Mark Your LLM: Detecting the Misuse of Open-Source Large Language Models via Watermarking

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Xu, Yijie, Liu, Aiwei, Hu, Xuming, Wen, Lijie, Xiong, Hui
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909538298363904
author Xu, Yijie
Liu, Aiwei
Hu, Xuming
Wen, Lijie
Xiong, Hui
author_facet Xu, Yijie
Liu, Aiwei
Hu, Xuming
Wen, Lijie
Xiong, Hui
contents As open-source large language models (LLMs) like Llama3 become more capable, it is crucial to develop watermarking techniques to detect their potential misuse. Existing watermarking methods either add watermarks during LLM inference, which is unsuitable for open-source LLMs, or primarily target classification LLMs rather than recent generative LLMs. Adapting these watermarks to open-source LLMs for misuse detection remains an open challenge. This work defines two misuse scenarios for open-source LLMs: intellectual property (IP) violation and LLM Usage Violation. Then, we explore the application of inference-time watermark distillation and backdoor watermarking in these contexts. We propose comprehensive evaluation methods to assess the impact of various real-world further fine-tuning scenarios on watermarks and the effect of these watermarks on LLM performance. Our experiments reveal that backdoor watermarking could effectively detect IP Violation, while inference-time watermark distillation is applicable in both scenarios but less robust to further fine-tuning and has a more significant impact on LLM performance compared to backdoor watermarking. Exploring more advanced watermarking methods for open-source LLMs to detect their misuse should be an important future direction.
format Preprint
id arxiv_https___arxiv_org_abs_2503_04636
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Mark Your LLM: Detecting the Misuse of Open-Source Large Language Models via Watermarking
Xu, Yijie
Liu, Aiwei
Hu, Xuming
Wen, Lijie
Xiong, Hui
Computation and Language
Artificial Intelligence
Cryptography and Security
Machine Learning
As open-source large language models (LLMs) like Llama3 become more capable, it is crucial to develop watermarking techniques to detect their potential misuse. Existing watermarking methods either add watermarks during LLM inference, which is unsuitable for open-source LLMs, or primarily target classification LLMs rather than recent generative LLMs. Adapting these watermarks to open-source LLMs for misuse detection remains an open challenge. This work defines two misuse scenarios for open-source LLMs: intellectual property (IP) violation and LLM Usage Violation. Then, we explore the application of inference-time watermark distillation and backdoor watermarking in these contexts. We propose comprehensive evaluation methods to assess the impact of various real-world further fine-tuning scenarios on watermarks and the effect of these watermarks on LLM performance. Our experiments reveal that backdoor watermarking could effectively detect IP Violation, while inference-time watermark distillation is applicable in both scenarios but less robust to further fine-tuning and has a more significant impact on LLM performance compared to backdoor watermarking. Exploring more advanced watermarking methods for open-source LLMs to detect their misuse should be an important future direction.
title Mark Your LLM: Detecting the Misuse of Open-Source Large Language Models via Watermarking
topic Computation and Language
Artificial Intelligence
Cryptography and Security
Machine Learning
url https://arxiv.org/abs/2503.04636