilon}(X\Vert Y)Smaxε​(X∥Y) and Smin⁡ε(X∥Y)S_{\min}^{\varepsilon}(X\Vert Y)Sminε​(X∥Y) as, respectively, the extended-real infimum of Dmax⁡(r∥Y)D_{\max}(r\Vert Y)Dmax​(r∥Y) and supremum of Dmin⁡(r∥Y)D_{\min}(r\Vert Y)Dmin​(r∥Y) over positive semidefinite rrr satisfying ∥r−X∥1≤ε\|r-X\|_1\le\varepsilon∥r−X∥1​≤ε and Re⁡Tr⁡r≤Re⁡Tr⁡X\operatorname{Re}\operatorname{Tr}r\le\operatorname{Re}\operatorname{Tr}XReTrr≤ReTrX, with ∥Z∥1=Re⁡Tr⁡Z†Z\|Z\|_1=\operatorname{Re}\operatorname{Tr}\sqrt{Z^\dagger Z}∥Z∥1​=ReTrZ†Z​. The six asserted identities give the named smooth quantities by these expressions:

Hmin⁡ε(ρ)=−Smax⁡ε(ρ∥1),Hmax⁡ε(ρ)=−Smin⁡ε(ρ∥1),Hmin⁡ε(ρAB∣σB)=−Smax⁡ε(ρAB∥1I⊗σB),Hmax⁡ε(ρAB∣σB)=−Smin⁡ε(ρAB∥1I⊗σB),Imin⁡ε(ρAB)=Smin⁡ε(ρAB∥ρA⊗ρB),Imax⁡ε(ρAB)=Smax⁡ε(ρAB∥ρA⊗ρB),\begin{aligned} H_{\min}^{\varepsilon}(\rho)&=-S_{\max}^{\varepsilon}(\rho\Vert1),& H_{\max}^{\varepsilon}(\rho)&=-S_{\min}^{\varepsilon}(\rho\Vert1),\\ H_{\min}^{\varepsilon}(\rho_{AB}\mid\sigma_B) &=-S_{\max}^{\varepsilon}(\rho_{AB}\Vert1_I\otimes\sigma_B),& H_{\max}^{\varepsilon}(\rho_{AB}\mid\sigma_B) &=-S_{\min}^{\varepsilon}(\rho_{AB}\Vert1_I\otimes\sigma_B),\\ I_{\min}^{\varepsilon}(\rho_{AB}) &=S_{\min}^{\varepsilon}(\rho_{AB}\Vert\rho_A\otimes\rho_B),& I_{\max}^{\varepsilon}(\rho_{AB}) &=S_{\max}^{\varepsilon}(\rho_{AB}\Vert\rho_A\otimes\rho_B), \end{aligned}Hminε​(ρ)Hminε​(ρAB​∣σB​)Iminε​(ρAB​)​=−Smaxε​(ρ∥1),=−Smaxε​(ρAB​∥1I​⊗σB​),=Sminε​(ρAB​∥ρA​⊗ρB​),​Hmaxε​(ρ)Hmaxε​(ρAB​∣σB​)Imaxε​(ρAB​)​=−Sminε​(ρ∥1),=−Sminε​(ρAB​∥1I​⊗σB​),=Smaxε​(ρAB​∥ρA​⊗ρB​),​

where (ρA)ij=∑b(ρAB)(i,b),(j,b)(\rho_A)_{ij}=\sum_b(\rho_{AB})_{(i,b),(j,b)}(ρA​)ij​=∑b​(ρAB​)(i,b),(j,b)​ and (ρB)bc=∑i(ρAB)(i,b),(i,c)(\rho_B)_{bc}=\sum_i(\rho_{AB})_{(i,b),(i,c)}(ρB​)bc​=∑i​(ρAB​)(i,b),(i,c)​. The six symbols on the left are defined by the corresponding expressions on the right; there is no optimization over σB\sigma_BσB​. The allowed ε\varepsilonε include 000 and values at least 111, and the smoothing matrices may have reduced trace or be zero.

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