1 paper · 1 filter
Shingo Kodama, Haya Diwan, Lucas Rosenblatt +2
The rapid spread of text generated by large language models (LLMs) makes it increasingly difficult to distinguish authentic human writing from machine output. Watermarking offers a…