Reproducibility audit · 2026-08-31

Status: reconstruction recipe is independently reproducible, but the dossier is not a complete E1 certificate package as currently linked.
Independent rerun under Python 3.13.5, NumPy 2.3.5, complex128, using the Frobenius norm reproduces the finite E47 core: rank(P)=47, ||P²−P||F=6.654705070535472e−15, ||KP||F=1.821462616389021e−12, and |Tr(P)−47|=1.4210854715202004e−14, all below the dossier tolerance 1e−9.
The dossier's displayed residual tuple (4.34e−14, 1.89e−11, 2.49e−13, pumping trace 1.67e−15, semigroup composition 6.84e−17) cannot be compared bit-for-bit because the claimed deterministic verification script and JSON certificate are not linked here and were not located by canonical Drive search. The exact projector-construction convention, pumping-test state, and any random seed are therefore missing from the accessible record.
Missing package metadata: issuance package/library versions beyond the independent rerun environment; explicit numerical datatype in the original package; named norm for every displayed residual; immutable hash/commit for the submission validator/results; canonical Google Drive links for the script and JSON certificate.
Registry boundary: this submission dossier does not create a new mathematical citizen. Its selector theorem maps to existing citizen C-38 and existing borough/certificate families; no new Machine Certificate or citizenship authority is inferred from this dossier.
No physical interpretation or empirical claim is promoted by this audit.

<aside> 🎯

Editorial decision: submit as a Physical Review A Regular Article, primary section A-2E — Quantum Information Science, secondary fit A-1E — open quantum systems and decoherence. PRX Quantum is not recommended at the present evidence level because the corpus does not yet establish a hardware resource advantage, experimental validation, or a field-level milestone.

</aside>

Submission thesis

The defensible paper is not a claim that decoherence-free subsystems or dissipative preparation are new. Its contribution is the constructive synthesis of four pieces:

An arbitrary selected set of total-spin labels is compiled into a Casimir polynomial whose kernel is exactly the chosen union of isotypic components.
The positive penalty gives an exact orthogonal-selector limit and a sharp optimal fixed step.
The distinctions among a Hilbert-space filter, a completely positive trace-nonincreasing operation, a CPTP dephasing channel, a dark subspace, and a quantum error-correcting code are made explicit.
A new, full-carrier GKSL reset construction pumps population into the selected subspace deterministically and has a closed-form semigroup and Liouvillian gap.

The proposed title is Casimir-polynomial compilation and dissipative preparation of symmetry-selected quantum subspaces.

Evidence map

Notion source	Submission use
The Mathematical City — Recursive Intelligence, E47 & K47/125	Canonical three-spin-2 carrier, rank-47 selected subspace, scope discipline, and publication route.
‣	General selector theorem, optimal constant step, and finite specialization checks.
‣	Correct CPTP block-dephasing channel, fixed algebra, and block dimensions.
‣	Correct separation of postselected projection, collective-noise noiseless multiplicities, and full Knill–Laflamme correction.
‣	Valid model-specific dark-subspace claim for jump `L=K`; excluded as a generic environmental-noise claim.
‣	Corrected `K(12)=-108`, current spectral endpoints, and evidence-provenance boundaries.
‣	Direct numerical residuals and reproducibility architecture.
‣ and ‣	Used to identify the missing obligation: neither establishes full 125-dimensional deterministic population transfer. The submission supplies that completion separately.

Core formal result

For H_(s,n)=V_s^(⊗n), total Casimir C_(s,n), and a nonempty selected label set S,