Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

§4.4.2 (4.91)–(4.93): Nussbaum–Szkoła representation

Open
QAlgorithms.Papers.Independent.Intakenext100.P07.nussbaum_szkola

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.

Given actual spectral data for normalized ρ,σ, P(x,y)=λx|<ex|fy>|²,Q(x,y)=μy|<ex|fy>|² are probabilities, Dbarα(ρ||σ)=Dα(P||Q), support inclusion is equivalent, and the constructed laws tensorize.

P(x,y)=λx∣⟨ex,fy⟩∣2,Q(x,y)=μy∣⟨ex,fy⟩∣2,D‾α(ρ∥σ)=Dα(P∥Q).P(x,y)=\lambda_x|\langle e_x,f_y\rangle|^2,\quad Q(x,y)=\mu_y|\langle e_x,f_y\rangle|^2,\quad\overline D_\alpha(\rho\|\sigma)=D_\alpha(P\|Q).P(x,y)=λx​∣⟨ex​,fy​⟩∣2,Q(x,y)=μy​∣⟨ex​,fy​⟩∣2,Dα​(ρ∥σ)=Dα​(P∥Q).

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#4.ns
§4.4.2 (4.91)–(4.93): Nussbaum–Szkoła representation, arXiv:1504.00233v5, PDF p.70. -/
theorem nussbaum_szkola [Nonempty I] [Nonempty J] (r s : paperMat I) (hr : IsDensityMatrix r) (hs : IsDensityMatrix s)
    (v w : I → ℝ) (U V : paperMat I) (hv : SpectralRep r v U) (hw : SpectralRep s w V) :
    paperProbability (paperNsP v U V) ∧ paperProbability (paperNsQ w U V) ∧
    (∀ a : ℝ, 0 ≤ a → paperRenyi false r s (ENNReal.ofReal a) = paperRenyi false
      (Matrix.diagonal (fun xy => (paperNsP v U V xy:ℂ))) (Matrix.diagonal (fun xy => (paperNsQ w U V xy:ℂ))) (ENNReal.ofReal a)) ∧
    (paperSupported r s ↔ ∀ xy, 0 < paperNsP v U V xy → 0 < paperNsQ w U V xy) ∧
    (∀ r' s' : paperMat J, IsDensityMatrix r' → IsDensityMatrix s' → ∀ v' w' : J → ℝ, ∀ U' V' : paperMat J,
      SpectralRep r' v' U' → SpectralRep s' w' V' → ∀ i j : I, ∀ k l : J,
      paperNsP (fun ik : I × J => v ik.1*v' ik.2) (Matrix.kronecker U U') (Matrix.kronecker V V') ((i,k),(j,l)) = 
        paperNsP v U V (i,j)*paperNsP v' U' V' (k,l)) := by sorry

end QAlgorithms.Papers.Independent.Intakenext100.P07

end
Source
Quantum Information Processing with Finite Resources -- Mathematical Foundations, arXiv:1504.00233v5, PDF p. 70, §4.4.2 (4.91)–(4.93): Nussbaum–Szkoła representation
Read-back

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

Let I,JI,JI,J be nonempty finite types with decidable equality; let r,sr,sr,s be positive semidefinite matrices of complex trace 111 on III; and let v,w:I→Rv,w:I\to\mathbb Rv,w:I→R and matrices U,VU,VU,V satisfy

vi,wi≥0,U†U=UU†=V†V=VV†=1,r=Udiag⁡(v)V0,s=Vdiag⁡(w)V†,v_i,w_i\ge0,\quad U^\dagger U=UU^\dagger=V^\dagger V=VV^\dagger=1,\quad r=U\operatorname{diag}(v)V_0,\quad s=V\operatorname{diag}(w)V^\dagger,vi​,wi​≥0,U†U=UU†=V†V=VV†=1,r=Udiag(v)V0​,s=Vdiag(w)V†,

where V0=U†V_0=U^\daggerV0​=U†, so the first decomposition is r=Udiag⁡(v)U†r=U\operatorname{diag}(v)U^\daggerr=Udiag(v)U†. Define

pij=vi∣(U†V)ij∣2,qij=wj∣(U†V)ij∣2.p_{ij}=v_i|(U^\dagger V)_{ij}|^2,\qquad q_{ij}=w_j|(U^\dagger V)_{ij}|^2.pij​=vi​∣(U†V)ij​∣2,qij​=wj​∣(U†V)ij​∣2.

