Repository navigation
Reproduce v2 post-hoc candidates and verify one BM25 repair - #14
Merged
Merged
Conversation
eluan216
changed the title
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
All 46 original v2 attempts and both empty prediction exports are preserved. Independent checksum-verified reconstruction and final-step-only revalidation exactly reproduced the user summary: BM25 2 applicable, 7 abstentions, 14 errors; graph 0 applicable, 6 abstentions, 17 errors.
Controlled SWE-bench 4.1.0 grading verified the pydicom-1694 BM25 candidate: one bug test and all 26 labeled regression tests passed. Its reference fix passed; its unchanged-program control failed the bug test while retaining all regressions. All three used the documented pytest 7.4.4 compatibility pin. The SQLFluff-2419 candidate failed its bug test, with qualified positive/negative controls under the unchanged official environment; this task has no labeled regression tests.
This is one test-verified BM25 repair under the post-hoc normalization protocol, not an original v2 success, a complete official repair score or evidence of graph superiority. No extra model generation was used.
Validation: all 51 CPU tests, bootstrap, retrieval study, input/holdout preparation, independent revalidation and both controlled grading jobs passed. Full run: https://github.com/eluan216/LocalCodeGraph/actions/runs/42483558413
The evaluation report, raw attempts, parsed controls, README and manuscript now record both positive and negative results.