Skip to content

Exclude rank saving for an arbitrary orthogonal basis and one bridge (no κ) - #7695

Open
aliromman wants to merge 2 commits into
CrocSwap:mainfrom
aliromman:codex/consecutive-result-035
Open

aliromman wants to merge 2 commits into
CrocSwap:mainfrom
aliromman:codex/consecutive-result-035

Conversation

@aliromman

@aliromman aliromman commented Nov 18, 2026 •

Copy link
Copy Markdown
Contributor

Result and scope

An arbitrary even orthogonal binary projector basis plus one fresh bridge cannot save paid rank at its actual pattern relation budget. Every old type and the bridge must occur. Coordinate old projectors are allowed; extra types are excluded. The exponent group is paid-generated, torsion-free and abelian, the saturated relation rank is at most the alphabet's off-pattern dimension k, and the full ambient endpoint is diagonal monomial.

Read the proof, exact receipt, verifier and scope record. All bridge calls form one consecutive block with exponent sum zero. That block cancels on every ambient input. Physical target exponents are sums of disjoint paid subsets, so total rational target rank is at most paid rank. The proof identifies the exact diagonal kernel by a Boolean partition, proves star independence for every old basis through an idempotence obstruction, and matches target multiplicities in the original coordinates.

As a source-dependent corollary, the complete no-edge pattern-five, irreducible-image-seven family cannot save rank with at most five exponent relations. This family is outside the extremal image hypothesis of the preceding edge-core theorem. The arbitrary-basis theorem also covers one indecomposable old block and coordinate-old cases with k=1. It is not a classification of all subextremal pools.

No new kappa, native supplier, rounding contract or complete multiplier invoice is supplied. The basis change is a proof device, not free physical transport. Extra types, omitted types and larger relation budgets remain open. The selected frontier and all five assumptions remain unchanged. This completes the advance claim.

Evidence and reproduction

Normal and optimized CPython 3.12.16 replay passed on macOS arm64. All 20166 GL2/GL4 bases pass exact Boolean-kernel, partition, physical-support and star-quotient checks. Twelve stated domains contain 130320 degree-two type orders. Three domains also check all 20880 full Laurent words in original coordinates: the coordinate four-basis, one orthogonal four-block and the two normal triples. Exactly 5280 have the predicted diagonal endpoint; every other full word has a nonzero off-diagonal entry.

Eight ambient fixtures through old dimension ten include repeated old types, nonsymmetric fresh factors, unused coordinates and dirty-spectator columns. They retain the exact relation budget and check complete operator equality, actual pattern/image dimensions and physical target assignments. Omitting one bridge changes a dirty column. A single-block control shows that a coarser allowed-entry quotient misses the obstruction. Eighty-four diagonal perturbations check the idempotence argument; the odd-dimension guard rejects.

Metadata, committed bytes, modes, source ancestry, eight-file export and archive integrity pass. A read-only optimized export replay passes without Git or a usable PATH; all nine payload/manifest hashes and modes stay unchanged. Finite checks corroborate the written all-width and all-length argument.

With CPython 3.12 and its standard library, from the repository root:

python3.12 -I -W error -B research/orthogonal-basis-bridge-obstruction/verify.py
python3.12 -I -W error -B -O research/orthogonal-basis-bridge-obstruction/verify.py
python3.12 -B scripts/research_package.py --package research/orthogonal-basis-bridge-obstruction
python3.12 -B scripts/check_research_commit.py --package research/orthogonal-basis-bridge-obstruction --check-source-history --check-modes

Expected markers:

PASS_ORTHOGONAL_BASIS_BRIDGE_OBSTRUCTION
PASS_REGISTERED_RESEARCH_PACKAGE
PASS_RESEARCH_COMMIT_BYTES

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. Full predecessor package replays, selected cold replay, Lean and Docker were not run. Independent human and formal-kernel review are unclaimed. The ambient identity is a formal scalar statement, not a finite-precision native implementation.

Sources and credit

Prepared by @aliromman with substantial OpenAI Codex assistance. Ali Romman's PR #7694, exact proof, supplies the prior extremal edge-core theorem and adapted matrix/Laurent routines.

@CubeJerry supplied the original bridge chronology and associated-graded argument in PR #6782, exact word proof, and the original Laurent recurrence. This adaptation is not an independent review. The Boolean partition is an elementary algebraic fact, not a claim of general mathematical priority.

The no-edge corollary uses Ali Romman's PR #7691, exact classification, and PR #7690, exact image bound. Those sources retain unpaidmoon's normal-triple and positive-link credit and Nick Petalas's nonsymmetric-link credit. Their constructions and pool geometry remain prior work. Credit details identify the new word result and source roles. New work is Apache-2.0.

Base: canonical main e54523517d0f03e58352e647c403a5af3e525a57, priority revision 8. Card bit-new-register-model, scope changed to this stated arbitrary-basis word class. Refreshed open PRs, Discussions and handoffs show no competing general orthogonal-basis bridge claim. Closed overlap uses the complete 6175-record snapshot with a fresh updated-order delta confirming no later closed changes. Sparse local and unpublished work remain coverage limits. No new native construction is proposed; a future saving requires a different stated alphabet or relation budget and a full paid/target invoice.

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