Skip to content

Conditional κ = .003648411294790 (+0.0121% over #7544; +0.0239% over #7542): fuse hosted role902 moves - #7546

Open
ToikY wants to merge 4 commits into
CrocSwap:mainfrom
ToikY:codex/p104-role902-direct-lift
Open

ToikY wants to merge 4 commits into
CrocSwap:mainfrom
ToikY:codex/p104-role902-direct-lift

Conversation

@ToikY

@ToikY ToikY commented Nov 13, 2026 •

Copy link
Copy Markdown
Contributor

Result

Conditional κ 0.003648411294790 under the inherited all-width native action, owned erasure, compiler, precision/tape and recovery premises with the exact signed #7113 caller at 07a76c5e2738882c0cb322278e56f6cae3c9885d.

Comparison Exact head κ Absolute gain Relative gain
#7544, strongest independently reproduced comparator d6be7d620c9c8337104daeb09fe01196eced5fe3 0.003647969120276 0.000000442174514 0.0121211145%
#7542, latest relevant peer claim at 22:19 UTC 71f5acb0409daf208a865482f3349049dbd3251c 0.0036475398277 0.00000087146709 0.0238919143%

#7542’s native and declared price replay passed. The separate complete signed and outer application is beyond that PR’s verification scope. #7544's complete native and signed replay passed locally and in hosted science.

Construction and scope

Fuse hosted Y2 role902 MOVEs 40950/40951 (618→789→913) into a direct rank-four lift, with first retained read ADD40952 into 985. #7542 removed the last intervening observation. Exact frames telescope with absolute phase and adjoint preserved. The signed invoice loses 34 rank-two and adds 17 rank-four positions across 17 calls. Paid total falls to 110,468, productive total is 22,461, ordinary work falls by 17,000, and the elementary tape upper bound falls by 2,898,976. The directed threshold is bounded by [0.00364841129479084532481164992024, 0.00364841129479084532481164992025].

Same-pool screen and controls. The original order by gross saving was the Type B square, role902, then helper1028. Only role902 survived exact overlap, first-read support and repair screening. Task scope is refined to this one literal pair on top of canonical main e545235. The first two commits are the existing #7257 onboarding prerequisite, separately scoped and credited.

Replay

With CPython 3.12 from a clean checkout:

python3.12 -B scripts/research_package.py --package research/p104-role902-direct-lift

Independent local physical replay checked all 512 signed forward and adjoint columns, including a rejected absolute-phase mutation. Independent grid and invoice replay covered all 1,569 endpoint rows and every owned CLEAR, COPY, lifetime and MOVE. Complete local native and signed replay and an independent directed recurrence check passed, including 3,040 catalogue maps, 104 boundary pairs, ordinary, precision and tape costs, and all 58 outer predicates. The clean registered local replay passed in 461.42 seconds. The first hosted run passed independent and native checks, then hit the combined command’s cap. The package now separates native and signed command budgets and checks the same scientific receipt. On exact head 99438c43df7aac1e5dbcc43c1578fcd886c119ec, hosted onboarding, the historical prerequisite and the complete scientific replay passed. The hosted scientific job certified κ 0.003648411294790.

Credit and next test

Credit DreamingOfClouds for #7542’s chart789 contact removal, Clear1nsight for #7481’s funded producer and #7257’s onboarding, the #7544 literal composition and complete replay, Rohan Arun for the signed #7113 caller, #7480 and #7334 for the native lineage, #7003 for chronological fusion and #7298 for directed pricing. The phase cross-check reuses the #7542 chart engine and the compiler and phase interpreter also pinned in #7526. Source notices are retained. OpenAI Codex materially assisted construction and verification.

The fixed chart912 role496 repairs were ruled out independently. The canonical discussion links the current follow-up work. The replay package retains the exact relative ratios.

Clear1nsight and others added 3 commits November 11, 2026 15:39
This is the exact documentation prerequisite retained by CrocSwap#7156. It points the onboarding check to the selected CrocSwap#7000/CrocSwap#7001 review. Credit to Rohan Arun and the source authors.
The PR CrocSwap#6292 rejection census and package plan are statements about 2485805. Load its exact sources and planner in a private sparse checkout, keeping all source hashes, 19 expected rejections, nine package selections and the empty-plan control. Subsequent edits to the live Makefile and CI guide no longer invalidate this historical check.

The original audit is by unpaidmoon, building on Dennis’s historical replay gate. The pinned repository view follows the precedent of Matarener’s PR CrocSwap#5434. Repair prepared with substantial OpenAI Codex assistance.
@ToikY ToikY changed the title Conditional κ = 0.003648411294790 (+11054362850/911992280069% over #7544): fuse hosted role902 moves Conditional κ = .003648411294790 (+0.0121% over #7544; +0.0239% over #7542): fuse hosted role902 moves Nov 13, 2026
DreamingOfClouds added a commit to DreamingOfClouds/integer-mult-bounds that referenced this pull request Nov 14, 2026
Research package research/p104-odd-unit-per-grading: one cost (the odd
unit, paid per grading) behind the decoder wall and the t = -1 park.
Part I lemmas with exact stdlib checks; Part II scoped negative results
by route; Part III unpaid words replayed exactly against CrocSwap#7546.
Corrections: the one-rank law is the Fredholm alternative; HALF18 rank.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.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