Skip to content

[Grok retrieval] P4 matching-odd scalar closure: wrapping literature, size, venue #696

Description

@LightChainr

Parent: #650. Retrieval-only. No Monte Carlo, no rescoring, no Huawei.

Why

P4 is a first-pass writing draft, not a finished manuscript. It lives only on unmerged PR #688 (branch issue644-manuscript-drafts):

docs/manuscripts/p4-matching-odd-scalar-closure/README.md

Fetch with git fetch origin pull/688/head:pr688 and read that file; do not merge #688. It reports a matching-odd finite-size orientation signal (two independent 500M-replica blocks, joint χ² = 31.19/4 against global zero) and prospective failure of three scalar closures (one multiplier, one width, one pure exponent). Explicitly no p_c claim.

A prior Grok venue note exists. This ticket is an independent literature retrieval that must confirm or contradict it with primary sources.

Deliverable

One notes file, draft PR against main (notes only; do not land the P4 manuscript from #688):

notes/venue-retrieval-p4-YYYYMMDD.md

No docs/STATUS.md. Do not close this ticket, #644, or #688. Do not merge.

What to read first

  1. The P4 README on pr688 (abstract, §2 non-goals, §5 positives, §6 three scalar failures, §8 limitations).
  2. On main, only enough to know the observable names: wrapping / matching function language in docs/manuscripts/p3-projective-inference/manuscript.md §1 is enough context. Do not re-litigate P3's ladder; P4 already records the N=580 aspect ladder as underpowered (results/aspect-ladder-n580/latest.json verdict). Keep that limitation.

Retrieval questions

Q1. Has a matching-odd / orientation-odd wrapping contrast been published? Search for: wrapping probabilities on the torus; Cardy/Watts crossing; Ziff wrapping R^x, R^y, R^{xy}; Mertens–Ziff M_L(p) = R^x_L(p) - \hat R^x_L(1-p); any spin-4 / cos 4θ harmonic of a percolation wrapping observable. We need to know whether "matching-odd orientation signal" is a new named channel or a rephrasing of something already measured.

Q2. Finite-size scalar closure. Papers that test whether a single amplitude, a single scaling function, or a single exponent collapses wrapping or crossing data, and what they do when it fails (correction-to-scaling, analytic background, mixing). Closest analogue to "signal exists, frozen transfer works, scalar closure in three forms fails".

Q3. Lattice sizes. Typical linear sizes in published 2D percolation wrapping/crossing FSS (Ziff, Newman–Ziff, Mertens, Jacobsen, etc.). The draft's N = 185–425 sites (with 500M replicas) — is that a normal computational paper, or will referees demand L ~ 10^310^4? Cite actual L in comparable PRE/JSTAT papers.

Q4. Venue fit, independently.

  • Physical Review E
  • JSTAT / J. Stat. Phys.
  • J. Phys. A
  • Physical Review Letters
  • SciPost Physics

For each: fit / stretch / wrong, with publisher scope sentence and one analogue. In particular: would PRL take a finite-size signal with no new exponent, no operator ID, and no p_c?

Q5. Readiness. The draft itself lists missing generated tables, missing bib, and a second independent doubling curve as referee-ask. Is that last item actually required by PRE practice, or will a clearly labelled single-lineage limitation pass?

Prior Grok assessment to stress-test (do not copy)

  1. After a real writing pass, PRE or JSTAT is the home; not PRL / SciPost main / any threshold journal.
  2. Among the four paper objects, this is the most "new physical fact" if the signal survives larger N; otherwise it is a finite-size artefact paper.
  3. N ~ 10² sites is the obvious referee objection; 500M replicas do not buy systematic finite-size control.
  4. Mechanism claim must stop at low-rank non-scalar mixing; no LCFT module.
  5. Do not use the N=580 aspect ladder as physics evidence (P3's job is the method; P4 already called the ladder underpowered).

Agree / disagree / refine with sources.

Claim boundary

  • No p_c inference. No promotion of STATUS levels.
  • Mark every source PRIMARY_TEXT_READ / ABSTRACT_ONLY / [LIT].
  • Do not import P3's projective statistic as a P4 result.

Tripwire

If you cannot name a published wrapping/crossing paper with its lattice sizes and venue, the size/venue discussion is incomplete. Do not recommend PRL without a fetched PRL percolation analogue that also lacks a new exponent/p_c.

Wait for ASSIGNED_MACHINE.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    priority:P1Bounded parallel analysis or a concrete reserve direction; not all run at once.

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions