Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

5.12

Open
QAlgorithms.Papers.Independent.Intakenext100.P07.conditional_down_chain

by Gohan · Oct 31, 2026 · Mathlib 0df444a (Lean v4.33.1)

prove2quantum

All quantum systems below are finite-dimensional complex Hilbert spaces of positive dimension, and logarithms are binary. A normalized state is positive semidefinite with trace one; a subnormalized state has trace at most one. Marginals are partial traces. Tildes denote sandwiched Rényi quantities and bars denote Petz quantities. Their trace functionals are Q~α=Tr⁡(σ(1−α)/(2α)ρσ(1−α)/(2α))α\widetilde Q_\alpha=\operatorname{Tr}(\sigma^{(1-\alpha)/(2\alpha)}\rho\sigma^{(1-\alpha)/(2\alpha)})^\alphaQ​α​=Tr(σ(1−α)/(2α)ρσ(1−α)/(2α))α and Q‾α=Tr⁡(ρασ1−α)\overline Q_\alpha=\operatorname{Tr}(\rho^\alpha\sigma^{1-\alpha})Q​α​=Tr(ρασ1−α); the divergences include division by Tr⁡ρ\operatorname{Tr}\rhoTrρ inside the logarithm. Matrix powers vanish on the kernel, including power zero. The notation ρ≪σ\rho\ll\sigmaρ≪σ means support inclusion. The upward conditional entropy optimizes the negative divergence over normalized reference states on the conditioning system; the downward entropy uses the actual marginal. The notation H(A∣B)=H(AB)−H(B)H(A|B)=H(AB)-H(B)H(A∣B)=H(AB)−H(B) denotes conditional von Neumann entropy.

For either fixed-reference entropy in its DPI range on normalized ABC: Hα↓(AB|C)≥Hα↓(A|BC)+Htilde_∞↓(B|C).

Hα↓(AB∣C)ρ≥Hα↓(A∣BC)ρ+H~∞↓(B∣C)ρ.H_\alpha^\downarrow(AB|C)_\rho\ge H_\alpha^\downarrow(A|BC)_\rho+\widetilde H_\infty^\downarrow(B|C)_\rho.Hα↓​(AB∣C)ρ​≥Hα↓​(A∣BC)ρ​+H∞↓​(B∣C)ρ​.

This result records the source’s operational or structural relation, with the stated endpoint and attainment clauses retained.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Intakenext100_P07
Formal statement
noncomputable section

open Filter

open scoped BigOperators Matrix ComplexConjugate ComplexOrder Matrix.Norms.L2Operator Topology

namespace QAlgorithms.Papers.Independent.Intakenext100.P07

attribute [local instance] Classical.propDecidable Classical.decEq

variable {I J A B C D X Y Z : Type*}

variable [Fintype I] [DecidableEq I] [Fintype J] [DecidableEq J]

variable [Fintype A] [DecidableEq A] [Fintype B] [DecidableEq B]

variable [Fintype C] [DecidableEq C] [Fintype D] [DecidableEq D]

variable [Fintype X] [DecidableEq X] [Fintype Y] [DecidableEq Y]

variable [Fintype Z] [DecidableEq Z]

/-- NODE paper:1504.00233#5.12
5.12, arXiv:1504.00233v5, PDF p.87. -/
theorem conditional_down_chain [Nonempty A] [Nonempty B] [Nonempty C] (r : paperMat ((A × B) × C))
    (hr : IsDensityMatrix r) (sand : Bool) (a : ENNReal) (ha : DPIOrder sand a) :
    conditionalDown sand a (paperRegroup r)+conditionalDown true ⊤ (paperMarginalBC r) ≤ conditionalDown sand a r := by sorry

end QAlgorithms.Papers.Independent.Intakenext100.P07

end
Source
Quantum Information Processing with Finite Resources -- Mathematical Foundations, arXiv:1504.00233v5, PDF p. 87, 5.12
Read-back

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

Let A,B,CA,B,CA,B,C be finite nonempty index types with decidable equality, and let r⪰0r\succeq0r⪰0 on (A×B)×C(A\times B)\times C(A×B)×C have trace one. Choose a Boolean flag ϵ\epsilonϵ, written 000 for false and 111 for true, and a∈[0,∞]a\in[0,\infty]a∈[0,∞], with a≥12a\ge\tfrac12a≥21​ if ϵ=1\epsilon=1ϵ=1, and a≤2a\le2a≤2 if ϵ=0\epsilon=0ϵ=0. Define the regrouped matrix and its BCBCBC marginal by

R(u,(b,c)),(u′,(b′,c′))=r((u,b),c),((u′,b′),c′),S(b,c),(b′,c′)=∑ur((u,b),c),((u,b′),c′).R_{(u,(b,c)),(u',(b',c'))} =r_{((u,b),c),((u',b'),c')},\qquad S_{(b,c),(b',c')}=\sum_u r_{((u,b),c),((u,b'),c')}.R(u,(b,c)),(u′,(b′,c′))​=r((u,b),c),((u′,b′),c′)​,S(b,c),(b′,c′)​=u∑​r((u,b),c),((u,b′),c′)​.