Then p,qp,qp,q are nonnegative functions on I×II\times II×I, each summing to 111, and for every real a≥0a\ge0a≥0,

Ra(r∥s)=Ra(diag⁡p∥diag⁡q).\mathcal R_a(r\|s)=\mathcal R_a(\operatorname{diag}p\|\operatorname{diag}q).Ra​(r∥s)=Ra​(diagp∥diagq).

Here the quantity Ra\mathcal R_aRa​ is defined as follows, including its endpoints: let X[t]X^{[t]}X[t] denote functional calculus with the function 000 at eigenvalue 000 and xtx^txt at nonzero eigenvalues, let Qt(X,Y)=Re⁡tr⁡(X[t]Y[1−t])Q_t(X,Y)=\operatorname{Re}\operatorname{tr}(X^{[t]}Y^{[1-t]})Qt​(X,Y)=Retr(X[t]Y[1−t]), let L(0)=−∞\mathcal L(0)=-\inftyL(0)=−∞ and L(z)=log⁡2z\mathcal L(z)=\log_2 zL(z)=log2​z for nonzero real zzz, and let D(X∥Y)D(X\|Y)D(X∥Y) be +∞+\infty+∞ unless ker⁡Y⊆ker⁡X\ker Y\subseteq\ker XkerY⊆kerX, in which case it is the extended-real embedding of

Re⁡tr⁡(X(L(X)−L(Y)))Re⁡tr⁡X,L(X)=(x↦log⁡2x)(X),\frac{\operatorname{Re}\operatorname{tr}(X(L(X)-L(Y)))} {\operatorname{Re}\operatorname{tr}X}, \qquad L(X)=(x\mapsto\log_2x)(X),RetrXRetr(X(L(X)−L(Y)))​,L(X)=(x↦log2​x)(X),

using the total real logarithm with value 000 at 000. For t>0t>0t>0, set R1=D\mathcal R_1=DR1​=D; for t≠1t\ne1t=1, set

Rt(X∥Y)={1t−1L ⁣(Qt(X,Y)Re⁡tr⁡X)(t<1 and Qt(X,Y)>0) or ker⁡Y⊆ker⁡X,+∞otherwise,\mathcal R_t(X\|Y)= \begin{cases} \displaystyle\frac1{t-1}\mathcal L\!\left(\frac{Q_t(X,Y)} {\operatorname{Re}\operatorname{tr}X}\right) &\bigl(t<1\ \text{and }Q_t(X,Y)>0\bigr)\ \text{or }\ker Y\subseteq\ker X,\\ +\infty&\text{otherwise}, \end{cases}Rt​(X∥Y)=⎩⎨⎧​t−11​L(RetrXQt​(X,Y)​)+∞​(t<1 and Qt​(X,Y)>0) or kerY⊆kerX,otherwise,​

and set R0(X∥Y)=inf⁡t>0Rt(X∥Y)\mathcal R_0(X\|Y)=\inf_{t>0}\mathcal R_t(X\|Y)R0​(X∥Y)=inft>0​Rt​(X∥Y). Thus a=0a=0a=0 is included by this infimum definition, and zero eigenvalues are permitted in both decompositions. In addition,

ker⁡s⊆ker⁡r⟺∀i,j, pij>0⟹qij>0.\ker s\subseteq\ker r \quad\Longleftrightarrow\quad \forall i,j,\ p_{ij}>0\Longrightarrow q_{ij}>0.kers⊆kerr⟺∀i,j, pij​>0⟹qij​>0.

Finally, for any density matrices r′,s′r',s'r′,s′ on JJJ, any nonnegative v′,w′:J→Rv',w':J\to\mathbb Rv′,w′:J→R, and any U′,V′U',V'U′,V′ satisfying both unitary identities and the decompositions r′=U′diag⁡(v′)U′†r'=U'\operatorname{diag}(v')U'^\daggerr′=U′diag(v′)U′†, s′=V′diag⁡(w′)V′†s'=V'\operatorname{diag}(w')V'^\daggers′=V′diag(w′)V′†, one has, for every i,j∈Ii,j\in Ii,j∈I and k,l∈Jk,l\in Jk,l∈J,

(vivk′)∣((U⊗U′)†(V⊗V′))(i,k),(j,l)∣2=pij vk′∣(U′†V′)kl∣2.(v_i v'_k)\left|( (U\otimes U')^\dagger(V\otimes V'))_{(i,k),(j,l)}\right|^2 = p_{ij}\,v'_k|(U'^\dagger V')_{kl}|^2.(vi​vk′​)​((U⊗U′)†(V⊗V′))(i,k),(j,l)​​2=pij​vk′​∣(U′†V′)kl​∣2.
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