fix: IVF GNO-IMI build fails recall threshold - #2744
Open
vsag-bot wants to merge 1 commit into
Open
Conversation
Signed-off-by: vsag-bot <276218163+vsag-bot@users.noreply.github.com> Assisted-by: Codex:gpt-5.6-sol
Collaborator
Author
|
/label status/waiting-for-review |
Collaborator
Author
|
AI PR state: Automatic merging is disabled. A maintainer performs the final merge. |
Contributor
Merge Protections🟢 All 3 merge protections satisfied — ready to merge. Show 3 satisfied protections🟢 Require kind label
🟢 Require version label
🟢 Require linked issue for feature/bug PRs
|
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
GNO-IMI now applies consistent cosine normalization during training, build/add routing, and search routing. A deterministic scale-invariance regression covers the failure mechanism.
Root Cause
GNO-IMI normalized cosine vectors during training and query routing but classified raw base vectors during build/add. This violated cosine scale invariance, assigned vectors to inconsistent buckets, and could lower recall below the functional-test threshold.
Changes
Validation
pre-fix scale-invariance regression: passed - Reproduced different bucket assignments before applying the production fix.cmake --build build-ai-fix-local --target unittests --parallel 4: passed - Affected library and unit tests built successfully.unittests "[GNOIMIPartition]": passed - All 3 test cases and 9 assertions passed.unittests "GNO-IMI cosine build routing is scale invariant": passed - All 512 routing assertions passed.functests "(PR) IVF GNO-IMI Build": passed - All 38,620 assertions passed.clang-format-15 --dry-run --Werror; git diff --check: passed - No formatting or whitespace errors.clang-tidy-15: not_run - clang-tidy 15 is unavailable in this environment.Residual Risks
Base branch:
mainFixes #2743
Prepared by
vsag-botafter automatic effort ordering and AI repairability preflight.