Skip to content

Construct complete no-edge pools without irreducible image bases (no κ) - #7688

Open
aliromman wants to merge 1 commit into
CrocSwap:mainfrom
aliromman:codex/consecutive-result-029
Open

aliromman wants to merge 1 commit into
CrocSwap:mainfrom
aliromman:codex/consecutive-result-029

Conversation

@aliromman

@aliromman aliromman commented Nov 18, 2026 •

Copy link
Copy Markdown
Contributor

Result and scope

A complete no-edge projector pool need not contain an irreducible image basis. For every n>=3, this construction has exactly n+1 binary rank-one idempotents, joint-image dimension n, irreducible full action, pattern dimension n+1 and minimum nonzero pattern weight six. Both of its proper full-image subcollections have n independent images and are reducible. The excess of one projector is the least possible for this property.

Read the proof, exact receipt, verifier and scope record. A cycle of row projectors plus one extra projector supplies the algebraic obstruction. A three-copy coordinate embedding preserves the exact complete pool of any binary matrix subspace and removes matrix units from its pattern span. The family uses width 3n and permits zero padding.

This rules out an unrestricted basis-extraction step for extending PR #7686. At image dimension seven, the new pattern dimension is eight. It does not settle the pattern-dimension-five question. That special case still needs an argument that uses its smaller pattern space.

No new kappa, paid BIT word, native operation or complete invoice is supplied. 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. For 3<=n<=12, the verifier checks all 16358 nonzero patterns and all 16368 projector subsets. Each complete pool is exactly the claimed family. Exactly three subsets per case have full image, including the full pool. The two proper subsets have explicit common-kernel or invariant-line witnesses. The latter has full covector rank, so a rank-only extraction test would miss the defect. Direct nonzero-vector orbit checks also cover 3<=n<=8.

The exact embedding lemma is checked on all 67 subspaces of M_2(F2), including coordinate and nilpotent cases, for 96 rank-one idempotent pool memberships. All 256 ordered matrix pairs satisfy the algebra law. Two-copy replication is a negative control: it sends an idempotent to a nilpotent. One-copy controls expose diagonal information loss and a matrix unit. The finite tests support the written all-dimension proof.

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. The proof and README are checked with GitHub Markdown rendering.

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

python3.12 -I -W error -B research/no-edge-basis-extraction-failure/verify.py
python3.12 -I -W error -B -O research/no-edge-basis-extraction-failure/verify.py
python3.12 -B scripts/research_package.py --package research/no-edge-basis-extraction-failure
python3.12 -B scripts/check_research_commit.py --package research/no-edge-basis-extraction-failure --check-source-history --check-modes

Expected markers:

PASS_NO_EDGE_BASIS_EXTRACTION_FAILURE
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. Predecessor scientific replays, the selected cold replay, Lean and Docker were not run. Independent human and formal-kernel review are unclaimed.

Sources and credit

Prepared by @aliromman with substantial OpenAI Codex assistance. Ali Romman's PR #7686, exact proof, supplies the scope question; its exact verifier supplies the reused binary matrix, span, orbit and lift primitives. @unpaidmoon supplied the earlier cyclic complete-pool example in PR #6962, exact proof. That example is itself an irreducible basis and is related context. Nick Petalas (@petalas) supplied earlier fiber work credited in the source record. No earlier cyclic example is claimed as new. New work is Apache-2.0; no predecessor package is needed at runtime.

Base: canonical main e54523517d0f03e58352e647c403a5af3e525a57, priority revision 8. Card bit-new-register-model, scope changed to this scalar proof-step counterexample. The original native construction gate remains unmet. Refreshed open/closed PR and Discussion overlap and live handoffs identify no conflicting claim for this extraction boundary; sparse local and unpublished work remain coverage limits. The next useful question is the restricted pattern-dimension-five case, not another instance of this family.

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