Published November 6, 2026 | Version v1

Code and Lean proofs for "On the estimation of linear-optical amplitudes and probabilities with entangled input states"

  • 1. ROR icon Quandela (France)

Description

Companion code and formal proofs of the paper "On the estimation of linear-optical amplitudes and probabilities with entangled input states" (J. Senellart, P.-E. Emeriau, 2026).

  • code/: gurvits-mps, a Python package (numpy, PyTorch; CPU or GPU) for Gurvits-type additive-error estimation of linear-optical amplitudes, probabilities and overlaps for sequentially generated photonic states (matrix-product states emitted by quantum emitters). It samples the importance distribution exactly and sequentially. It also includes a Rademacher variant, a brute-force reference, validation scripts, tests and the GPU timings quoted in the paper. 
  • Lean/: machine-checked proofs (Lean 4, Mathlib) of Proposition 1 (Gaussian identity), Theorem 1 (i)–(ii) and the exactness part of (iii), and Proposition S1 (vacuum emissions), with no sorry. > Each folder's README gives the build, install and test commands.

Files

gurvits-ebs-code-lean.zip

Files (168.8 kB)

Name Size
md5:887017183be6487114316630fcf99bd7
168.8 kB Preview Download