Repository navigation
Conversation
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
This scoped proof excludes interleaving in a six-call word on the shared-hub projector pool when each projector is used once at a nonzero exponent. Over any torsion-free abelian exponent group, a diagonal endpoint requires two consecutive triple blocks, in either order, with labels
SandTconstant within the respective blocks. The endpoint isdiag([S],[S],[S+T],[T],[T]). Its targets use disjoint sets of paid labels, so total rational target rank is at most total paid rank.The first nonzero augmentation degree of each triple supplies a common nonzero initial form. Their product lets the cross-commutator lemma force the block order even at five relations, when one triple can vanish in degree one. This answers the six-nonzero-call question in the advance claim.
Credit aliromman for the pool in #7707 at
57e24923b609d2588b83321f530fa2a1e75ed364and the cross-commutator lemma in #7711 atac185e4529a01156f7b574555b50827f8cb645bb. Credit CubeJerry for the augmentation chronology method in #6782.Evidence
The proof is in
research/shared-hub-single-pass/README.md. The registered verifier multiplies all 527,040 full Laurent matrices across 720 orders and 732 signed label tuples at five relations. It found 792 diagonal endpoints and checked every diagonal entry against the formula. The direct command took 2.61 seconds locally. The registered replay and committed-byte, mode and source-history checks passed.Run from the repository root with CPython 3.12.15 and the standard library.
The direct check prints
PASS_SHARED_HUB_SINGLE_PASS cases=527040 diagonal=792.Scope and open obligations
The proof covers precisely six nonzero calls in the fixed original register coordinates. The finite replay uses an infinite cyclic exponent group. The written proof handles other torsion-free abelian groups. Repeated types, zero exponents and changed registers require separate arguments. Native realization, physical storage and restoration, and the complete recursive invoice remain open. Predecessor replays, the selected cold replay, Lean and Docker were unrun. Independent human and formal proof-kernel review are unclaimed.
Open next questions
Can a seven-call word with five relations interleave the triples with all calls nonzero and the repeated type separated by an opposite-triple call? Start with one repeated A type. Compare the sum of its initial forms in the private square with their separate contributions to the first mixed degree.
Related open work
#7707 supplies the six-projector pool and studies its geometry.
#7711 treats arbitrary repetitions at four relations. Its larger-budget control uses zero calls.
#7710 studies a separate fresh-bridge alphabet and explicitly leaves out this six-projector pool.
#7749 constructs a four-projector one-edge maximizer from the same family and leaves its paid-word question open.
CI notes
This PR carries the shared CI-doc and replay-gap sync used by other open PRs.
The six current-status tests and registered historical replay-gap package passed locally with CPython 3.12.15. All three checks passed on
9394615e9fd3a30b66e23d422da4945258d7923f. The hosted run covers planning and the historical replay-gap and shared-hub package replays.The base CI guide named
make frontier6030-verifywhile the selected record requiredmake frontier7000-verify. The same navigation assertion failed in the #7711 plan job. The shared sync updates the guide and runs the replay-gap verifier against its pinned historical snapshot.