Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem A.14: Compatible observables Xobs,Yobs valued in compatible Xalg,Yalg satisfy I(Xobs:Yobs)≤I(Xalg:Yobs)≤I(Xalg:Yalg).

Open
QAlgorithms.Papers.Independent.Renew46.P43.observableInformation

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

prove2quantum

Fix one normalized ambient state. Let countable observables Xobs,YobsX_{\rm obs},Y_{\rm obs}Xobs​,Yobs​ take values in the compatible finite intrinsic subalgebras X,Y\mathcal X,\mathcal YX,Y. Classical information is defined by relative entropy of the actual joint outcome law, and hybrid information by mean posterior-state divergence; thus no difference of two infinite entropies is used.

I(Xobs;Yobs)≤I(X;Yobs)≤I(X;Y).I(X_{\rm obs};Y_{\rm obs})\le I(\mathcal X;Y_{\rm obs})\le I(\mathcal X;\mathcal Y).I(Xobs​;Yobs​)≤I(X;Yobs​)≤I(X;Y).

This is the source result with its stated operational or information-theoretic domain.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Renew46_P43
Formal statement
noncomputable section

namespace QAlgorithms.Papers.Independent.Renew46.P43

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

open Filter MeasureTheory

open QAlgorithms.Papers.Independent.Information100.P08 (IsPure)

open QAlgorithms.Papers.Independent.Information100.P06 (Below)

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

open QAlgorithms.Papers.Independent.Dependency50.P19 (count both)

open QAlgorithms.Papers.Independent.Information100.P22 (coherentInformation)

attribute [local instance] Classical.propDecidable Classical.decEq

open QAlgorithms.Papers.Independent.Claimed50.P40 (InBlocks)

attribute [local instance] QAlgorithms.Papers.Independent.Renew46.P43.matrixMeasurableSpace

/-- Theorem A.14, PDF p.72: both inequalities, using countable outcome laws and
divergence-defined hybrid information rather than infinity-minus-infinity. -/
theorem observableInformation {X Y A : IntrinsicSystem} {B D : Type*} [Countable B] [Countable D]
    (E : Embedding X A) (F : Embedding Y A) (hcompat : CommutingImages E.map F.map)
    (U : B → Matrix (Fin X.dim) (Fin X.dim) ℂ) (V : D → Matrix (Fin Y.dim) (Fin Y.dim) ℂ)
    (hU : IsCountablePOVM U) (hV : IsCountablePOVM V)
    (hUX : ∀ b, InBlocks X.block (U b)) (hVY : ∀ b, InBlocks Y.block (V b))
    (rho : Matrix (Fin A.dim) (Fin A.dim) ℂ) (hrho : A.IsState rho) :
    measuredInformation (fun b => E.map (U b)) (fun b => F.map (V b)) rho ≤
      hybridInformation X E (fun b => F.map (V b)) rho ∧
    hybridInformation X E (fun b => F.map (V b)) rho ≤
      ENNReal.ofReal (intrinsicMutualInformation X Y E.map F.map rho) := by sorry

end QAlgorithms.Papers.Independent.Renew46.P43

end
Source
Coding Theorems of Quantum Information Theory, arXiv:quant-ph/9907077v1, PDF p. 72, Theorem A.14
Read-back

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

Let X,Y,AX,Y,AX,Y,A be systems with positive integer dimensions dZd_ZdZ​ and coordinate block labels βZ\beta_ZβZ​; matrices respecting a system's blocks have zero entries between differently labelled indices. Let E:MdX(C)→MdA(C)E:M_{d_X}(\mathbb C)\to M_{d_A}(\mathbb C)E:MdX​​(C)→MdA​​(C) and F:MdY(C)→MdA(C)F:M_{d_Y}(\mathbb C)\to M_{d_A}(\mathbb C)F:MdY​​(C)→MdA​​(C) be unital complex-linear embeddings which preserve the block algebras, preserve products and adjoints and are injective on those algebras, and preserve positivity under every finite amplification restricted to matrices supported within equal source-block labels. Assume their images of block-respecting matrices commute. Let B,DB,DB,D be arbitrary countable types, and let (Ub)b∈B(U_b)_{b\in B}(Ub​)b∈B​ and (Vd)d∈D(V_d)_{d\in D}(Vd​)d∈D​ be families of positive semidefinite matrices on XXX and YYY, respectively, each respecting its system's blocks, such that the families have sums IXI_XIX​ and IYI_YIY​, respectively. These are countable sums in the finite-dimensional matrix topology; finiteness of the outcome types is not required. Let ρ\rhoρ be a positive semidefinite, trace-111, block-respecting matrix on AAA. Put U^b=E(Ub)\widehat U_b=E(U_b)Ub​=E(Ub​), V^d=F(Vd)\widehat V_d=F(V_d)Vd​=F(Vd​), and define the restriction map

