Y
Hacker News
new
|
ask
|
show
|
jobs
by
yorwba
113 days ago
Section 6, "Removing Watermarks," of the paper you cite makes it very clear that detecting LLM-generated text is not solved if the user takes measures to avoid detection.