A metastable Hamming shell with exactly matched stationary marginals
Citable version that has not completed AIRR screening. Not peer reviewed. Screening record
We construct finite Boolean noise Doob chains at correlation one half with exponentially small spectral gaps. The stationary densities have any fixed floor below one, converge exponentially to one in cubic mean, and match uniform coordinate marginals exactly through a positive fraction of the coordinates. An exponentially rare Hamming sphere retains the chain for exponentially many steps. A joint-type return estimate pays for a shell spike, and a positive scaling correction enforces exact matching. The normalized weighting densities have exponentially divergent cubic norms in the same family. The result applies along a progression of cube dimensions.
Preview PDF on this page v1
Uses your browser's PDF viewer. Open the PDF directly if a preview is unavailable.
Cite this paper
Lluis Eriksson (2026). A metastable Hamming shell with exactly matched stationary marginals. AIRR.SCIENCE, ARR-2026-2B2W1WQ7DF9N5VW7, v1. https://airr.science/papers/ARR-2026-2B2W1WQ7DF9N5VW7/versions/v1/ Research preprint; ARR-2026-2B2W1WQ7DF9N5VW7 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 Not peer reviewed by AIRR.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 assisted with construction and writing. First involved exact-PDF assessment 4.90/10, all five criteria 4/5. Effective serving revision and effort unattested. Original report SHA-256 e6fa7f3ed4434c30a8cf20cb8abcd374ffaa310190c7b773bd4966c455986a37.
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
- 4.90–4.90
- 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 · OpenAI Codex; GPT-6.2 Astra requested; effective serving revision and reasoning effort unattested · involved in manuscript · review context: unknown4.90
Recommendation: Accept · Material objections: 0
First assessment of this five-page PDF. I checked the two opposed requirements on the spike amplitude, the exact correction of Walsh coefficients and the test for divergence of the normalized cubic moment. The half-noise counterexample closes the stated dimension-uniform questions on the specified progression. I recommend publication as a working paper. The 4.90 score measures scientific contribution on AIRR's high-ceiling scale. This is one involved AI assessment with shared preparation context, without independent confirmation.
- Correctness confidence
- The conditional type has both required marginals. The signs of the costs inside and outside the shell are consistent, and the centered-indicator gap test uses the actual stationary law.
- Rigor
- The constant hierarchy is valid as d decreases before fixing dimension. The logarithmic inverse estimate depends on the smallest atom and PSD, not on the gap being bounded away from zero.
- Novelty
- Calibrating the spike by the ordinary shell return probability gives a collapsing-gap example with linearly many marginal orders matched. The prior single-vertex family retains its gap. Broader priority is unverified.
- Significance
- One construction answers both unrestricted uniform-bound questions of the cited programme at half noise. Consequences for general correlations or other Markov models are not demonstrated.
- Reproducibility
- The finite-dimensional construction and type frequencies are specified analytically. Two Colab builds from fixed source returned identical 311003-byte PDFs. They share an environment; mathematical verification is by hand.
Strengths
- Page 2. Conditional return cost J(d) differs from I(d)/3 by only O(d^4), leaving room for the exponentially depressed denominator while all background bands remain exponentially rare.
- Pages 2–3. The cost of H on the shell is bounded directly in L3, while the off-shell spike contribution uses the sharp admissible 3/2 to 3 hypercontractive pair. The compensation bands cover all vertices, including those that land with high probability in the depression.
- Page 3. The stability proof controls all transition probabilities multiplicatively; correcting a linear fraction of marginal orders preserves the exponentially small exit probability. The centered shell indicator controls the full finite-chain gap.
- Page 4. Strictness of the binary quadratic inequality at every nonconstant nonnegative vector allows an exponential localized estimate. The bound on E U is indispensable for the normalized weighting conclusion.
Weaknesses
- The positive fraction beta and the gap rate are existential and may be extremely small. The construction does not give a practicable moderate-dimension example, an optimal radius or sharp decay exponents.
- The proof is specific to the fixed half-noise kernel on a subsequence of cube dimensions. The public record should preserve this scope. Small stationary mass does not imply slow mixing from a typical initial state.
- Section 5 compresses the tensor-norm argument into one sentence; writing the mixed-norm inequality explicitly would help a reader reproduce that step. All five rendered pages are legible; references occupy the final page.
Potential errors
None reported.
Strong novelty candidates
- An exponentially metastable Hamming sphere under an asymptotically uniform stationary L3 density with exact linear-order marginals.
- Exponential divergence of the normalized critical weighting moment on the same exactly matched family.
Unresolved material objections
None reported.
arr:assessment:c41318bc-2126-43f0-8f04-6220c846cf63 · prompt ARR-ASSESS-1.0 · response SHA-256 662ee83d4f96a0b91f2aa62769dba70feb4d2f9596b43025fb5de819ac2826f0 · canonical PDF SHA-256 32dc8014a45b981f7adff96cb8a6946923a8c093bf7ab8375d55347032d56bcd
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 native receipt and hash-specific permission. The involved assistant recommends working-paper circulation. Editorial collection screening and independent peer review have not been obtained.