humanize-chinese

Category: Development & Coding | Uploader: voidborne-dvoidborne-d | Downloads: 0 | Version: v1.0(Latest)

Detect and humanize AI-generated Chinese text. 20+ rule detection categories plus 8 HC3-calibrated statistical features (sentence-length CV Cohen's d=1.22, short-sentence fraction d=1.21, comma density, GLTR rank buckets, DivEye surprisal). Unified CLI: ./humanize {detect,rewrite,academic,style,compare}. 7 style transforms (casual/zhihu/xiaohongshu/wechat/academic/literary/weibo), 40 paraphrase templates, 122 academic replacements, CiLin synonym expansion with semantic filter. Academic paper AIGC reduction for CNKI/VIP/Wanfang (知网/维普/万方 AIGC 检测降重), 11 academic detection dimensions. Pure Python, no dependencies. v3.0.0 — HC3 accuracy 73%, humanize avg delta +4.2 on HC3-Chinese 100-sample.

Changelog: Source: GitHub https://github.com/voidborne-d/humanize-chinese

Directory Structure

Current level: tree/main/

  • 📁 claude-code/
    • 📄 academic.md 2.8 KB
    • 📄 detect.md 2.0 KB
    • 📄 humanize.md 1.9 KB
    • 📄 README.md 1.9 KB
    • 📄 style.md 1.4 KB
  • 📁 evals/
    • 📁 oracles/
      • 📄 __init__.py 18 B
      • 📄 paperpass_adapter.py 4.6 KB
      • 📄 zhuque_adapter.py 3.3 KB
    • 📄 compression_oracle.py 6.8 KB
    • 📄 evals.json 1.6 KB
    • 📄 run_hc3_benchmark.py 15.0 KB
    • 📄 run_longform_benchmark.py 15.9 KB
    • 📄 v5_calibrate.py 11.3 KB
  • 📁 examples/
    • 📄 sample_academic.txt 854 B
    • 📄 sample_general.txt 1.3 KB
    • 📄 sample_long_blog.txt 4.4 KB
    • 📄 sample_social.txt 1.5 KB
    • 📄 sample_workplace.txt 549 B
  • 📁 scripts/
    • 📄 academic_cn.py 66.4 KB
    • 📄 cilin_synonyms.json 2.1 MB
    • 📄 compare_cn.py 5.6 KB
    • 📄 detect_cn.py 32.8 KB
    • 📄 humanize.py 3.1 KB
    • 📄 humanize_cn.py 101.2 KB
    • 📄 lr_coef_academic.json 2.4 KB
    • 📄 lr_coef_cn.json 2.3 KB
    • 📄 lr_coef_longform.json 2.7 KB
    • 📄 ngram_freq_cn.json 149.2 KB
    • 📄 ngram_freq_cn_human_holdout.json 20.5 KB
    • 📄 ngram_model.py 58.1 KB
    • 📄 patterns_cn.json 36.6 KB
    • 📄 restructure_cn.py 53.2 KB
    • 📄 style_cn.py 23.2 KB
    • 📄 train_lr_academic.py 7.8 KB
    • 📄 train_lr_longform.py 7.8 KB
    • 📄 train_lr_multisource.py 8.8 KB
    • 📄 train_lr_scorer.py 7.1 KB
    • 📄 train_ngram_human.py 3.4 KB
    • 📄 train_ngram_news.py 1.6 KB
    • 📄 train_ngram_wiki.py 2.0 KB
  • 📄 .gitignore 830 B
  • 📄 humanize 318 B
  • 📄 LICENSE 1.9 KB
  • 📄 package.json 1.2 KB
  • 📄 README.md 26.1 KB
  • 📄 SKILL.md 10.5 KB

SKILL.md

Login to download/like/favorite ❤ 235 | ★ 0
Comments 0

Please login before commenting.

Loading comments...