Skip to content

Bound parity-sketch erasure of two stage-17 body lanes - #7443

Open
ToikY wants to merge 3 commits into
CrocSwap:mainfrom
ToikY:codex/p104-stage17-parity-sketch-erasure-bound-20261112
Open

ToikY wants to merge 3 commits into
CrocSwap:mainfrom
ToikY:codex/p104-stage17-parity-sketch-erasure-bound-20261112

Conversation

@ToikY

@ToikY ToikY commented Nov 12, 2026

Copy link
Copy Markdown
Contributor

On the two exact #7113 stage-17 Type B body lanes at clocks 4646 and 4660, allow any 2×2 erased matrix depending on up to five linear Walsh parities. At width two the polarized quadratic form has rank 16 and every sketch cell retains rank at least 6. The target needs at least 7/8 of one old lane’s displacement capacity. One normalized rank-16 core supplies at most 3/4 per charged volume, requiring at least 7/6 old lane volumes. The directed tariff funds less than 1.003. The optimistic gross upper is below −20.798 raw. NO VERIFIED GAIN.

From a clean POSIX checkout with CPython 3.12, run python3.12 -I -B research/p104-stage17-parity-sketch-erasure-bound/replay.py. Two exact methods check the rational bounds and one extremal five-parity control over all 65,536 addresses. The registered replay, exact commit and mode check, Git-free read-only export replay, and all 27 onboarding tests passed locally. Address-mixing erasure, larger sketches, and a different core remain open.

#7334/#7113 supply the complete comparator and literal caller, #7416 the two-lane stage-17 setting, #7406 the projector and #7397 the volume tariff. OpenAI Codex and its subagents materially assisted the proof, implementation and independent replay.

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.
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