I Cleaned 455 Articles, 474 Numbers Stayed in the English Twin

说明与用途

自动化扫描清理了455篇土耳其语文章，删除了3,222行伪造内容，但474个百分比仍保留在英文双语版本中。

问题

Automated content pipeline accumulated fabricated figures, invented scenes, and fake command output in a bilingual blog, requiring a sweep to de‑fab the content.

结果

The sweep cleaned 455 Turkish articles, deleted 3,222 Turkish lines, and identified 474 percentages that persisted in English twins.

工作流

• Run automated sweep script (scripts/measure-defab-coverage.mjs) to identify fabricated content.
• Commit pilot with rule definitions.
• Execute four batch commits to delete fabricated lines.
• Run second wave commit to finalize cleanup.
• Measure coverage and report 455 Turkish files cleaned, 474 numbers remaining in English twins.

证据

Six commits: a pilot (09:53), four batches (10:19–10:55) and a second wave (13:42).

来源

The sweep touched 455 Turkish article files. The number of English files it touched: zero.

来源

474 numbers, across 213 articles.

来源

The pilot commit's message names four classes: fake numbers, fake command output, piles of anecdotes, repeated credits.

来源