Skip to content

Prove adjacent swaps of the #7041 block word miss mask (4) - #291

Draft
unpaidmoon wants to merge 1 commit into
cursor/upstream-e545235-base-67f2from
cursor/c7-boundary-swap-mask-four-67f2
Draft

unpaidmoon wants to merge 1 commit into
cursor/upstream-e545235-base-67f2from
cursor/c7-boundary-swap-mask-four-67f2

Conversation

@unpaidmoon

Copy link
Copy Markdown
Owner

Change and mathematical scope

Negative result, exhaustive finite census, plus a coordinate proof for one boundary. No κ change. The selected record remains κ = 363005817/100000000000.

On the length-42 block word 350122222224444444000000111111333333555555, no pair reads as the weight-1 mask (4). Every adjacent transposition misses it as well. The named case is the exchange of string indices 17 and 18, the last 4 and the following 0:

350122222224444440400000111111333333555555

An adjacent exchange moves only one prefix. Pairs that avoid that prefix keep the original sums, and the original 903 pairs miss mask (4). The 42 pairs that use the moved prefix are evaluated directly. Nine indices have unequal letters; the other 32 adjacent swaps are the original word.

For index 17 the moved prefix is q'_18 = (2, 1, 7, 1, 6, 1). A sum against that prefix reads as mask (4) only for the absent count vectors (5, 6, 0, 6, 2, 6) and (5, 6, 7, 6, 2, 6). Both target sums do read as mask (4), so the predicate is not what rejects them. Exactly two readings change, and both become non-readings.

This does not decide whether some other floor word realizes mask (4), and it does not decide the six-source depth-two question at characteristic 7.

Task and overlap

Base commit: e54523517d0f03e58352e647c403a5af3e525a57.

Session r280, item 3 of the ordered eligible list. After dropping delivered handoffs, open-PR artifacts, and the claimed CI-document repair, the remaining prose handoffs are the seven-call signed-frame request, the eight-source singleton request, and the six-source c = 7 request (discussion comment 19384981, updated by comment 19537594). This package is that third item.

Task card: connector-alphabet-change, scope changed. The alphabet stays the unsigned dirty connector. The piece is open question 1 of PR CrocSwap#7226, which names this boundary swap and does not evaluate it. The other eight block boundaries are the same one-prefix census.

Nearest work: PR CrocSwap#7041 records the word; PR CrocSwap#7226 inventories its other 61 masks and leaves this edit open; PR CrocSwap#6941 is the diagonal equality case. PR CrocSwap#7037, PR CrocSwap#7016, PR CrocSwap#7032 and PR CrocSwap#7045 are different claims. Overlap refresh at 2026-11-11T15:22:29Z covered 589 open PR bodies and 7,253 discussion records. No open body or local note contains the swapped literal.

Validation

python3 -B research/six-source-c7-boundary-swap-mask4/verify.py
python3 -O -B research/six-source-c7-boundary-swap-mask4/verify.py
python3 -B scripts/research_package.py --package research/six-source-c7-boundary-swap-mask4

Both direct commands and the registered replay print PASS_SIX_SOURCE_C7_BOUNDARY_SWAP_MASK4 on CPython 3.12. Metadata pins passed. The census is 42 words times 903 pairs.

Not run: make frontier7000-verify, any SAT search, non-adjacent transpositions, and the seventeen weight-3 masks on the index-10 swap.

Credit

The depth-two interface is Nick Petalas's (PR CrocSwap#4231, PR CrocSwap#4248, PR CrocSwap#6107), with the Ross Richardson PR CrocSwap#934 / PR CrocSwap#943 lineage. JosieElliston's PR CrocSwap#6141 leaves characteristic 7 open. The block word is PR CrocSwap#7041. The question is open question 1 of PR CrocSwap#7226.

Prepared by unpaidmoon with substantial Cursor Grok assistance. Apache-2.0.

Open next questions

  • Which of the 17 missed weight-3 masks appear after exchanging indices 10 and 11? First step: evaluate those masks on that one word. This package only tracks mask (4).
  • Does a non-adjacent transposition realize mask (4)? First step: one distance-2 swap of a 4 with a 0. The one-prefix reduction uses adjacency. The full domain has 861 transpositions.

Workflow feedback (optional)

The CrocSwap#7226 open question names one boundary and the letters on it, so the follow-up is a 42-pair check rather than another floor-word search. The same one-prefix identity covers the other eight block boundaries at no extra mathematical cost.

Open in Web Open in Cursor 

The index-17 boundary edit asked by PR CrocSwap#7226, and every other adjacent
transposition of that length-42 floor word, has no pair that reads as
the weight-1 mask (4). No kappa change.

Co-authored-by: unpaidmoon <unpaidmoon@users.noreply.github.com>
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.

2 participants