Skip to content

Classify full-endpoint words on the shared-hub six-projector pool - #7711

Open
aliromman wants to merge 1 commit into
CrocSwap:mainfrom
aliromman:codex/consecutive-result-046
Open

aliromman wants to merge 1 commit into
CrocSwap:mainfrom
aliromman:codex/consecutive-result-046

Conversation

@aliromman

@aliromman aliromman commented Nov 18, 2026 •

Copy link
Copy Markdown
Contributor

This scoped proof excludes paid-rank saving on the exceptional six-projector pool from PR #7707, at its four-relation budget. Every full diagonal word must consist of two triple blocks. Exact type totals then give targets (S,S,S+T,T,T), each a sum of a disjoint subset of paid exponents. No new kappa follows.

Read the proof, receipt, checker and credits.

The exact binary image/covector alphabet is (6,5),(5,6),(7,7),(12,20),(20,12),(28,28) on five coordinates. Every type occurs, with arbitrary repetitions and no extra types. Calls are I+([a]-1)P over the paid-generated torsion-free abelian exponent group, with a saturated relation lattice of rank at most four. The endpoint must be diagonal monomial on every original ambient input. Coordinate permutations and unused spectators are included.

The new lemma gives rank eight for the nine cross-commutators modulo the first-order pattern space. The only dependence is their total sum. In degree two, this forces a common ordering polynomial in the intersection of nine independent variable-product spaces. That intersection is spanned by the product of the two triple totals, so all calls of one triple precede all calls of the other. Private rows and columns then force exact equality of the three totals in each triple. Parity alone does not prove that step.

The extra sixth projector changes the first-order equations. It is expressly excluded by PR #7710; its separate-bridge proof does not cover this case. Other maximizing pools, extra or omitted types, larger relation budgets and arbitrary ambient changes of basis remain outside the result. Physical implementation, owned storage, restoration, native costs and the complete recursive invoice remain separate. The selected kappa is still 0.00363005817, with all five assumptions retained.

With CPython 3.12 and its standard library:

python3 -B scripts/research_package.py --package research/shared-hub-triple-word-obstruction

Both registered normal and optimized commands print:

PASS_SHARED_HUB_TRIPLE_WORD_OBSTRUCTION
68760 complete Laurent words; 1584 two-block endpoints
24 larger fixtures; disjoint paid-subset targets and all boundary controls pass

All 511 nonempty commutator subsets are checked. Five multiplicity domains cover every type order at lengths six through nine in those domains. Each order has a degree-two and full Laurent check, with independent multinomial passing counts. Twenty-four larger fixtures reach 17 ambient coordinates and 15 calls, with complete target-subset checks and explicit input witnesses for crossed-block mutants. The unbounded proof covers arbitrary repetitions; the finite census uses the universal exact common-total lattice, not every integral lattice with the same parity reduction.

Controls retain the distinction between parity and exact totals, and between the paid-generated group and a larger ambient group. A five-relation interleaved diagonal example has three zero calls; a six-relation interleaved identity has every call nonzero. These boundary examples do not save rank. Omitted-type and wrong-hub-target controls also fail the full endpoint check.

Normal/optimized replay, committed bytes and modes, source history, export and sealed Git-free replay pass. 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 suites, selected cold replay, native/physical replay, Lean, Docker, independent human and formal-kernel review were not run. The adapted checker is not an independent source review.

Base: current main e54523517d0f03e58352e647c403a5af3e525a57. Ali Romman's PR #7707, head 57e24923b609d2588b83321f530fa2a1e75ed364, supplies the six-member pool. Its geometry remains prior work. The preceding exclusion and open question are pinned at c396e9ee03b331f3b46aa87526817b237f5ad6f8.

Credit @CubeJerry for the augmentation chronology method in PR #6782, head 685457252b51b9c7932ece363365bdf0b12c9c31: word proof and original Laurent recurrence. Ali Romman's PR #7695, head 996c002d82420141c19797e4e3b699ed88aee330, supplies the prior paid-subset proof and adapted exact primitives. The linked credits retain its edge-family predecessor and the normal-triple geometric contributors. Prepared by Ali Romman with substantial OpenAI Codex assistance, Apache-2.0.

The scoped claim records the changed alphabet and stop rule. Live handoffs, source-reference intake, phrase overlap and the closed delta were checked. Sparse local and unpublished work remain coverage limits. Open next questions require a specified different pool or relation budget and a complete rank bill; no follow-up is reserved.

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