Characteristic function of :
ProvedBosonicLearn.fugacity_characteristicbosonic-gaussian-statesquantum-informationquantum-learning
The Weyl characteristic function of the passive Gaussian state with fugacity matrix .
Let be a positive semidefinite complex matrix with positive definite, and . For put , so that . Then the iterated series defining converge and
Role. This is the computation behind clause 1 of Lemma 2 of the paper: is Gaussian, with covariance read off from (BosonicLearn.fugacity_isGaussianState).
Formalization Note The left-hand side is the published Support50.P14.oscillatorCharacteristic ( with interleaved quadrature index , ), and the convergence statement is OscillatorCharacteristicSummable. The quadratic form is written as the double sum .
Preamble
import Mathlib import Definitions.Def_BosonicLearn_fugacityCoeff import Definitions.Def_QAlgorithms_Papers_Independent_Support50_P14 open QAlgorithms.Papers.Independent.Claimed50.P47 (Occupation) open QAlgorithms.Papers.Independent.Support50.P14 (oscillatorCharacteristic OscillatorCharacteristicSummable) open scoped Matrix ComplexOrder
Formal statement
namespace BosonicLearn
/-- **Characteristic function of `det(I − R)Γ(R)`.**
Chen, Anna Mele, Anna Mele, Preskill, *When are bosonic Gaussian states classical to learn?*, arXiv:2609.26705, Lemma 2, Eq. (21), p. 20: "the corresponding passive Gaussian state is ρ = det(I − R)Γ(R)"; §1.1, Eq. (2), pp. 4–5
(characteristic function of a Gaussian state).
Let `R` be positive semidefinite with `1 − R` positive definite and `ρ_R = det(1 − R)Γ(R)`. For `ξ ∈ ℝ^{2n}` let
`ζⱼ = (−ξ_{pⱼ} + i ξ_{qⱼ})/√2`, so that the Weyl operator is `e^{iξ·r̂} = D(ζ)`. Then
`Tr(ρ_R e^{iξ·r̂}) = exp(−½ ζ† (1 + R)(1 − R)⁻¹ ζ)`,
and the iterated sums defining the left-hand side converge.
**Formalization Note.** The left-hand side is the published `Support50.P14.oscillatorCharacteristic`
(`∑ᵤ ∑ᵥ ρ_{uv} W(ξ)_{vu}`, interleaved quadrature index `(j, 0) = qⱼ`, `(j, 1) = pⱼ`), and the convergence statement
is `OscillatorCharacteristicSummable`. -/
theorem fugacity_characteristic {n : ℕ} (R : Matrix (Fin n) (Fin n) ℂ) (hR : R.PosSemidef)
(hR1 : (1 - R).PosDef) :
OscillatorCharacteristicSummable (fugacityCoeff R) ∧
∀ ξ : Fin n × Fin 2 → ℝ,
oscillatorCharacteristic (fugacityCoeff R) ξ =
Complex.exp (-(1 / 2 : ℂ) *
∑ j, ∑ k, star ((-(ξ (j, 1) : ℂ) + Complex.I * (ξ (j, 0) : ℂ)) / (Real.sqrt 2 : ℂ)) *
((1 + R) * (1 - R)⁻¹) j k *
((-(ξ (k, 1) : ℂ) + Complex.I * (ξ (k, 0) : ℂ)) / (Real.sqrt 2 : ℂ))) := by sorry
end BosonicLearn
Source
Chen, Anna Mele, Anna Mele, Preskill, When are bosonic Gaussian states classical to learn?, arXiv:2609.26705, Lemma 2 and Eq. (21), p. 20 (passive Gaussian states as det(I − R)Γ(R)); §1.1, Eqs. (2) and (4), pp. 4–5 (Gaussian states, characteristic function, passive Gaussian states)