all InfoSec news
Topic-based Watermarks for LLM-Generated Text
April 3, 2024, 4:10 a.m. | Alexander Nemecek, Yuzhou Jiang, Erman Ayday
cs.CR updates on arXiv.org arxiv.org
Abstract: Recent advancements of large language models (LLMs) have resulted in indistinguishable text outputs comparable to human-generated text. Watermarking algorithms are potential tools that offer a way to differentiate between LLM- and human-generated text by embedding detectable signatures within LLM-generated output. However, current watermarking schemes lack robustness against known attacks against watermarking algorithms. In addition, they are impractical considering an LLM generates tens of thousands of text outputs per day and the watermarking algorithm needs to …
algorithms arxiv attacks cs.cl cs.cr cs.lg current generated human language language models large llm llms offer robustness signatures text tools topic watermarking watermarks
More from arxiv.org / cs.CR updates on arXiv.org
Jobs in InfoSec / Cybersecurity
EY- GDS- Cybersecurity- Staff
@ EY | Miguel Hidalgo, MX, 11520
Staff Security Operations Engineer
@ Workiva | Ames
Public Relations Senior Account Executive (B2B Tech/Cybersecurity/Enterprise)
@ Highwire Public Relations | Los Angeles, CA
Airbus Canada - Responsable Cyber sécurité produit / Product Cyber Security Responsible
@ Airbus | Mirabel
Investigations (OSINT) Manager
@ Logically | India
Security Engineer I, Offensive Security Penetration Testing
@ Amazon.com | US, NY, Virtual Location - New York