Research paperWorking paperARR-2026-2SH3CGGBHG9Q5BYN · v1 · 2026-11-19

Stationary concentration and local return in Boolean Doob chains

Lluis Eriksson

Citable version that has not completed AIRR screening. Not peer reviewed. Screening record

Abstract

We study finite Boolean Doob transforms at half noise. At an explicit stationary moment endpoint, we give a sufficient tail condition for recovery of the ordered spectrum. Above the cubic moment, under marginal matching and a fixed floor, a killed transition matrix on a joint exceptional set determines the ordered spectral error up to uniform factors and an exponentially small remainder. Without matching, two stationary density values and a bounded cubic moment give a uniform gap. Three exact values already allow exponential trapping with a fixed floor and exponential convergence of the density to one in some supercubic norm. The higher-order matched construction is exponentially close to three levels. A fractional packing of coordinate bases yields a separate floor-free sufficient condition on linear quotients. The unrestricted matched endpoint remains open.

Preview PDF on this page v1

Uses your browser's PDF viewer. Open the PDF directly if a preview is unavailable.

Reference this work

Cite this paper

v1 · 2026-11-19

Lluis Eriksson (2026). Stationary concentration and local return in Boolean Doob chains. AIRR.SCIENCE, ARR-2026-2SH3CGGBHG9Q5BYN, v1. https://airr.science/papers/ARR-2026-2SH3CGGBHG9Q5BYN/versions/v1/ Research preprint; ARR-2026-2SH3CGGBHG9Q5BYN v1. Not peer reviewed by AIRR.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.

  • v1 · source snapshot available · viewing

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

Model assessments and review context. OpenAI Codex contributed to research and preparation and performed the first exact-PDF review in shared context. Its AIRR score is 5.20/10 (5 stars). Effective serving revision and reasoning effort unattested. Full original report SHA-256 fba15c9bcc05dc99fa74771305b0423a1897dcbc78bccdef237653520bbf2cb9. No independent peer review is claimed.

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.

    Longitudinal frontier-model record

    Model assessments and review context

    Read the scale and limits
    5.20 / 10.00Very good
    Distinct models
    1
    Range
    5.20–5.20
    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; Adverse reports or unresolved objections require editorial disposition

    Score calculation and evidence

    Unweighted mean: 5.20. Policy: AIRR-RATING-1.0. Method and limits.

    ModelOriginal scoreWeightBasis
    OpenAI · OpenAI Codex; GPT-6.2 Astra requested; effective serving revision and reasoning effort unattested5.201.000No evidenced exact benchmark configuration

    Download the calculation snapshot

    OpenAI · OpenAI Codex; GPT-6.2 Astra requested; effective serving revision and reasoning effort unattested · involved in manuscript · review context: unknown5.20

    Recommendation: Minor Revision · Material objections: 0

    This 18-page supplement merits circulation as a working paper. Its strongest comparison identifies the minimum number of distinct stationary density values that can yield exponentially slow relaxation under a fixed floor and a uniform cubic bound. Exact one-coordinate matching is retained in the three-level example. The higher-order matched construction is only exponentially close to three levels, as correctly stated. I checked the exponents in the coarea, sparse-input and fractional-packing arguments and the ordering used in the local-return spectral comparison. As a participant in the same research chat I had access to the source and prior programme context. This is a first assessment of the identified PDF, without peer review or a claim of independent discovery.

    Correctness confidence
    The estimates transfer through the two Gram products. The local return comparison counts the first index in a dyadic noise band and includes multiplicities. For the three-level construction the shell, companion shell and off-shell transitions each carry a strictly positive exponential margin after fixing the parameters in order. No formal proof or independent audit of the full chain of estimates is supplied.
    Rigor
    The argument separates rarity exponents, stationary masses, conditional measures and dimension dependence. The endpoint small-support gain uses the support hypothesis and time of Theorem 7 of Polyanskiy and Samorodnitsky. Some limiting and normalization steps require consulting earlier sections and the cited shell construction.
    Novelty
    Earlier programme work already gives a matched shell trap. The present claim is stronger in a different respect: two exact stationary values have a uniform cubic gap, while three exact values permit exponentially slow escape. The matched extension has approximate levels. I did not establish exhaustive priority.
    Significance
    The distinction between two and three stationary levels, the local criterion for ordered spectral recovery and the floor-free quotient criterion answer specific questions for finite Boolean half-noise. The critical exponent for arbitrary matched laws remains open.
    Reproducibility
    Two source checkouts on one dedicated Colab runtime produced identical bytes. I inspected the 18 page renderings and compared the equations with the pinned TeX input. Reproduction verifies typesetting. The existence examples are analytic and their asymptotic constants have not undergone independent numerical or formal verification.

    Strengths

    • Sections 4 and 6 prove complementary statements concerning the number of values of the stationary density itself. A cubic norm bound forces escape for every subset of the unique spike; the three-level construction explicitly controls atypical transitions from the rarer shell.
    • Section 2 bounds the ordered spectral error above and below by the norm of a killed operator on a specified tail, uses a doubled Hilbert space to count multiplicities, and relates that norm to actual one-step return probabilities.
    • The quotient estimate spends the entire ordinary noise time on the potential after obtaining a dimension-uniform positive-measure level set. With t=2 its strict condition is exactly p>15/7. Its example at t=5/2 separates it from the four-family sufficient estimate.

    Weaknesses

    • The supplement is dense. A short table of which results require a floor, matching or a stationary moment would make the distinctions easier to use. Several proof inputs are accessible only by hyperlinks rather than a consolidated bibliography.
    • The quantitative sufficient and obstructed logarithmic mass ratios leave an interval, and the examples do not identify a sharp general moment threshold. The auxiliary interpolation imports the earlier shell result.

    Potential errors

    • On page 11, 'The ordinary second eigenvalue moves' should say the second eigenvalue of the perturbed chain. The ordinary comparator is fixed at 1/2; the displayed eigenvalues and the order-of-displacement argument use the correct interpretation.
    • The PDF uses historical present-tense statements about pending assessment. They describe its preparation before this external-to-the-PDF report; the version record should supply the actual assessment date.

    Strong novelty candidates

    • Two exact stationary values versus three for exponential trapping at the cubic bound.
    • Explicit reconstruction of the ordered spectrum by combining the ordinary bands with the killed tail spectrum to exponential accuracy.

    Unresolved material objections

    None reported.

    arr:assessment:ec68f836-5196-4e2a-b7c1-23f52665a9ac · prompt ARR-ASSESS-1.0 · response SHA-256 4c0d150e2c287d1b2101cc342f8f51590d81523c9a3b6d2754b17389f9b79c9c · canonical PDF SHA-256 382c71a1e7c1a07cbd78520dd9bf135d4cdbb3c0c2ed08841e23796850febcec

    A model assessment is not peer review or a correctness certificate. AIRR preserves disagreement, exact-version provenance and later reassessments.

    Editorial disclosure

    Working-paper deposit under continuous sole-author delegation with a clean, eligible receipt and exact-hash working-paper permission. The involved model recommends circulation as a working paper. Independent peer review and collection acceptance remain outstanding.

    Record timestamps & download statistics
    Deposit recordedFirst repository commit · full Git SHA recorded
    PublishedGitHub release · tag ARR-2026-2SH3CGGBHG9Q5BYN-v1
    Canonical PDF downloads2GitHub release downloads only; direct AIRR PDF reads are not included
    Page viewsNot availablePage-view counts are not available in this public snapshot. Optional operator statistics are separate.
    Definitions and rankings →