# ReScience C official submission metadata
# Schema: ReScience C official metadata v1.0 (parsed by article.py)

title: '[Re] Partial Replication of the Willow Fano-Factor Results in "The Rotation Gap Is Not An Error"'

authors:
  - name: Ivan Nestorov
    orcid: 0009-0006-7940-9539
    email: volmax.core@gmail.com
    affiliations: 1,*

affiliations:
  - code: 1
    name: VolMax Studio Lab
    address: Belgrade, Serbia

keywords: quantum error correction, surface codes, google willow, fano factor, replication, python

code:
  - url: https://github.com/VolMax-Studio/rotation-gap-fano-s1
  - doi: 10.5281/zenodo.22309778
  - swh:

data:
  - url: https://doi.org/10.5281/zenodo.13273331
  - doi: 10.5281/zenodo.13273331

replication:
  - cite: "Stenberg, S. (2026). The Rotation Gap Is Not An Error — Ternary Structure in IBM Quantum Hardware. arXiv:2604.11963v1 [quant-ph]."
    bib: stenberg2026rotation
    url: https://arxiv.org/abs/2604.11963
    doi: 10.48550/arXiv.2604.11963

abstract: |
  We present a pre-registered partial replication of the empirical quantum error-correction
  detector statistics reported in Section II F and Table III of Stenberg (2026, arXiv:2604.11963v1).
  Analyzing raw detector bitstream records from the 105-qubit Google Willow archive (Zenodo DOI:
  10.5281/zenodo.13273331, 420 experiments across code distances d in {3, 5, 7}), we recompute
  all four primary and secondary claims under strict cryptographic and algorithmic verification gates.
  We confirm that: (1) the aggregate empirical Fano factor reproduces to F = 2.4157 ≈ 2.42 ± 0.36 (Claim W1);
  (2) distance-stratified Fano factors reproduce to d=3 -> 2.2942, d=5 -> 2.5935, and d=7 -> 2.7978 (Claim W2);
  (3) the one-way ANOVA across distances yields F = 59.13, p = 2.46e-23 (Claim W3); and (4) the one-sample
  t-statistic against Poisson expectation (F = 1) reproduces to t = +80.15 (Claim W4).
  We explicitly document our three-layer mixed implementation provenance: the primary detector decoding path
  is independently implemented (read_detection_counts_fast), while the statistical estimator on the analytical path
  deliberately transcribes the author's published operations under the preregistered design, with reference
  author decoding operations retained off-path for 10/10 exact bitwise equivalence verification.
  Finally, we document two epistemic limitations: a structural zero-power defect in the preregistered
  QEC round-sensitivity test resulting from Google's balanced experimental design (N_r = 28), and an exploratory
  finding demonstrating that equal-distance weighting shifts the recomputed mean Fano factor from 2.4157 to 2.5618
  (|Delta| = 0.1461, with the corresponding source-reported aggregate being 2.42) due to the predominant representation
  of d=3 experiments (64.3%) in the public repository.

bibliography: bibliography.bib
type: Replication
domain: Quantum Computation and Statistical Physics
language: Python

# -----------------------------------------------------------------------------
# Post-acceptance metadata (to be assigned/filled by editorial staff upon acceptance)
# -----------------------------------------------------------------------------

review:
  - url:
    doi:

contributors:
  - role: editor
    name:
    orcid:
  - role: reviewer
    name:
    orcid:
  - role: reviewer
    name:
    orcid:

dates:
  - received:
    accepted:
    published:

article:
  - number:
    doi:
    url:

journal:
  - name: "ReScience C"
    issn: 2430-3658
    volume:
    issue: