feat(templates): promote alias-expansion-sweep aliases into frontmatter - #24
Merged
Merged
Conversation
Promotes the 12 alias sets (182 aliases) from attune-rag's aliases_override.json — their designated temporary home — into the corresponding template frontmatter. Purely additive: existing aliases are preserved verbatim, promoted aliases appended in override-file order, no collisions introduced (221 distinct aliases corpus-wide, verified globally unique; every promoted alias clears the retriever's 2-stemmed-token minimum). Part of #9 (the attune-rag override retirement closes it).
Codecov Report✅ All modified and coverable lines are covered by tests. 📢 Thoughts on this report? Let us know! |
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Part of #9 — first of two PRs. The attune-rag override-retirement PR closes the issue.
What
Promotes the 12 alias sets (182 aliases) from attune-rag's
aliases_override.json— always a temporary home perdocs/specs/archive/alias-expansion-sweep/— into the corresponding template frontmatter. Purely additive: existing frontmatter aliases preserved verbatim, promoted aliases appended in override-file order.Validation
attune_rag.retrieval._tokenize(clearsMIN_ALIAS_OVERLAP=2).Benchmarks (disclosed, gates pass)
scripts/benchmark_all_fixtures.py, before → after: overall P@1 75.1% → 73.4% (gate ≥73% ✔), R@3 83.8% → 83.2%. Same delta with attune-rag 0.1.23 (help CI's pin) and repo HEAD 0.8.0.Why it moves at all: the retriever's
content_previewtokenizes the first 500 chars of the raw file — bigger frontmatter displaces body text from the window. The alias-credit mechanism itself is behavior-neutral (main+override scores identically to main on this fixture set). Drop concentrates in the heavily-aliased templates: code-quality 56→48, planning 56→48, refactor-plan 68→60.Measured follow-up (separate issue, attune-rag): stripping frontmatter from
contentbefore preview tokenization recovers to 74.8% on the promoted corpus. Filed as a fast-follow rather than expanding this PR.Paraphrased coverage the sweep bought is preserved end-to-end: with this branch installed and the override stubbed, attune-rag's golden thresholds pass and paraphrased R@3 is 100% (diagnostic-1).
After merge
attune-helppin, retire the sweep entries fromaliases_override.json(mechanism stays), regenerate bundled-corpus goldens — closes Promote attune-rag alias-expansion-sweep aliases into upstream frontmatter #9.