Hi HN,
I built a tool that rewrites AI-generated text to pass detection systems like Turnitin and CNKI (知网). Unlike simple paraphrasing tools, it uses semantic understanding to maintain meaning while removing AI fingerprints.
Technical approach:
– Analyzes sentence structure patterns typical of LLM outputs
– Rebuilds content using varied syntax and vocabulary distributions
– Preserves academic tone while reducing AI detection scores from 80%+ to <20%
Tested on: CNKI (知网), VIP (维普), WanFang (万方), and Western detection systems
Try it: https://www.yanbiai.com/tools/paper-pass?ly=hackernews
Would love feedback from the community, especially on:
1. Ethical considerations (I know this is controversial)
2. Technical improvements to the rewriting algorithm
3. Detection evasion vs. genuine writing assistance
Free tier available for testing.
Comments URL: https://news.ycombinator.com/item?id=45862583
Points: 1
# Comments: 0
Source: www.yanbiai.com
