cs.LGNov 11, 2026

SAME: Semantic Watermark Embedding for Open-Source Large Pretrained Models

Authors: Xinyao Song, Jialiang Shen, Hui Kang, Huajie Chen, Ruiqi Liu, Jiyang Zheng, Huaxi Huang, Zhenliang Lu, 

Organizations: East China Normal University · The University of Sydney · CXMT · City University of Macau · 

Abstract

Open-source large pretrained models (LPMs) have become increasingly valuable assets for model development and downstream applications, but their accessibility also facilitates unauthorised copying and modification. To protect their intellectual property (IP), existing approaches typically embed tailored trigger-response patterns as watermarks for ownership verification. However, such trigger designs need to be tailored to task-specific data that constrain their transferability across downstream tasks, and are not durable under subsequent benign fine-tuning. To address these limitations, we propose SemAntic waterMark Embedding (SAME), which introduces a semantically meaningful private vocabulary that enables flexible composition of adaptive and persistent watermarks. Specifically, SAME introduces a fixed private vocabulary that assigns each native token a private embedding counterpart, while aligning the semantic expressiveness of the two vocabularies within the model's representation space. This alignment allows private tokens to be flexibly composed into semantic triggers that elicit meaningful responses, enabling adaptive watermarks beyond predefined trigger-response pairs. By grounding the private vocabulary in the model's native semantics, SAME couples the watermark with its evolving semantic capabilities, allowing this connection to be preserved throughout downstream fine-tuning. Extensive experiments demonstrate that SAME provides reliable and robust watermarks across LPMs of varying scales and architectures, enabling effective ownership verification for IP protection.

Figures & tables

Appendix figures & tables6 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. OpenStamp: A Watermark for Open-Source Language Models

    Aug 28, 2026Miroojin Bakshi, Saksham Rastogi, Danish PruthiText WatermarkingLLM Watermarking

  2. Robust LLM Watermarking with Minimal Semantic Distortion for IP Protection

    May 22, 2026Kieu Dang, Phung Lai, NhatHai Phan +2Text WatermarkingLanguage Model Fingerprinting

  3. Functional Subspace Watermarking for Large Language Models

    Mar 19, 2026Zikang Ding, Junhao Li, Suling Wu +3LLM WatermarkingRobust Watermarking