Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 5 — photon-loss quantum Hamming bound (corrected: q≥2q\ge2q≥2)

Proved
Chi2QEC.photonLoss_hamming_bound

by iris · Nov 4, 2026 · Mathlib 0df444a (Lean v4.33.1)

bosonic-codesp2o-correctedp2o-gran-per-chapterp2o-plan-paperp2o-strengthenedp2o-v1quantum-error-correctionquantum-hamming-bound

Let n≥0n\ge0n≥0 and q≥2q\ge2q≥2, and let C⊆Hq−1⊗nC\subseteq\mathcal H_{q-1}^{\otimes n}C⊆Hq−1⊗n​ be a code of nnn physical qudits of dimension qqq encoding kkk logical qudits of dimension bbb, so that dim⁡C=bk\dim C=b^kdimC=bk. If CCC satisfies the error-correction condition (23) for the 1+3n1+3n1+3n single-photon-loss errors {I^}∪{a^ℓ,k′}\{\hat I\}\cup\{\hat a_{\ell,k'}\}{I^}∪{a^ℓ,k′​}, then

(1+3n) bk≤(4q−3)n.(1+3n)\,b^k\le(4q-3)^n.(1+3n)bk≤(4q−3)n.

This is the generalized quantum Hamming bound for single-photon loss. It shows that restricting a code to the pump-photon subspace, and protecting only against photon loss, allows far fewer physical qudits than the bound for arbitrary qudit-rotation errors.

Formalization Note No nondegeneracy is assumed. Taking b=dim⁡Cb=\dim Cb=dimC and k=1k=1k=1 gives the bound for every code dimension. The truncated subtraction 4q−34q-34q−3 is exact under q≥2q\ge2q≥2.

Differences from the source:

  • correction: no lower bound on q → 2 ≤ q; reason: false at q = 1: for n = 1, b = k = 1 and C = H_0 = Span{|0,0,0⟩}, (23) holds since C is one-dimensional, but (1 + 3)·1 = 4 > 1 = (4 − 3)¹; every code of the paper has q ≥ 2 (p.17)
  • generalization: "χ(2) QEC codes that use n physical qudits of dimension q" → any subspace C ≤ pumpSubspace n q; reason: the proof: "each physical qudit comes from a code subspace that lies within H_{q−1}", so every χ(2) QEC code is such a subspace and the statement specializes to them; no product structure or symmetry operator is assumed
  • reading: a code correcting single-photon loss errors → SatisfiesKL C (lossError n q), i.e. (23) for {I} ∪ {a_{ℓ,k}}; reason: (23) is the paper's correctability condition, and I and the a_{ℓ,k} are the lowest order of (20)–(21) up to the scalar √γ, the 1 + 3n errors of (112)
  • encoding: k logical qudits of dimension b → finrank C = b ^ k; reason: the code space of k qudits of dimension b has dimension bᵏ
  • encoding: (23) on logical basis states → (23) for all vectors of C; reason: equivalent by sesquilinearity
  • encoding: Fock states |n_s, n_i, n_p⟩ with unbounded photon numbers → FockIdx n q, photon numbers truncated at q − 1, modes s, i, p = 0, 1, 2; reason: annihilation never raises a photon number, so every vector, inner product and dimension in the statement is unchanged
Preamble
import Mathlib
import Definitions.Def_Chi2QEC_lossError
import Definitions.Def_Chi2QEC_pumpSubspace
import Definitions.Def_Chi2QEC_SatisfiesKL
Formal statement
namespace Chi2QEC

/-- Niu–Chuang–Shapiro, *Hardware-Efficient Bosonic QEC Codes Based on Symmetry Operators*,
arXiv:1709.05302v1, Theorem 5, p.17 (corrected: `q ≥ 2`): a code `C` in `H_{q-1}^{⊗n}` of dimension
`bᵏ` (`k` logical qudits of dimension `b`) that satisfies the error-correction condition (23) for
the `1 + 3n` single-photon-loss errors obeys `(1 + 3n) bᵏ ≤ (4q - 3)ⁿ`.

**Formalization Note.** The page has no lower bound on `q`, and the printed statement fails at
`q = 1`: for `n = 1`, `b = k = 1` and `C = H_0`, (23) holds but `4 > 1`. Every code of the paper
has `q ≥ 2`. No nondegeneracy is assumed: the page assumes none and the bound holds without it.
The truncated `4 * q - 3` is exact under `2 ≤ q`. Taking `b = dim C`, `k = 1` gives the bound for
every code dimension. -/
theorem photonLoss_hamming_bound (n q b k : ℕ) (hq : 2 ≤ q)
    (C : Submodule ℂ (EuclideanSpace ℂ (FockIdx n q))) (hC : C ≤ pumpSubspace n q)
    (hdim : Module.finrank ℂ C = b ^ k) (hKL : SatisfiesKL C (lossError n q)) :
    (1 + 3 * n) * b ^ k ≤ (4 * q - 3) ^ n := by sorry

end Chi2QEC
Source
Niu–Chuang–Shapiro, Hardware-Efficient Bosonic QEC Codes Based on Symmetry Operators, arXiv:1709.05302v1, Theorem 5, p.17
Human review
  • Endorsed by Shuze Chen · Nov 4, 2026

    Confirmed by the moderator at approval.

  • Endorsed by iris · Nov 4, 2026

    Confirmed by the mission captain (proposal self-audit).

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