Reviewed tier overrides for python.zh-cn's probable whole-file AI passes - #12
Merged
Conversation
There was a problem hiding this comment.
Pull request overview
Adds a reviewed per-commit tier override list for lecture-python.zh-cn and wires it into the QuantEcon python.zh-cn configuration so whole-file rewrite commits can be classified consistently for the first multi-edition report workflow.
Changes:
- Link the python.zh-cn scan config to a new per-commit tier overrides file.
- Add
configs/quantecon/overrides/python-zh-cn.ymlcontaining reviewed overrides for specific commits (tier + explanatory note). - Document the addition in
CHANGELOG.md.
Reviewed changes
Copilot reviewed 3 out of 3 changed files in this pull request and generated no comments.
| File | Description |
|---|---|
configs/quantecon/python-zh-cn.yml |
Points the python.zh-cn configuration at the new overrides file. |
configs/quantecon/overrides/python-zh-cn.yml |
Introduces reviewed per-commit tier overrides (with notes) for lecture-python.zh-cn. |
CHANGELOG.md |
Records the addition of the reviewed tier-override file as part of prep for Issue #1. |
💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.
… passes Prepares the python.zh-cn scan for the first three-edition report (#1): the editor-edit-corpus plan in project-translation segregated whole-file rewrites inside the 2025 hand-translation campaign as probable further AI passes, and without overrides they classify as roster tiers and inflate human churn. Candidates were derived by rule — a roster-tier commit replacing >=50% of a document's prose against the parent revision, renames followed — and each was adjudicated from its before/after pairs. Seven commits show uniform equivalent-meaning rephrasing with no glossary direction and are overridden to ai-assisted; one English source re-seed is overridden to seed; commits whose pairs carry a systematic glossary direction or targeted fixes (政策→策略, 求职搜索→工作搜寻, 蛋糕食用→吃蛋糕) are judged genuine editing and left alone. Effect at 6a4b427: corpus human share 11.3% -> 6.2%, ai-assisted 0.3% -> 4.6%, seed 0.9%; the pair count drops by exactly the re-seed commit's 1,966 pairs. The plan's derivation is described in QuantEcon/project-translation#22 and plans/editor-edit-corpus.md. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
mmcky
force-pushed
the
python-zh-cn-overrides
branch
from
August 27, 2026 23:51
1263356 to
ee3d43d
Compare
This was referenced Aug 28, 2026
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.
Prepares the python.zh-cn scan for the first three-edition report (#1, Phase 2 of #8). The editor-edit-corpus plan segregated whole-file rewrites inside the 2025 hand-translation campaign as probable further AI passes (see QuantEcon/project-translation#22 and plans/editor-edit-corpus.md); without overrides they classify as roster tiers and inflate human churn in F1 and the composition shares in S1.
Derivation and adjudication
Candidates were derived by rule — a roster-tier commit replacing ≥50% of a document's prose lines, measured against the parent revision with renames followed — over a scan of lecture-python.zh-cn @
6a4b427. Each candidate was then adjudicated from its before/after pairs, which split the set cleanly:ai-assisted(7): uniform equivalent-meaning rephrasing across everything touched, no glossary direction — the March 2025 rewrite wave in the days after the bulk translation (ab0498b5,0f3e2336,5990a84b), two whole-lecture single-commit fresh translations (778e3e11,f22cdff8), and two whole-file uploads with sentence-scale restructuring throughout (ef47483d,091ed6c1).seed(1):a6644817re-seeds three lectures to updated English source in a toc restructure — the removed Chinese is displaced by English source text, not edited.1822c976, 求职搜索→工作搜寻 in806803f8, 蛋糕食用→吃蛋糕 in the cake-eating uploads) — judged genuine human editing, which is why this list is shorter than the plan's instance count of 12.Effect
At
6a4b427the corpus human share moves 11.3% → 6.2%,ai-assisted0.3% → 4.6%,seed0.9%; the pair count drops by exactly the re-seed commit's 1,966 pairs, confirming only theseedreclassification removes pairs from the corpus.Every entry carries a note in the file per the overrides convention — each is a reviewable claim, and the #1 hand-adjudication pass can overturn any of them.
🤖 Generated with Claude Code