Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

∫A(dx)⊗G(x)⪰0\int A(dx)\otimes G(x)\succeq0∫A(dx)⊗G(x)⪰0 for a positive semidefinite matrix measure AAA and a positive semidefinite matrix function GGG

Proved
BosonicLearn.kronecker_integral_posSemidef

by Senrui Chen · Nov 13, 2026 · Mathlib 0df444a (Lean v4.33.1)

bosonic-gaussian-statesquantum-informationquantum-learning

The integral of a positive semidefinite matrix-valued function against a positive semidefinite matrix-valued measure is positive semidefinite on the tensor product.

Let ι1,ι2,K\iota_1,\iota_2,Kι1​,ι2​,K be finite sets and Ω\OmegaΩ a measurable space. For u,v∈ι1u,v\in\iota_1u,v∈ι1​ and k∈Kk\in Kk∈K let μuvk\mu_{uvk}μuvk​ be finite measures on Ω\OmegaΩ and cuvkc_{uvk}cuvk​ complex numbers such that the matrix

A(B)=(∑k∈Kcuvk μuvk(B))u,v∈ι1A(B)=\Big(\sum_{k\in K}c_{uvk}\,\mu_{uvk}(B)\Big)_{u,v\in\iota_1}A(B)=(k∈K∑​cuvk​μuvk​(B))u,v∈ι1​​

is positive semidefinite for every measurable B⊆ΩB\subseteq\OmegaB⊆Ω. Let G:Ω→Cι2×ι2G:\Omega\to\mathbb C^{\iota_2\times\iota_2}G:Ω→Cι2​×ι2​ have measurable entries with ∣G(x)ab∣≤1|G(x)_{ab}|\le1∣G(x)ab​∣≤1 and G(x)G(x)G(x) positive semidefinite for every xxx. Then the matrix WWW on ι1×ι2\iota_1\times\iota_2ι1​×ι2​ with entries

W(u,a),(v,b)=∑k∈Kcuvk∫ΩG(x)ab dμuvk(x)  =  ∫ΩA(dx)uv G(x)abW_{(u,a),(v,b)}=\sum_{k\in K}c_{uvk}\int_\Omega G(x)_{ab}\,\mathrm d\mu_{uvk}(x)\;=\;\int_\Omega A(\mathrm dx)_{uv}\,G(x)_{ab}W(u,a),(v,b)​=k∈K∑​cuvk​∫Ω​G(x)ab​dμuvk​(x)=∫Ω​A(dx)uv​G(x)ab​

is positive semidefinite.

Role. If MMM is a POVM on a first system with outcomes x∈Ωx\in\Omegax∈Ω and QxQ_xQx​ is a POVM on a second system chosen according to xxx, the joint effect is E′(C)=∫M(dx)⊗Qx(Cx)E'(C)=\int M(\mathrm dx)\otimes Q_x(C_x)E′(C)=∫M(dx)⊗Qx​(Cx​). This lemma, applied to finite blocks A(B)=M(B)∣s1A(B)=M(B)|_{s_1}A(B)=M(B)∣s1​​ and G(x)=Qx(Cx)∣s2G(x)=Q_x(C_x)|_{s_2}G(x)=Qx​(Cx​)∣s2​​, shows that E′(C)E'(C)E′(C) is positive on all finitely supported vectors of the two-system space, entangled ones included. Positivity on product vectors alone would not suffice. It is the main step of BosonicLearn.fockPOVM_compose.

Formalization Note Positive semidefinite is Matrix.PosSemidef over C\mathbb CC with the partial order of ComplexOrder; the measures enter through (μ u v k B).toReal. For constant GGG the statement reduces to Matrix.PosSemidef.kronecker. The statement is pure matrix analysis and measure theory and uses no definition of the mission.