For a matrix XXX on E×FE\times FE×F, write (XF)v,w=∑uX(u,v),(u,w)(X_F)_{v,w}=\sum_uX_{(u,v),(u,w)}(XF​)v,w​=∑u​X(u,v),(u,w)​ and Hϵ↓(X;t)=−Dtϵ(X,IE⊗XF)H_\epsilon^\downarrow(X;t)=-D_t^\epsilon(X,I_E\otimes X_F)Hϵ↓​(X;t)=−Dtϵ​(X,IE​⊗XF​). Here DDD is defined explicitly as follows: let X[p]X^{[p]}X[p] be functional calculus with the function zero at zero and xpx^pxp elsewhere; let log⁡2X=cfc⁡(log⁡2,X)\log_2X=\operatorname{cfc}(\log_2,X)log2​X=cfc(log2​,X), L(0)=−∞L(0)=-\inftyL(0)=−∞, and L(x)=log⁡2xL(x)=\log_2xL(x)=log2​x otherwise; and set

Qt0(X,Y)=Re⁡tr⁡(X[t]Y[1−t]),Qt1(X,Y)=Re⁡tr⁡[(Y[(1−t)/(2t)]XY[(1−t)/(2t)])[t]].Q_t^0(X,Y)=\operatorname{Re}\operatorname{tr}(X^{[t]}Y^{[1-t]}),\qquad Q_t^1(X,Y)=\operatorname{Re}\operatorname{tr} \left[\left(Y^{[(1-t)/(2t)]}XY^{[(1-t)/(2t)]}\right)^{[t]}\right].Qt0​(X,Y)=Retr(X[t]Y[1−t]),Qt1​(X,Y)=Retr[(Y[(1−t)/(2t)]XY[(1−t)/(2t)])[t]].

For real t>0t>0t>0, D1ϵ(X,Y)D_1^\epsilon(X,Y)D1ϵ​(X,Y) equals Re⁡tr⁡[X(log⁡2X−log⁡2Y)]/Re⁡tr⁡X\operatorname{Re}\operatorname{tr}[X(\log_2X-\log_2Y)]/\operatorname{Re}\operatorname{tr}XRetr[X(log2​X−log2​Y)]/RetrX if ker⁡Y⊆ker⁡X\ker Y\subseteq\ker XkerY⊆kerX, and +∞+\infty+∞ otherwise; when t≠1t\ne1t=1, Dtϵ(X,Y)D_t^\epsilon(X,Y)Dtϵ​(X,Y) equals

1t−1L ⁣(Qtϵ(X,Y)Re⁡tr⁡X)\frac1{t-1}L\!\left(\frac{Q_t^\epsilon(X,Y)} {\operatorname{Re}\operatorname{tr}X}\right)t−11​L(RetrXQtϵ​(X,Y)​)

if (t<1 and Qtϵ(X,Y)>0)(t<1\text{ and }Q_t^\epsilon(X,Y)>0)(t<1 and Qtϵ​(X,Y)>0) or ker⁡Y⊆ker⁡X\ker Y\subseteq\ker XkerY⊆kerX, and equals +∞+\infty+∞ otherwise. Set D0ϵ=inf⁡t>0DtϵD_0^\epsilon=\inf_{t>0}D_t^\epsilonD0ϵ​=inft>0​Dtϵ​, D∞0=sup⁡t>0Dt0D_\infty^0=\sup_{t>0}D_t^0D∞0​=supt>0​Dt0​, and D∞1(X,Y)=inf⁡{λ∈R:X⪯2λY}D_\infty^1(X,Y)=\inf\{\lambda\in\mathbb R:X\preceq2^\lambda Y\}D∞1​(X,Y)=inf{λ∈R:X⪯2λY}, with extended-real infima and suprema and X⪯YX\preceq YX⪯Y meaning Y−X⪰0Y-X\succeq0Y−X⪰0. Then the following extended-real inequality holds:

Hϵ↓(R;a)+H1↓(S;∞)≤Hϵ↓(r;a).H_\epsilon^\downarrow(R;a)+H_1^\downarrow(S;\infty) \le H_\epsilon^\downarrow(r;a).Hϵ↓​(R;a)+H1↓​(S;∞)≤Hϵ↓​(r;a).

In particular, the leftmost conditional treats AAA as its first subsystem and BCBCBC as its second, while the rightmost treats ABABAB as its first subsystem and CCC as its second; the endpoint orders use precisely the definitions just given.

Human review
  • Endorsed by Shuze Chen · Oct 31, 2026

    Confirmed by the moderator at approval.

  • Endorsed by Gohan · Oct 31, 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