Repository navigation
Audit retained-helper independence in scatter parity - #3128
Merged
Mister-Meeseeks merged 1 commit intoOct 17, 2026
Merged
Conversation
Mister-Meeseeks
pushed a commit
that referenced
this pull request
Oct 17, 2026
… native COPY counts (no new kappa) Runs PR #3128's four full-size fixtures on the selected PR #589 bit word (selected baseline; balanced alias, which appends helper 480 as retained slot 9 and [8,8,9,9] to scatter row 64; unpaired even slot; distinct helpers) and a new unused-slot control through the unchanged symbolic checker, record exporter, simple-scatter-v1 guard and native legality and five-stage column checkers. Compiled for ten COPY/ERASE pairs, the balanced alias passes legality and all 883 formal columns; the selected nine-copy builds reject it at their count checks, and the guard rejects its repeated row before any output is written. Its complete local invoice keeps the extra rank-seven COPY/ERASE pair and two scatter ADDs: records 27,234 to 27,238, rank mass 5,129 to 5,136, five-stage ADDs 121,525 to 121,535 and fresh copies 45 to 50. The prefix maxima (49,347 forward, 1,113,070 reverse) and the 93-bit payload bound are unchanged. An unused tenth alias with unchanged simple rows passes both guard entry points and the ten-copy native checks, while the nine-copy builds reject it, so the guard enforces row support, not retained-list injectivity or the selected copy count. The unpaired and changed-helper controls keep symbolic acceptance and native legality but fail the formal columns at output 175, where dirty input 480 is a witness. Three native mutants separate pricing, lifecycle and endpoint checks. Finite audit evidence only: NCOPY=10 is an explicitly changed contract, so no ten-total supplier is admitted and no frontier or kappa changes. Cold replay, bank/projector/hosting admission, nine-stage allocation, all-size compilation, precision/tape and analytic transfer stay open. Adds research/scatter-alias-native-audit/ only. Submitted by @commdata2338 with substantial AI assistance (OpenAI Codex research and implementation; packaged with Claude Code). Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01X86sCSXdoDd61SDFfC1PKG
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Change and mathematical scope
State the result, changed hypotheses and status: construction, proof, bounded
search, audit or tooling. Distinguish a scalar saving from a paid-profile or κ
improvement. List remaining obligations.
Task and overlap
Give the base commit, exact dependency heads, concrete object and bounded domain.
For donated-agent work, include the task ID/scope key or link the task card from
autoresearch.md. Link the nearest overlapping PRs and explain the added value;
different random seeds alone are not separate contributions.
Declare whether scope is unchanged, refined or changed. If the card's premise
changed, explain why and how the actual result connects to the bottleneck.
Validation
Give reproduction commands, environment, exact evidence and meaningful failure
controls. Distinguish finite tests from proof arguments and list checks not run.
For a proposed bound, include the full paid budget and source bindings.
Credit
Credit predecessor and parallel work, retain source licenses, and disclose
substantial AI assistance. Useful independent or superseded work still merits credit.
Open next questions
For a research package, link its next questions and give the smallest useful
follow-up, why it matters and any blocker. “None justified” is acceptable.
Workflow feedback (optional)
What helped, what caused friction or an unintended detour, and one suggested
improvement? A link to the feedback discussion or an existing report is enough.