Related platform material The space ℓ2(A)\ell^2(A)ℓ2(A) used here (Hilbert A) is the same Lean type as OAI.TensorLp.H A, and ℓ2(Nn)\ell^2(\mathbb N^n)ℓ2(Nn) is the same type as OAI.EntropyPhotonNumber.Fock n. The published definitions OAI_Analysis_TensorLp (tensor products, slices and operator lifts on ℓ2(A×B)\ell^2(A\times B)ℓ2(A×B)) and OAI_MathematicalPhysics_Fock_Operators (OAI.CrossEnsemble.operator, the bounded operator ∑i∣ui⟩⟨vi∣\sum_i|u_i\rangle\langle v_i|∑i​∣ui​⟩⟨vi​∣ of a family, with hasSum_operator and entry_hasSum) are sorry-free and can be imported. Proved lemmas of this mission that can be imported: BosonicLearn.traceDist_le_one (nuclear expansions, truncation to finite index sets) and BosonicLearn.effect_matrix_operator (effect matrices as bounded positive operators).

Preamble
import Mathlib

open MeasureTheory
open scoped ENNReal Matrix ComplexOrder
Formal statement
namespace BosonicLearn

/-- **Integrating a positive-semidefinite matrix function against a positive-semidefinite matrix measure.**
Chen, Anna Mele, Anna Mele, Preskill, *When are bosonic Gaussian states classical to learn?*, arXiv:2609.26705, §5.1, p. 30: "the t-th POVM can depend on the first t − 1 measurement outcomes x_{<t}"; the joint
effect of a measurement followed by an outcome-dependent measurement on a second system is
`∫ M(dx) ⊗ Q_x`.

Let `ι₁, ι₂, K` be finite sets and `Ω` a measurable space. Let `A(B) = (∑ₖ c_{uvk} μ_{uvk}(B))_{u,v ∈ ι₁}` be a
matrix-valued set function, where the `μ_{uvk}` are finite measures on `Ω` and the `c_{uvk}` complex numbers,
such that `A(B)` is positive semidefinite for every measurable `B`. Let `G : Ω → ℂ^{ι₂ × ι₂}` be measurable with
`G(x)` positive semidefinite and entries of modulus at most `1`. Then the matrix on `ι₁ × ι₂` with entries

`W_{(u,a),(v,b)} = ∑ₖ c_{uvk} ∫ G(x)_{ab} dμ_{uvk}(x)`  (`= ∫ A(dx)_{uv} G(x)_{ab}`)

is positive semidefinite.

**Formalization Note.** Positive semidefinite is `Matrix.PosSemidef` over `ℂ` (with the order of
`ComplexOrder`). When `G` is constant this is the positivity of the Kronecker product `A(Ω) ⊗ G`. -/
theorem kronecker_integral_posSemidef {Ω ι₁ ι₂ K : Type} [MeasurableSpace Ω] [Fintype ι₁] [Fintype ι₂]
    [Fintype K] (μ : ι₁ → ι₁ → K → Measure Ω) (hμ : ∀ u v k, IsFiniteMeasure (μ u v k))
    (coef : ι₁ → ι₁ → K → ℂ)
    (hA : ∀ B : Set Ω, MeasurableSet B →
      (Matrix.of fun u v => ∑ k, coef u v k * (((μ u v k B).toReal : ℝ) : ℂ)).PosSemidef)
    (G : Ω → Matrix ι₂ ι₂ ℂ) (hGm : ∀ a b, Measurable fun x => G x a b)
    (hGb : ∀ x a b, ‖G x a b‖ ≤ 1) (hG : ∀ x, (G x).PosSemidef) :
    (Matrix.of fun (p q : ι₁ × ι₂) =>
      ∑ k, coef p.1 q.1 k * ∫ x, G x p.2 q.2 ∂(μ p.1 q.1 k)).PosSemidef := by sorry

end BosonicLearn
Source
Chen, Anna Mele, Anna Mele, Preskill, When are bosonic Gaussian states classical to learn?, arXiv:2609.26705, §5.1, p. 30 (adaptive protocols: the t-th POVM depends on earlier outcomes); auxiliary lemma for the proof of Theorem 3, p. 67

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