Skip to content

Prove the sharp joint dirty-word interval envelope (no new κ) - #6770

Merged
Mister-Meeseeks merged 2 commits into
CrocSwap:mainfrom
CubeJerry:research/20261106-sharp-dirty-intervals
Nov 10, 2026
Merged

Mister-Meeseeks merged 2 commits into
CrocSwap:mainfrom
CubeJerry:research/20261106-sharp-dirty-intervals

Conversation

@CubeJerry

Copy link
Copy Markdown
Contributor

Summary

Prove the sharp height-uniform interval envelope F(h)=floor(h^2/2)+h+2 for the unchanged #628 balanced word: every ordered pair of ideal-leaf cuts, both orientations and shared dirty scratch. The new upper proof matches #1374's credited lower family and closes #5872's gap. #1085's entry invariant and #1278's phase/squarefree results remain explicit premises.

Canonical base and source pin: 7c49fe3. Exact source-file hashes are in SOURCES.json.

Evidence

python3 -B -W error -X dev research/chronological-dirty-intervals/verify.py

Expected: PASS_CHRONOLOGICAL_DIRTY_INTERVALS. Normal/optimized registered and read-only Git-free replays passed: 95,735 prefix boundaries, 89,528 intervals, 2,118 phase identities and 15 controls. A separately written compressed scalar engine checks attaining witnesses through h=32. Separate proof/code review found no unresolved gap under the imported premises.

Limitations and next questions

No new κ. The bound is sharp across each height, not necessarily at every n; fixed-n sharpness remains open. Rounded callers, actual error support and the complete precision/physical invoice need separate integration. Five selected assumptions remain. Full-history validation was unavailable in the shallow checkout.

Attribution

@CubeJerry, with substantial OpenAI Codex assistance. Full proof, predecessor credit and exact review hashes: research package. Original notices retained; no independent human or proof-kernel acceptance claimed.

Closes the gap between the retained lower family and interval upper bound.
Includes exact finite corroboration and scoped independent AI review.
Prepared for @CubeJerry with OpenAI Codex assistance. No new kappa.
Retain exact reviewed scientific hashes and distinguish the specified-word theorem from caller admission. No scientific files changed.

Prepared for @CubeJerry 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