Citation

Cite this

This is a public draft, not a peer-reviewed paper. Cite v0.3 (current). v0.2 is superseded: its headline G1 result was a wrong-basis readout artifact, caught by the live sampler and corrected with a real QFT in v0.3. If you must reference v0.2 (e.g. a rebuttal), cite the superseded DOI explicitly and link to the changelog so the lineage stays clean.

v0.2 superseded · The v0.2 "tension_with_g1" verdict has been retracted. The G1 kernel used Walsh–Hadamard + Z-basis readout while the docstring claimed Fourier-basis output; the Phase 3 live sampler caught the disagreement. v0.3 ships a real QFT (controlled-phase ladder + bit-reversal SWAPs), gives the closed form, and replaces the verdict with modal_projector_works_when_p_divides_N. See /notes/v0-3-changelog for the full story.

DOI placeholder · The Zenodo DOIs above (10.5281/zenodo.XXXXXXX for v0.3, 10.5281/zenodo.YYYYYYY for the superseded v0.2) are placeholders. Both deposits are pending; this page updates once the DOIs are minted. Until then, cite the URL + version line and we will retro-pin.

Plain text

Nadarasa, A. (2026). The Nadarasa Reduction: a windowed,
modally-restricted dihedral coset sieve (Draft v0.3, unreviewed).
https://arunquantum.lovable.app · doi:10.5281/zenodo.XXXXXXX
(v0.2 superseded — basis bug; see https://arunquantum.lovable.app/notes/v0-3-changelog · doi:10.5281/zenodo.YYYYYYY)

BibTeX

@misc{nadarasa2026reduction,
  author       = {Nadarasa, Arun},
  title        = {The Nadarasa Reduction: a windowed, modally-restricted
                  dihedral coset sieve (Draft v0.3, unreviewed)},
  year         = {2026},
  howpublished = {\url{https://arunquantum.lovable.app}},
  doi          = {10.5281/zenodo.XXXXXXX},
  note         = {Selene shot data and Guppy source at /reproduce.
                  v0.3 corrects the v0.2 G1 readout (Walsh--Hadamard basis bug
                  caught by the live sampler) and replaces the v0.2
                  'tension_with_g1' verdict with
                  'modal_projector_works_when_p_divides_N'. G2 is
                  reclassified 'kernel_cannot_test_g2'.
                  v0.2 superseded: doi:10.5281/zenodo.YYYYYYY.}
}

CITATION.cff

cff-version: 1.2.0
message: "If you cite this draft, please cite v0.3 (current). v0.2 is superseded."
title: "The Nadarasa Reduction (Draft v0.3, unreviewed)"
authors:
  - family-names: Nadarasa
    given-names: Arun
url: "https://arunquantum.lovable.app"
doi: "10.5281/zenodo.XXXXXXX"
version: "0.3"
date-released: 2026-06-28
type: software
identifiers:
  - type: doi
    value: "10.5281/zenodo.YYYYYYY"
    description: "Superseded v0.2 (retracted G1 verdict; see /notes/v0-3-changelog)."
keywords:
  - quantum cryptanalysis
  - dihedral hidden subgroup problem
  - Kuperberg sieve
  - Quantinuum Guppy
  - Selene emulator

Norms we ask reviewers to apply

Cite the version. v0.3 is current. G1 verdict is modal_projector_works_when_p_divides_N; G2 is kernel_cannot_test_g2 (kernel rebuild pending). A future v0.x may reclassify G2 once a coherent two-coset combiner ships.
Cite the data. Numbers on the GAP pages come from Selene shots — reproducible via /reproduce. Please regenerate before quoting them.
Derive predictors from gates, not docstrings. That is the v0.3 methodology rule and the reason v0.2 was superseded. The live sampler is the cheapest enforcement mechanism we have.
Chen 2024 is the warning. That preprint was withdrawn within a week. Treat any claim here with the same scepticism.

v0.3 changelog G1 (corrected)G2 (not yet testable)Reproduce locally