Alphabeta Math
Physics theoremStatement: AI-adaptedProof: AI-generatedPipeline-generatedprecheck passverified 2026-10-22 (gpt-6.1-sol)
How statement and proof provenance work

The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.

  • Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
  • AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
  • AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.

These labels describe origin, not correctness: citations and verification chips remain separate evidence.

Pure loss channel and binomial thinning

Physical scope

An exact conditional one-mode pure-loss model with a vacuum environment mode, a passive two-mode unitary and a trace over the environment. The channel, binomial thinning and coherent-state attenuation are model deductions on the full infinite-dimensional finite-mode Fock space. No detector efficiency, actual loss record, apparatus realization, continuum bath, weak-coupling limit or experimental data are asserted.

Statement

Assume AC and ACω, the assigned finite-mode model, one input mode and a second mode prepared independently in vacuum ∣0⟩⟨0∣. Write F:=Fs(C) for each one-mode Fock space and F(2):=Fs(C2) for the two-mode Fock space with ordered coordinate basis e1=(1,0),e2=(0,1). In the completed product F⊗^F, define I(∣m⟩⊗∣n⟩):=∣m,n⟩ on occupation vectors, where ∣m,n⟩ is the two-mode occupation basis in that ordered mode basis. The product-basis result and finite-mode Fock occupation-basis result listed in Statement references make I extend uniquely to a unitary I:F⊗^F→F(2). For 0≤η≤1, define Cη:=(η−1−η1−ηη) and let Uη:=UCη be the two-mode beam-splitter unitary specified by the beam-splitter result in Statement references on F(2). Transport this unitary to the product space as U~η:=I∗UηI. Define the vacuum insertion isometry V0:F→F⊗^F by V0ψ=ψ⊗∣0⟩. For every trace-class operator ρ on F, the vacuum product operator ρ⊗∣0⟩⟨0∣ means V0ρV0∗, a trace-class operator by the trace-class ideal and tensor-map suppliers. Define Φη(ρ):=Tr⁡2 ⁣(U~η(ρ⊗∣0⟩⟨0∣)U~η∗), where Tr⁡2 is the trace-class partial trace over the second one-mode factor. Here CPTP means complex-linear, trace-preserving, and positive after every finite-dimensional ancilla amplification on trace-class operators. The map Φη is CPTP. For 0<η<1 and an input number state ∣n⟩, its output state is Φη(∣n⟩⟨n∣)=∑k=0n(nk)ηk(1−η)n−k∣k⟩⟨k∣, so ideal number detection has the binomial-thinning law. At η=0 this number input is mapped to vacuum, and at η=1 it is unchanged. For a coherent input ∣α⟩, its output is ∣η α⟩⟨η α∣ for every 0≤η≤1. On arbitrary input states, Φ0 maps to vacuum and Φ1 is the identity.

Facts & Assumptions

Given: A one-mode trace-class input ρ, an independent vacuum environment mode, 0≤η≤1, the passive unitary Uη with real amplitudes η,1−η, and the trace-class partial trace over the second mode.

[A1]

The finite-mode assignment adopts the mode Fock spaces and ideal photon-counting model; the independent vacuum environment is a stated model preparation, not an observed state (Finite-mode dipole and detector assignment).

[A2]

Assume AC for the finite-mode Bogoliubov implementation and ACω for the trace-class partial trace and countable Fock constructions. These are mathematical assumptions, not physical postulates (The Axiom of Choice, The Axiom of Countable Choice (ACω)).

[F1]

For the passive unitary with first-column amplitudes (t,r) and vacuum input on the second mode, UC∣n,0⟩=∑k=0n(nk) tkrn−k∣k,n−k⟩ and UC∣α,0⟩=∣tα,rα⟩ on the two-mode Fock space (Passive beam-splitter photon statistics). Under I, these equations become U~C(∣n⟩⊗∣0⟩)=∑k=0n(nk) tkrn−k∣k⟩⊗∣n−k⟩ and U~C(∣α⟩⊗∣0⟩)=∣tα⟩⊗∣rα⟩ on F⊗^F.

[F2]

For arbitrary complex Hilbert spaces H,K, the trace-class partial trace Tr⁡K:S1(H⊗^K)→S1(H) is trace-norm contractive, preserves trace and positivity, and on rank-one tensors satisfies Tr⁡KTu⊗v,u′⊗v′=⟨v,v′⟩Tu,u′; its proof assumes ACω (Separable partial trace and its basic properties).

[F3]

On trace-class operators, complete positivity means that every finite matrix amplification preserves positive trace-class block operators, and trace preservation means equality of input and output traces (Declared ideal measurements and instruments of the framework). Only these mathematical definitions are used; no measurement update is inferred.

[F4]

The one-mode occupation vectors are a complete orthonormal family, and the coherent vector is normalized with expansion ∣α⟩=e−∣α∣2/2∑nαn∣n⟩/n! (Fixed-mode number law).

[F5]

For trace-class T and bounded B, tr⁡(BT)=tr⁡(TB), and every trace-class input has a nuclear rank-one series convergent in trace norm (Nuclear bounded cyclicity for trace-class operators).

[F6]

Bounded maps tensor to bounded maps on the completed Hilbert tensor product (Bounded tensor maps: extension, norm estimate, adjoint and commutation). For finite-dimensional Cq, expand in an orthonormal basis of the first factor to see Iq⊗J preserves norm whenever J does; the reassociation of three factors preserves the elementary-tensor inner products and therefore extends by density to a unitary, using the completion construction (Hilbert tensor construction: inner product, completion and product bases).

