I Know You Did Not Write That! A Sampling Based Watermarking Method for Identifying Machine Generated Text

Kaan Efe Keleş, Ömer Kaan Gürbüz, Mucahid Kutlu

Primary Area: generative models

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Keywords: large language models, detecting machine generated text, watermarking text

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.

TL;DR: embedding a watermark into machine generated text to reliably detect it afterwards

Abstract: Potential harms of Large Language Models such as mass misinformation and plagiarism can be partially mitigated if there exists a reliable way to detect machine generated text. In this paper, we propose a new watermarking method to detect machine-generated texts. Our method embeds a unique pattern within the generated text, ensuring that while the content remains coherent and natural to human readers, it carries distinct markers that can be identified algorithmically. Specifically, we intervene with the token sampling process in a way which enables us to trace back our token choices during the detection phase. We show how watermarking affects textual quality and compare our proposed method with a state-of-the-art watermarking method in terms of robustness and detectability. Through extensive experiments, we demonstrate the effectiveness of our watermarking scheme in distinguishing between watermarked and non-watermarked text, achieving high detection rates while maintaining textual quality.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 7265