Repository navigation
Tooling: accelerate exact seven-owner Fano replay with row tables - #3219
Merged
Mister-Meeseeks merged 1 commit intoOct 17, 2026
Merged
Mister-Meeseeks merged 1 commit into
Mister-Meeseeks merged 1 commit into
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The PR3168 verifier spends most of its time multiplying the same finite GF(2) row patterns. This opt-in adapter precomputes their exact right actions, preserving all 1,379,952 reduced triples, full 289-bit equality keys, relation words, Fano searches and the complete original receipt. No new bound, supplier or κ is claimed.
Three alternating fresh-process pairs on the same complete input measured:
Raw timings and memory, exact source/blob pins, the row/packing equivalence argument and a registered replay are included in
research/fano-census-row-tables/.Validation: every actual triple key agrees exactly with the original multiplication; every complete benchmark receipt matches the committed predecessor bytes. Seven test groups pass under Python 3.12 (normal and optimized) and 3.13, including a separate dot-parity oracle, four semantic mutants, source mutation/restoration and output-reuse controls. Package metadata, selected-source inventory and whitespace checks pass.
Reproduce:
Expected:
PASS_ALL_EXACT_TRIPLE_KEYS 1379952 rows 18388, thenPASS_FANO_ROW_TABLESwith receipt hashe33a4809ad6ec2025aa46ecb5b2f9d790295379d4794bd911e0b67cf7e6aa652.Commit:
209d20d9da025c90110465f9a1f18186a6160147. Base:463113d675af7410ac91d53a4fed51ec6494eb9a; OS assignment seed 91, verification engineering. No pinned sprint issue existed; claim and overlap coverage are recorded locally. Frozen predecessors,upstream/, selected κ and default commands are unchanged. The ledger's partial-review status of #3097/#3168 is not promoted by this tooling result. Full supplier replay, repository-wide suites, Lean and hosted CI were not run; observation/representation/support and inherited multiplication interfaces remain assumptions.Credit Sudarsh Chaturvedi's #3168 census/Fano verifier, Gonzalo Mier's #3097 support-seven work, Ross Richardson's chart-swap derivation, Aaron Wang/navinor, ywz65 and the cumulative selected supplier lineage. Prepared by Charan with substantial OpenAI Codex assistance; no independent human review. Apache-2.0.