[RE(Z)]ij={tr⁡(ZE(eji)),βX(i)=βX(j),0,otherwise,[R_E(Z)]_{ij}= \begin{cases} \operatorname{tr}\bigl(ZE(e_{ji})\bigr),&\beta_X(i)=\beta_X(j),\\ 0,&\text{otherwise}, \end{cases}[RE​(Z)]ij​={tr(ZE(eji​)),0,​βX​(i)=βX​(j),otherwise,​

with RFR_FRF​ defined analogously, and define

J(i,u),(j,v)={tr⁡(ρE(eji)F(evu)),βX(i)=βX(j) and βY(u)=βY(v),0,otherwise.J_{(i,u),(j,v)}= \begin{cases} \operatorname{tr}\bigl(\rho E(e_{ji})F(e_{vu})\bigr), &\beta_X(i)=\beta_X(j)\ \text{and}\ \beta_Y(u)=\beta_Y(v),\\ 0,&\text{otherwise}. \end{cases}J(i,u),(j,v)​={tr(ρE(eji​)F(evu​)),0,​βX​(i)=βX​(j) and βY​(u)=βY​(v),otherwise.​

For real numbers x,yx,yx,y, define the extended nonnegative quantity

K(x,y)={max⁡{y,0},x=0,+∞,x≠0, y=0,max⁡ ⁣{xlog⁡2(x/y)+(y−x)/ln⁡2,0},x≠0, y≠0.K(x,y)= \begin{cases} \max\{y,0\},&x=0,\\ +\infty,&x\ne0,\ y=0,\\ \max\!\left\{x\log_2(x/y)+(y-x)/\ln2,0\right\},&x\ne0,\ y\ne0. \end{cases}K(x,y)=⎩⎨⎧​max{y,0},+∞,max{xlog2​(x/y)+(y−x)/ln2,0},​x=0,x=0, y=0,x=0, y=0.​

Also define D2(S∥T)=0D_2(S\|T)=0D2​(S∥T)=0 if either matrix fails to be positive semidefinite of trace 111; if both are such matrices, define it to be +∞+\infty+∞ when ker⁡T⊈ker⁡S\ker T\not\subseteq\ker SkerT⊆kerS, and otherwise define

D2(S∥T)=max⁡ ⁣{Re⁡tr⁡(cfc⁡[t↦tln⁡t](S)−S cfc⁡[ln⁡](T)),0}ln⁡2.D_2(S\|T)= \frac{\max\!\left\{\operatorname{Re}\operatorname{tr} \bigl(\operatorname{cfc}[t\mapsto t\ln t](S)-S\,\operatorname{cfc}[\ln](T)\bigr),0\right\}}{\ln2}.D2​(S∥T)=ln2max{Retr(cfc[t↦tlnt](S)−Scfc[ln](T)),0}​.

Write H(M)=−Re⁡tr⁡(cfc⁡[t↦tlog⁡2t](M))H(M)=-\operatorname{Re}\operatorname{tr}\bigl(\operatorname{cfc}[t\mapsto t\log_2t](M)\bigr)H(M)=−Retr(cfc[t↦tlog2​t](M)), and set qd=Re⁡tr⁡(ρV^d)q_d=\operatorname{Re}\operatorname{tr}(\rho\widehat V_d)qd​=Retr(ρVd​) and Sd=RE(V^d ρ V^d)S_d=R_E\bigl(\sqrt{\widehat V_d}\,\rho\,\sqrt{\widehat V_d}\bigr)Sd​=RE​(Vd​​ρVd​​), where the square root is given by continuous functional calculus. Then the following inequalities hold in [0,∞][0,\infty][0,∞]:

∑(b,d)∈B×DK ⁣(Re⁡tr⁡(ρU^bV^d),Re⁡tr⁡(ρU^b) Re⁡tr⁡(ρV^d))≤∑d∈Dmax⁡{qd,0} D2 ⁣(qd−1Sd ∥ RE(ρ))≤max⁡{H(RE(ρ))+H(RF(ρ))−H(J),0}.\begin{aligned} &\sum_{(b,d)\in B\times D} K\!\left( \operatorname{Re}\operatorname{tr}(\rho\widehat U_b\widehat V_d), \operatorname{Re}\operatorname{tr}(\rho\widehat U_b)\, \operatorname{Re}\operatorname{tr}(\rho\widehat V_d) \right)\\ &\qquad\le \sum_{d\in D}\max\{q_d,0\}\, D_2\!\left(q_d^{-1}S_d\,\middle\|\,R_E(\rho)\right) \le \max\{H(R_E(\rho))+H(R_F(\rho))-H(J),0\}. \end{aligned}​(b,d)∈B×D∑​K(Retr(ρUb​Vd​),Retr(ρUb​)Retr(ρVd​))≤d∈D∑​max{qd​,0}D2​(qd−1​Sd​​RE​(ρ))≤max{H(RE​(ρ))+H(RF​(ρ))−H(J),0}.​

The sums are the extended nonnegative sums of the displayed terms. The inverse is total, so qd−1=0q_d^{-1}=0qd−1​=0 when qd=0q_d=0qd​=0, and multiplication in [0,∞][0,\infty][0,∞] uses 0⋅+∞=00\cdot+\infty=00⋅+∞=0.

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