2025 Other https://fordatis.fraunhofer.de/handle/fordatis/469
http://dx.doi.org/10.24406/fordatis/426

Supplementary Data for "Security and Detectability Analysis of Unicode Text Watermarking Methods Against Large Language Models"

Supplementary Data of the Conducted Experiments in the Conference Paper

Hellmeier, Malte (Fraunhofer-Institut für Software- und Systemtechnik ISST)

ISST Fraunhofer-Institut für Software- und Systemtechnik
Files in This Item:
FileDescriptionSizeFormat 
cover-text.txtUsed Lorem Ipsum cover text1,3 kBTextDownload/Open
prompts.txtUsed prompts for experiments 1, 2, and 31,71 kBTextDownload/Open
results_experiments.xlsxResulting model outputs of experiments 1, 2, and 318,78 kBMicrosoft Excel XMLDownload/Open

Abstract

This dataset includes Large Language Models (LLM) prompts and the cover text, used for the three experiments conducted in the conference paper "Security and Detectability Analysis of Unicode Text Watermarking Methods Against Large Language Models". It also includes the results of all three experiments.
Classification
005 Computerprogrammierung, Programme, Daten
Keywords
Large Language Models
Digital Watermarking
Experimental Data
Funder
Fraunhofer-Gesellschaft FhG
Fraunhofer Group
IUK-Technologie
Show full item record


Items in Fordatis are protected by copyright, with all rights reserved, unless otherwise indicated.