Skip to content

Referee-proofread and harden the Erdős 625 manuscript - #58

Draft
SamPetkov wants to merge 20 commits into
agent/625-self-contained-bulletproof-manuscriptfrom
agent/625-referee-readable-tier-one-pass
Draft

Referee-proofread and harden the Erdős 625 manuscript#58
SamPetkov wants to merge 20 commits into
agent/625-self-contained-bulletproof-manuscriptfrom
agent/625-referee-readable-tier-one-pass

Conversation

@SamPetkov

Copy link
Copy Markdown
Owner

Purpose

This PR is stacked on #57. It performs a second, referee-oriented pass over the self-contained Erdős 625 manuscript. The goal is not cosmetic expansion: it reduces the amount of notation a reader must hold at once, replaces several compressed proof interfaces by exact statements, and makes the quantitative constant independently reproducible.

The theorem status remains fail-closed. This PR improves publication readiness but does not promote the full-sequence theorem until the mathematical and Lean gates in Appendix A are closed on one integrated commit.

Readability pass

  • rewrites the abstract around the theorem, obstruction, method, and consequence;
  • replaces the boxed repository warning with a short typographic verification note;
  • presents an unnumbered main theorem rather than Theorem 0.1;
  • reorganizes the introduction into previous work, four main ideas, the role of the four-size support, and a compact proof guide;
  • condenses the proof-object preliminaries and removes duplicated graph definitions;
  • replaces the technical proof synopsis with a reader-oriented four-stage guide;
  • converts the generated legacy paragraph proofs and manual terminal squares to AMS proof environments;
  • normalizes log 2 notation and several dense transitions without changing hypotheses or summation domains.

Section 8 hardening

The high-cell section now follows the valid order of operations explicitly:

  1. sum the complete physical matching fiber;
  2. extract exact one-cell deficit ratios;
  3. pay the ambient falling-factorial loss once;
  4. sum all positive deficits;
  5. regroup full-containment references by endpoint table;
  6. transport the table sum to the partial-diagonal weights;
  7. insert the phase estimates.

The square-free endpoint comparison now has a proof. It derives the local atom ratio, the identity

g(t) / g(s) = 2^(d s + binom(d,2)),

the ambient falling-factorial inequality, and the cancellation producing the Q_ij factors. The section also isolates the reusable finite core that works for any finite endpoint alphabet with matching-supported selected cells.

Section 9 hardening

The former informal “thresholds of F” phrase is replaced by an exact function Phi_F(r'), and the cycle-space expansion is an equality before expectation.

The joint prescribed-cell estimate is derived from factorial moments. This makes explicit why demands in cells sharing rows or columns may be expanded simultaneously.

The estimate

q_ab <= C theta_ab^2

is now a named lemma with a complete proof based on log-convexity and endpoint control. The upper endpoint contributes

-U^2/24 + O(U log U)

in the exponent. The complementary residual regime now derives separately the local-reward and cycle-space bounds before combining them.

Exact constant ledger

Adds 625/experiments/check_constant_ledger_v3.py, which verifies the strengthened four-support certificate using exact rational arithmetic. It encloses 2^(1/20) by two rationals whose twentieth powers bracket 2, bounds log 2 by a rational series with an exact remainder, and checks the mean bracket, four omitted-tail bounds, and the final 139/500 ledger without floating-point decisions.

The final section now explains this certificate in the paper and uses a deterministic error sequence in the phase-resolved probability statement.

Publication-rank boundary

The accompanying audit records the distinction between presentation quality and theorem status. If the full-sequence theorem closes as stated, the contribution resolves the exceptional phase left by the existing approximately-95%-of-integers result and gives a quantitative n/(log n)^3 lower bound. The remaining rank-limiting work is therefore closure and independent replay of:

  • the concrete phase/chromatic-tail package;
  • the signed four-size first-moment assembly;
  • the complete partial-diagonal theorem;
  • the global high-skeleton/residual assembly;
  • the final theorem-facing Lean root.

Validation

The dedicated workflow now:

  • compiles all manuscript tools;
  • runs the exact rational constant checker;
  • regenerates the manuscript from the frozen canonical TeX blob;
  • rejects legacy boxes, paragraph proofs, manual terminal squares, placeholders, duplicate labels, and implementation-status prose in Section 8;
  • checks the exact Phi_F interface and quadratic-activity lemma;
  • compiles the AMS/BibTeX manuscript;
  • rejects undefined references, citations, and material overfull boxes;
  • verifies full-paper length and text extractability;
  • uploads representative renders of the front matter, Sections 8 and 9, and the final page.

A green workflow validates the manuscript assembly, exact constant ledger, and editorial invariants. It does not substitute for the remaining mathematical closure gates.

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

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant