Exponential Sampling (EXP) is a state-of-the-art text watermarking method for large language models (LLMs), designed to embed hidden signals for authorship attribution and misuse detection. It achieves high detection accuracy under benign conditions with minimal impact on text quality.
Exponential Sampling (EXP) is a method to embed invisible watermarks in text generated by large AI models. It works well for identifying AI-generated content under normal circumstances but fails when the text is translated back and forth between languages. New research suggests combining multiple watermarking techniques can make it more robust.
EXP
Was this definition helpful?