Skip to content

Prove a strict column-defect bound for mixed-cycle matrices - #7768

Open
aliromman wants to merge 1 commit into
CrocSwap:mainfrom
aliromman:codex/consecutive-result-080
Open

aliromman wants to merge 1 commit into
CrocSwap:mainfrom
aliromman:codex/consecutive-result-080

Conversation

@aliromman

@aliromman aliromman commented Nov 19, 2026 •

Copy link
Copy Markdown
Contributor

Scoped negative theorem: if q>=5 is prime, b>=1 and an n-by-m matrix over roots of order 3q^b has orthogonal columns, with n=q+3 and d=n-m, then m^2 < d^2(4d-3). Thus eight rows permit at most five columns, and 104 rows permit at most ninety. These are necessary upper bounds; attainment is not claimed. No new kappa follows.

Read the proof, exact checks and bound table, checker, source pins and credits. This extends PR #7762 from one missing column to an arbitrary column defect.

The projected star gives a Hermitian q-power-root block K and a d-row remainder R with K^2-3K-qI=-R*R. The quadratic is irreducible over the q-power cyclotomic field, so its two root multiplicities agree. Trace, squared trace and the rank of R give the non-strict inequality. Equality would make the rows of R orthogonal, which forces q to divide m and then q=9. This contradiction makes the bound strict. In particular, q=101,m=91 satisfies the non-strict spectral relaxation but cannot occur.

Python 3.12 and the standard library:

python3.12 -B scripts/research_package.py --package research/mixed-cycle-column-defect

Both registered modes print:

PASS_MIXED_CYCLE_COLUMN_DEFECT
Strict bound m^2 < d^2(4d-3), d=q+3-m: written proof
36 exact block fixtures; 125 polynomial grid points; 24 prime bounds
Eight rows permit at most five columns; 104 rows permit at most ninety
Equality obstruction, positive controls, fault controls and unchanged source replay pass

Checks cover 2,828 exact block entries and all 11,193 multiplicity cases for primes 5 through 101. Three polynomial identities are verified on a grid that determines all coefficients. Valid two-column controls, a deleted-root control and four changed block formulas distinguish the hypotheses. Synthetic block fixtures check an identity; they are not asserted to be orthogonal matrices. The general theorem is the written proof, with classical cycle and quadratic-subfield inputs.

All sixteen retained source files are authenticated. The unchanged PR7762 and nested PR7744 verifiers pass, as do normal/optimized registered replay, syntax/indentation, committed bytes/source history/modes, export and sealed read-only Git-free serial replay. Hosted CI did not certify this package. The plan job failed in the shared test_current_ci_section_matches_selected_record: the base guide names frontier6030 while the selected record requires make frontier7000-verify. The log reports 760 tests, one failure and 24 skipped; package replay was skipped. This package changes no shared CI file. Local replay remains the scientific evidence.

Ali Romman (@aliromman) supplied the projected-star proof and source checker. Nick Petalas (@petalas) supplied the two-prime projection and released question in PR #7744, with its pinned proof and prime-power arithmetic. All predecessor notices remain intact. Classical inputs are Lam–Leung Theorem 3.3 and Corollary 3.4 and Sharifi Lemma 3.2.2.

Sharp attainable column counts, other row counts, larger powers of three, more prime factors, approximate orthogonality, physical preparation and complete caller pricing remain open. The selected cold replay, Lean and Docker were not run. All five framework assumptions and the selected frontier remain unchanged. No independent human/formal-kernel acceptance or global literature priority is claimed.

Base: canonical main e54523517d0f03e58352e647c403a5af3e525a57, priority revision 8. Source commits are content pins, not claims of ancestry. Prepared by Ali Romman with substantial OpenAI Codex assistance in research, implementation, verification and writing. Apache-2.0. Canonical claim.

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