Baselines for Identifying Watermarked Large Language Models

2023/05/29 by Leonard Tang, Tang, Leonard, Gavin Uberti +3
Computer Science · #Artificial Intelligence (cs.AI) #Computers and Society (cs.CY) #Cryptography and Security (cs.CR) #FOS: Computer and information sciences #Hate Speech and Cyberbullying Detection #Machine Learning (cs.LG) #Natural Language Processing Techniques #Topic Modeling

paper · pdf · doi:10.48550/arxiv.2305.18456

openalex publication_date 2023/05/29 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We consider the emerging problem of identifying the presence and use of watermarking schemes in widely used, publicly hosted, closed source large language models (LLMs). We introduce a suite of baseline algorithms for identifying watermarks in LLMs that rely on analyzing distributions of output tokens and logits generated by watermarked and unmarked LLMs. Notably, watermarked LLMs tend to produce distributions that diverge qualitatively and identifiably from standard models. Furthermore, we investigate the identifiability of watermarks at varying strengths and consider the tradeoffs of each of our identification mechanisms with respect to watermarking scenario. Along the way, we formalize the specific problem of identifying watermarks in LLMs, as well as LLM watermarks and watermark detection in general, providing a framework and foundations for studying them.

Related

• Three Bricks to Consolidate Watermarks for Large Language Models
• Yet Another Watermark for Large Language Models
• Mark Your LLM: Detecting the Misuse of Open-Source Large Language Models via Watermarking
• Learnable Linguistic Watermarks for Tracing Model Extraction Attacks on Large Language Models
• Mark My Words: Analyzing and Evaluating Language Model Watermarks