Undetectable Watermarks for Language Models

Abstract Recent advances in the capabilities of large language models such as GPT-4 have spurred increasing concern about our ability to detect AI-generated text. Prior works have suggested methods of embedding watermarks in model outputs, by $\textit{noticeably}$ altering the output distribution. We ask: Is it possible to introduce a watermark without incurring $\textit{any detectable}$ change to

[wktk_msum]

https://twitter.com/__kokumoto/status/1665500508784656385 生成AIによる出力に電子透かし加えて出典わかりやすくする暗号技術、らしい