Moment transfer and spectral bounds for Boolean Doob kernels
Citable version that has not completed AIRR screening. Not peer reviewed. Screening record
We study positive potentials for Boolean noise and the associated reversible Doob kernels. Under a stationary floor, a critical moment cap and exact low-degree marginal matching, we construct potentials that grow exponentially at a vertex where their noise average decays exponentially. Despite an unbounded pointwise ratio, the entire spectrum of this family converges exponentially to that of Boolean noise. A quantitative inverse estimate for symmetric stationary scaling permits exact correction of the low Walsh coefficients. We also establish weak relaxation estimates, full-chain bounds under a positive normalized potential minimum, and separate potential smoothing from stationary-density smoothing. The unrestricted stationary spectral-gap problem remains open.
Preview PDF on this page v3
Uses your browser's PDF viewer. Open the PDF directly if a preview is unavailable.
Cite this paper
Lluis Eriksson (2026). Moment transfer and spectral bounds for Boolean Doob kernels. AIRR.SCIENCE, ARR-2026-4T35VQ6MZB8F99F4, v3. https://airr.science/papers/ARR-2026-4T35VQ6MZB8F99F4/versions/v3/ Research preprint; ARR-2026-4T35VQ6MZB8F99F4 v3. Not peer reviewed by AIRR. Working paper; not admitted to the AIRR accepted collection.
Use this version link to cite the manuscript you read. A Working paper label means the version has not completed AIRR screening. AIRR screening is not journal peer review. Import BibTeX, RIS or CSL JSON into your reference manager to apply a citation style.
Version history
The paper identifier remains stable. Each version has its own immutable release, timestamp and version identifier.
Revision statement
Major revision. Sections 9-14 add weak relaxation, positive-background gap, exact matched vanishing minima and denominators, full-spectrum stability for constructed families, and limits of smoothing and variance/form comparisons.
Verification record
- Frontier-model screening
- Not assessed
- Source integrity
- Pass
- Bibliographic integrity
- Partial
- Reproducibility
- Partial
- Lean 4
- Not applicable
Recorded under ARR-VERIFY-1.0. AIRR verification and screening are not peer review.
AI assistance statement
OpenAI Codex assisted with preparation. Model assessments and review context: first exact-PDF review, involved and shared context, 4.50/10; criteria 4/4/3/4/4 out of 5. Exact serving revision and effort not independently verified. Full report SHA-256 2faff257f5d38ef41a239a0f462e187db486baa7c6bdbe26e9c55ab59fc42a80.
Frontier-model screening
Status: not_assessed. Any listed reports correspond to this exact version under ARR-SCREEN-1.0; no absent assessment is represented as a pass.
Model assessments and review context
- Distinct models
- 1
- Range
- 4.50–4.50
- Method
- bounded capability-weighted mean · exact version only
Limited evidence · 0/1 benchmark matches. This is an evidence description, not a probability of correctness.
Only one distinct model; One provider; shared blind spots possible; Fresh blinded review context not documented for every model; Benchmark match incomplete
OpenAI · Codex; requested GPT-6.2 Astra; effective serving revision and reasoning effort not independently attested · involved in manuscript · review context: unknown4.50
Recommendation: Accept · Material objections: 0
First exact-PDF assessment of the 40-page revision dated 17 November 2026. Recommend circulation as a mathematical working paper with its historical version note. A stationary density can converge to one at the distinguished vertex while the potential grows exponentially there and its noise average decays exponentially. The full spectrum stays exponentially close to Boolean noise in the constructed family. This excludes a plausible mechanism for proving a uniform gap. The result is model-specific. The reviewer shares the preparation context, had access to earlier results and issued this assessment once. Editorial acceptance and independent peer review are outside this recommendation.
- Correctness confidence
- The concentration exponents, stationary floor, ratio critical norm and multiplicative stability in Theorem 11.1 survive checks. Confidence is limited by shared context and selective analytic checking.
- Rigor
- Lemma 10.4 gives a quantitative inverse with an explicit dimension factor. Section 11 absorbs that factor after killing the low Walsh coefficients. The four-character calculation in Theorem 12.2 does not require a numerical nonzero test.
- Novelty
- The exactly matched shell construction and the coexistence of a pointwise diverging ratio with exponentially stable full spectrum are specific novelty candidates. Classical noise analysis, concentration and matrix scaling are prior ingredients. Priority remains unestablished.
- Significance
- The construction rules out an entire pointwise strategy for the stationary-gap problem while retaining all original constraints. The general full gap is unresolved.
- Reproducibility
- Two fresh source checkouts produced identical 569555-byte PDFs on one Linux runtime. Both PDFs, full source, logs, all 40 page images and PDF text are retained. No independent theorem checking was executed.
Strengths
- Pages 25-28: the two rare Hamming shells have separated one-step and two-step means. The enlarged exceptional set controls the stationary floor off both shells. The bound G >= exp(-epsilon L) is sufficient because the shell probabilities beat its cost in the critical norm.
- Page 22: for T=(I-K)/2 the absolute row sum is at most one, and its stationary L2 norm is at most one half. Splitting the series at the logarithmic crossover controls reconstruction without assuming a top spectral gap.
- Pages 27-28: the low Walsh correction costs at most exp(o(L)) times an exponentially small coefficient. Positive reconstruction gives multiplicative control of both numerator and denominator even where the denominator tends to zero.
- Pages 28-30: after smoothing the potential the first-degree coefficient is proportional to ef times (tau^(2m+1)-tau), with ef nonzero. This distinguishes destruction of exact matching from any quantitative claim about the size of the defect.
Weaknesses
- The first-page abstract extends onto page two and is difficult to navigate. Three overfull lines are logged, the largest 30.7 pt. Contact sheets showed no missing pages or content extending past a page edge.
- Section 16 is explicitly about the source state of 17 November. It cites v1 as the predecessor and predates compilation of this PDF. The version landing page must identify the actual v2 parent and date the build and this separate exact-PDF report.
- No uniform control near rho endpoints, global novelty determination, unrestricted stationary-gap proof, human peer review or Lean check. The full scientific inspection emphasized Sections 10-12 and did not rederive every earlier theorem.
Potential errors
None reported.
Strong novelty candidates
- Exactly matched admissible examples with vanishing noise denominator, diverging pointwise ratio and stable full spectrum.
- Explicit separation of potential smoothing from stationary-density smoothing under labelled matching.
Unresolved material objections
None reported.
arr:assessment:d9309a08-9112-4024-a1c3-45c39891a7df · prompt ARR-ASSESS-1.0 · response SHA-256 463bd285428cdd74cc35a4f672e6ed8e2face03071949f0500c3cbdcf84bf934 · canonical PDF SHA-256 5a7a3ddb54080e00ad0bd25e86633a37aba6b680821eb10955848c892a762207
A model assessment is not peer review or a correctness certificate. AIRR preserves disagreement, exact-version provenance and later reassessments.
Editorial disclosure
Native submission SUB-E9080D9DEB2E1425 is clean and eligible. Exact-version working-paper permission is recorded. The sole involved AI assessment assigns 4.50/10. The author owns AIRR. Editorial admission and independent peer review have not occurred.