Skip to content

Sharp axial geometry for the full birth-time law: necessity beyond the union bound - #736

Draft
LightChainr wants to merge 1 commit into
mainfrom
research/axial-quantile-criterion-20260913
Draft

LightChainr wants to merge 1 commit into
mainfrom
research/axial-quantile-criterion-20260913

Conversation

@LightChainr

Copy link
Copy Markdown
Owner

Result delivered, not a new enumeration assignment

For honest axial square-site tori with 2 <= w_n <= m_n and w_n*m_n -> infinity, under the explicitly cited critical square-site RSW, site sharpness and matching-duality inputs,

all fixed interior Q_n(u) -> p_c
    iff convergence is uniform on compact u-intervals
    iff F_n tends to the threshold step
    iff log(m_n)/w_n -> 0.

The new part is necessity, including logarithmic-width sequences. An upper union bound failing to vanish was never a proof of necessity. Here a seam-closed ring of crossing events has probability at least exp(-eta*w) at a FIXED p_eta < p_c; independent vertex-disjoint transverse bands amplify it. This forces P0 -> 0 and keeps each fixed lower quantile away from p_c on a witnessing subsequence. The median/balance root is deliberately NOT included in that obstruction.

This complements #718/#735: on m=ceil(exp(d*w)), w -> infinity still suffices for the balance root, but not for the full law. No new numerical p_c, exponent, or continuum identification. No necessity theorem for arbitrary oblique periods. The missing uniform oblique-corridor lemma is stated precisely as the next question.

Files and validation

  • notes/axial-quantile-geometry-criterion-20260913.md: proof, quantifier order, source reading boundaries and prior-art caveat.
  • scripts/axial_ring_gluing.py: free-rectangle crossing BFS versus an independent lifted-graph winding detector.
  • tests/test_axial_ring_gluing.py: 3 mathematical tests, passed locally.
  • results/research-control-20260913/axial-ring-controls.json: 21,760 exhaustive configurations, 2,365 deterministic uneven-cell controls, zero implication failures; exact Fraction FKG checks at 1/3, 1/2, 2/3.

At 4x2 the sufficient ring event has 17 configurations, while winding has 35: the construction is explicitly not an event identity. Controls do not prove RSW or the all-size theorem.

Python compilation passed. Full repository CI not run by this delivery. No existing code, production data, frozen design or old result is replaced. Independent all-size proof review and systematic novelty comparison remain publication work. This proof is supplied as a consequence of named inputs, not as a claim to have invented RSW/strip gluing.

Owner-delegated research redirection on 2026-09-13. Advances #613; #276 stays completed. Related #716/#718/#735 and #650.

… convergence

Add the missing necessity via fixed-subcritical RSW rings and independent bands.
Separate full-law consistency from balance-root consistency; no oblique necessity,
new pc, or novelty claim. Exact finite controls and three local tests executed;
full repository CI not run.
LightChainr added a commit that referenced this pull request Sep 12, 2026
…ocation

Keep one active acceptance package on #735, incorporating the #613/#736 axial
full-law boundary. Source-visible state is a visible reserve, not a second
active dispatch. Preserve the newer #650 focus instead of overwriting it.
Oblique full-law necessity is an optional strengthening, not a new prerequisite.
Documentation and allocation JSON only; no research or experimental changes.
LightChainr added a commit that referenced this pull request Sep 12, 2026
…ull-law criterion

Add orientation-uniform staircase crossings with explicit periodic closure and
finite-group translation packing. This proves log N / ell -> 0 is necessary
and sufficient for the whole birth law on arbitrary honest integer-period tori,
while consolidating #735's weaker geometry for balance-root consistency.

Five additive files; five local tests and 135168 finite configurations checked.
No full repository CI or publication novelty claim. No existing research asset
or navigation file modified. Refs #613 #735 #736 #650.
LightChainr added a commit that referenced this pull request Sep 13, 2026
…emma text

Fourth handoff for this branch: an independent write-up of the same theorem,
prepared by a session without repository write access and delivered as a
package. It is adopted alongside the existing consolidation rather than
replacing it; the two are kept deliberately in parallel.

Added:
- docs/manuscripts/geometric-consistency/README.md   full statement and proof
- scripts/oblique_winding_necklace.py
- tests/test_oblique_winding_necklace.py
- results/geometric-consistency/oblique-necklace-controls.json

Changed:
- docs/ROADMAP.md  the active-package paragraph now points at the consolidation
  instead of describing the uniform oblique-corridor lemma as an unproved
  strengthening. That text was written before #739 and is now stale.

What this version adds over docs/manuscripts/geometric-balance/manuscript.md:
1. Damron--Lam, Section 2, is compared explicitly as the closest retrieved
   quantitative mechanism for tall thin rectangles, with the pointer to
   Grimmett's 1981 sponge-dimension work. The #613 roadmap asks for a
   strip-percolation prior-art comparison and the balance write-up has none.
2. Necessity in the bounded-ell case is argued by greedy packing of disjoint
   translates (a k-site support has at most k^2 intersecting translates, hence
   at least floor(N/k^2) disjoint ones) rather than by an exponential union
   bound. Same mechanism, different bookkeeping.
3. Ten sections including a separate discussion of what lies outside the
   result, and a nearest-source table with hypotheses and limit orders.

The two write-ups prove the same statement and cite #736 at the same commit
64d809b. Keeping both is an explicit
allocation choice, not an oversight: this is an exploratory repository and the
parallel text carries prior-art material the other lacks. Neither is presented
as an independently refereed acceptance, and no novelty is certified by
non-retrieval.

Executed here (2026-09-13):
- six local tests pass in this repository tree
- all five applied files byte-identical to the packaged copies, ROADMAP included
- the geometric control regenerates byte-identically: 8 period cases and
  12,990 crossing-event checks
- the applied ROADMAP blob base f6868c5 matches
  the pre-patch file exactly

Not established: no new p_c value, no near-critical rate, no exponent, no
external peer review. Full Matching-One repository CI has not been run.
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