Proof

1.1A1A2F5F6algebra

For 0≤η≤1, direct multiplication gives Cη∗Cη=CηCη∗=I2, so the full unitary-matrix hypothesis of the beam-splitter theorem holds. The product occupation vectors ∣m⟩⊗∣n⟩ form a complete orthonormal family by the product-basis result of Hilbert tensor construction: inner product, completion and product bases, and the two-mode occupation vectors ∣m,n⟩ form a complete orthonormal family by Finite-mode Fock to Schrödinger identification. Therefore the basis map defining I preserves inner products on a dense subspace and has dense range; it extends uniquely to a surjective isometry, hence a unitary. Also, the vacuum insertion V0ψ=ψ⊗∣0⟩ is an isometry by the elementary-tensor norm identity of Hilbert tensor construction: inner product, completion and product bases. For each finite-dimensional ancilla Cq, Wq:=Iq⊗V0 is an isometry by [F6]. If T≥0 is trace class, then WqTWq∗≥0 because ⟨WqTWq∗x,x⟩=⟨TWq∗x,Wq∗x⟩≥0. For arbitrary trace-class T, write T=∑jTuj,vj as a nuclear series by [F5]. Directly from the rank-one definition and the adjoint identity, WqTuj,vjWq∗x=⟨Wq∗x,vj⟩Wquj=⟨x,Wqvj⟩Wquj=TWquj,Wqvjx. The resulting rank-one series converges in trace norm since ∑j∥Wquj∥∥Wqvj∥=∑j∥uj∥∥vj∥<∞, and its operator-norm sum is WqTWq∗. Since Wq preserves inner products, [F5]'s rank-one trace formula gives tr⁡(WqTWq∗)=∑j⟨Wquj,Wqvj⟩=∑j⟨uj,vj⟩=tr⁡T. Thus adjoining vacuum is completely positive and trace-preserving.

1.2A2F2F3F6algebra

For each finite ancilla Cq, the amplification of the partial-trace map is, after the canonical unitary reassociation from [F6], the partial trace over the environment Fock space on Cq⊗^Fs(C)⊗^Fs(C). By [F2] it maps positive trace-class operators to positive trace-class operators and preserves their trace. Therefore the partial trace is completely positive and trace-preserving under [F3].

1.3A1F1F2algebra

For 0<η<1 and input ∣n⟩⟨n∣, [F1] gives the joint pure state ∑k=0nck∣k⟩⊗∣n−k⟩ with ck=(nk)ηk/2(1−η)(n−k)/2. Expanding its rank-one density and applying the rank-one partial-trace formula [F2], cross terms with k≠ℓ vanish because ⟨n−ℓ,n−k⟩=δkℓ. The reduced state is ∑k∣ck∣2∣k⟩⟨k∣, which is the displayed mixture; its ideal number-PVM law is the binomial distribution by the declared outcome model [A1].

1.4F1F2algebra

For input ∣α⟩⟨α∣, [F1] gives the product output ∣tα⟩⟨tα∣⊗∣rα⟩⟨rα∣. The rank-one partial-trace formula [F2] leaves the first factor unchanged; with t=η, this is ∣η α⟩⟨η α∣.

1.5F1F4algebra

At η=0, the transported beam-splitter expansion [F1] gives U~0(∣m⟩⊗∣0⟩)=∣0⟩⊗∣m⟩ for every m∈N0; at η=1 it gives U~1(∣m⟩⊗∣0⟩)=∣m⟩⊗∣0⟩. The finite linear span of the one-mode occupation basis is dense by [F4], so boundedness of the unitaries extends these identities to U~0(ψ⊗∣0⟩)=∣0⟩⊗ψ and U~1(ψ⊗∣0⟩)=ψ⊗∣0⟩ for all one-mode vectors ψ.

2.1F1F5F6step 1.1algebra

Conjugation by U~η is completely positive because unitary conjugation preserves positivity on every finite ancilla amplification. It preserves trace for every trace-class input by [F5], since tr⁡(U~ηTU~η∗)=tr⁡(TU~η∗U~η)=tr⁡T; by [F6], Iq⊗U~η is unitary and the same identity holds on every finite ancilla amplification. Composition with step 1.1 preserves both properties.

3.1F2F3step 1.1step 1.2step 2.1algebra

By steps 1.1, 1.2 and 2.1, the composition defining Φη maps trace-class operators to trace-class operators, is completely positive under every finite ancilla, and preserves trace. It is CPTP on the full infinite-dimensional one-mode trace-class space; no finite-dimensional Choi theorem is used.

4.1F2F5step 1.5algebra∎

For rank-one Tu,v, the partial-trace formula [F2] applied to the two isometries in step 1.5 yields Φ0(Tu,v)=⟨u,v⟩∣0⟩⟨0∣=tr⁡(Tu,v)∣0⟩⟨0∣ and Φ1(Tu,v)=Tu,v. By [F5], every trace-class input has a nuclear series of rank-one terms converging in trace norm. The partial trace is trace-norm contractive by [F2], and bounded isometric and unitary conjugations are trace-norm continuous by [F5]; the trace is trace-norm continuous by [F5]. Passing to the limit gives Φ0(ρ)=tr⁡(ρ)∣0⟩⟨0∣ and Φ1(ρ)=ρ. In particular these are vacuum and identity on density operators.

Depends on

Dependency tree · 2 levels

Sources