Skip to content

Prove exact shared-hub two-pentagon query dimension nine (no κ) - #7531

Open
petalas wants to merge 1 commit into
CrocSwap:mainfrom
petalas:research/w5-1113c-frame-transport
Open

petalas wants to merge 1 commit into
CrocSwap:mainfrom
petalas:research/w5-1113c-frame-transport

Conversation

@petalas

@petalas petalas commented Nov 13, 2026 •

Copy link
Copy Markdown
Contributor

Summary

Scoped written proof with exact finite conformance, produced by Nick Petalas. Two pentagons sharing one hub have exact fitting rank nine when every vertex carries two field coordinates, over every field. All permitted directed 2-by-2 edge blocks are unrestricted. The independent-set bound gives eight, and the new saturation argument excludes it. Nine coordinate pairs attain the minimum.

independent vertices {1,3,5,7} occupy eight query coordinates
    -> terminal subspaces saturate both arms
    -> the hub decoder vanishes if rank is eight
    -> contradiction to its identity diagonal
nine disjoint coordinate pairs attain rank nine

This original changed-mask decoder experiment stops at that first sharp result. It does not change the selected #7000 frontier. The six new files are isolated in research/shared-hub-pentagon-query/.

Credit JoyboyBrian #5023 and tekuuu #3362 for the odd-cycle block-query interface and the retained P112/native lineage. Berliner and Langberg's classical outerplanar theorem covers scalar coding; this contribution supplies the two-coordinate arbitrary-block result and direct proof. The older Clebsch puncturing package is related background, with a different mask and fixed matrix. No predecessor executable or scientific result is audited or replayed, and no global literature-priority claim is made.

Evidence

The direct check prints PASS_SHARED_HUB_PENTAGON_QUERY and minimum dimension nine in its generated, file-hash-bound JSON result. Normal and optimized checks, registered metadata/pins and both registered replays pass locally with CPython 3.12.15.

  • 376 complete factorizations over F2, F3, F5, F7, F101 and Q, checking all 121,824 ordered entries and all factor/matrix ranks. Asymmetric and noncommuting fixtures are included.
  • All 262,144 binary inputs, with 4,718,592 coordinate recoveries.
  • All 187,200 normalized local branch instances over F2 and F3 used in the written proof. These are not an exhaustive census of arbitrary fitting matrices.
  • Twelve illegal/omitted-query mutants reject. A rank-eight code becomes valid after adding only edge 0-2, and the original mask rejects it.

From the repository root:

/Users/nick/git/integer-mult-bounds/.venv/bin/python3.12 -W error -B research/shared-hub-pentagon-query/verify.py
/Users/nick/git/integer-mult-bounds/.venv/bin/python3.12 -W error -O -B research/shared-hub-pentagon-query/verify.py
/Users/nick/git/integer-mult-bounds/.venv/bin/python3.12 -B scripts/research_package.py --package research/shared-hub-pentagon-query
/Users/nick/git/integer-mult-bounds/.venv/bin/python3.12 -B scripts/check_research_commit.py --package research/shared-hub-pentagon-query --check-source-history

Another CPython 3.12 executable may replace that absolute path. The two registered commands were timed locally at 1.79 s and 1.54 s, below their 120-second limits. These times document replay feasibility, not a speedup. Registered replay returns PASS_REGISTERED_RESEARCH_PACKAGE; committed-byte/source-history preflight returns PASS_RESEARCH_COMMIT_BYTES for all six files and ancestor base e54523517d0f03e58352e647c403a5af3e525a57. Syntax inspection and git diff --check pass.

The all-field lower bound is the written proof, not an inference from the finite corpus. Independent human and formal proof-kernel review are unclaimed.

Limitations and unrun obligations

Only this graph, two-coordinate payload and linear fitting contract are settled. Larger payloads, nonlinear coding and other topologies remain unproved. No physical source reads, side-information acquisition, native maps, dirty restoration, frames, coefficient preparation, ownership, hosting, precision/tape/stock, uniform compiler, full recursive invoice or supplier admission is supplied. No new kappa or complete multiplication theorem is claimed. Selected cold replay, Lean and Docker were not run.

Open next questions

  • Determine the exact rank with three coordinates per vertex, starting with the possible rank-thirteen case. The saturation argument and ordinary edge pairing leave that case unresolved.
  • Bind the nine-query code to an actual chronological caller, then charge every source read, side-information read, query, restoration and delivery. No physical profitability follows from dimension alone.

Related open work

No open PR found in the publication overlap check studies this two-pentagon shared-hub mask or its unrestricted two-coordinate fitting rank. Merged #5023 supplies the single-cycle background; the older Clebsch puncturing package studies a different mask. Open #7198 studies persistent stock in a chronological common-query stream, and #7269 studies denominator-constrained helper capacity; neither computes the fitting rank of this mask. Their results were not replayed or numerically compared.

  • #7533 reduces a separate-term A7+A1 COPY family to coding on a different 44-vertex graph. It does not compute this two-pentagon graph's fitting rank; no numerical comparison is asserted.

The pre-DONE refresh read this account's seven other open PRs created since the advance claim: #7527, #7528, #7530 and #7532 through #7535. The remaining subjects are suffix precision, a Euclidean resolvent, permutation tensor layouts, a source-count floor, orthogonal Schur residuals and capped tensor splitting. None computes the shared-hub two-coordinate fitting rank. No extra computations were run to check another contribution.

CI notes

At unchanged head a965bcb5b88e2e3a9856e1cc53639b43a896b663, Registered research packages run 43467848165 finished with plan failed and replay skipped. CI did not certify this package. The local normal/optimized verifier, registered metadata/replay and committed-byte/source-history checks passed as recorded above.

The plan job 132031654829 failed in Onboarding helper tests on exactly one test: test_research_current_status.CurrentCiNavigationTests.test_current_ci_section_matches_selected_record. Its log reports Ran 760 tests and FAILED (failures=1, skipped=24). The assertion says that `make frontier7000-verify` was not found in the CI verification document.

I read the job log and the corresponding files at main base e54523517d0f03e58352e647c403a5af3e525a57. certificates/selected-result.json names make frontier7000-verify, while the current section of docs/ci-verification.md still describes #6030 and make frontier6030-verify. Those files and the failing test are unchanged by this PR. This is a repository-level documentation mismatch. The package replay job was skipped after the plan failure.

The same single test, missing-command assertion, 760-test count and failures=1, skipped=24 result were independently confirmed from the failed-job logs in rohanarun's run 42905094314, dillona's run 42903313790, and petalas's #7527 run 43467013583, #7528 run 43467095967 and #7530 run 43467444862. No fix, retrigger commit or job rerun was made for this shared failure. Only the local checks certify the automated package conformance reported here; the written theorem remains without independent human or formal proof-kernel review.

Advance CLAIM: #1166 (comment). Delivered mathematical scope matches it.

Merge Danger

Door: two-way.

Blast Radius: isolated.

Only one six-file research package is added. Shared code, frozen evidence and selected-frontier files are unchanged.

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