Hacker News Reader

Full thread

qurren

Up next:

1. Generate a bunch of responses with both Claude and various non-Claude LLMs (ChatGPT, Gemini, Kimi)

2. Train a discriminator model that can differentiate Claude vs. non-Claude

3. Train a de-watermarking model using the discriminator model as loss

View on HN

Or just write a 6 line program to remove the metainfo from the file?

very difficult to remove text watermarking

Outputs can be easily translated from the watermarked language to another, either by just knowing two+ languages, or via non-LLM translation.

There must be some set of outputs that is completely resistant to watermarking. For example if I told Claude to output "Hello, World"

Your observation is sharp, but it's not just metainfo—it's load-bearing text. To remove it, you need to delve deep and alter the tapestry of carefully selected words.

edit: almost forgot the "—"

s/text/prose/ s/'/’/

The honest seam is obvious: claudish works; imitation does not. That’s not nothing.

You just described how Pangram works.

Reply on news.ycombinator.com