Eq. (207) — heterodyne detection on yields
ProvedBosonicLearn.eq_207bosonic-gaussian-statesp2o-gran-per-chapterp2o-plan-paperp2o-v1quantum-informationquantum-learning
Let be an -mode Gaussian state (grouped quadratures, vacuum covariance ). The outcome of heterodyne detection on , whose law has density , is Gaussian:
This is the starting point of heterodyne tomography: learning reduces to estimating the mean and covariance of a classical Gaussian. The paper takes it from [BMEM25, Eq. (39)].
Differences from the source:
- encoding: the heterodyne outcome law N(µ, Σ + ½I) → the measure with density (2π)^{-n}⟨α(β)|ρ|α(β)⟩ equals Mathlib's multivariateGaussian on EuclideanSpace ℝ (Fin n ⊕ Fin n); reason: same law; equality of measures also fixes the normalization
Preamble
import Mathlib import Definitions.Def_BosonicLearn_heterodyneLaw import Definitions.Def_BosonicLearn_IsGaussianState open QAlgorithms.Papers.Independent.Dependency50.P33 (State traceSize) open QAlgorithms.Papers.Independent.Claimed50.P47 (Occupation) open MeasureTheory ProbabilityTheory open scoped ENNReal Matrix
Formal statement
namespace BosonicLearn
/-- **Eq. (207) — heterodyne detection on `ρ(µ, Σ)` yields `N(µ, Σ + ½I)`.**
Chen, Anna Mele, Anna Mele, Preskill, *When are bosonic Gaussian states classical to learn?*, arXiv:2609.26705, §7.2, proof of Theorem 8, Eq. (207), p. 61: "In our covariance convention, a heterodyne measurement on
ρ(µ, Σ) produces a classical Gaussian outcome Y ∼ N(µ, C), C := Σ + 1/2 I. This is the standard heterodyne law
[BMEM25, Eq. (39)], after the fixed permutation from the interleaved quadrature ordering used there to the grouped
ordering used in this manuscript."
For every `n`-mode Gaussian state `ρ(µ, Σ)` (grouped quadratures, vacuum `Σ = ½I`), the heterodyne outcome law,
defined from the Husimi density `(2π)^{-n}⟨α(β)|ρ|α(β)⟩`, equals the multivariate Gaussian measure on `ℝ²ⁿ`
with mean `µ` and covariance `Σ + ½I`.
**Formalization Note.** Mathlib's `multivariateGaussian` is the Gaussian law on `EuclideanSpace ℝ (Fin n ⊕ Fin n)`.
The paper imports this identity from [BMEM25]; equality of measures also says the Husimi density has total mass one. -/
theorem eq_207 {n : ℕ} (ρ : State (Occupation (Fin n))) (μ : Fin n ⊕ Fin n → ℝ)
(S : Matrix (Fin n ⊕ Fin n) (Fin n ⊕ Fin n) ℝ) (hρ : IsGaussianState ρ μ S) :
heterodyneLaw ρ.coefficients = multivariateGaussian (WithLp.toLp 2 μ) (S + (1 / 2 : ℝ) • 1) := by sorry
end BosonicLearn
Source
Chen, Anna Mele, Anna Mele, Preskill, When are bosonic Gaussian states classical to learn?, arXiv:2609.26705, §7.2, Eq. (207), p. 61
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.