Oct. 16, 2023, 1:10 a.m. | Ryoma Sato, Yuki Takezawa, Han Bao, Kenta Niwa, Makoto Yamada

cs.CR updates on arXiv.org arxiv.org

We propose Easymark, a family of embarrassingly simple yet effective
watermarks. Text watermarking is becoming increasingly important with the
advent of Large Language Models (LLM). LLMs can generate texts that cannot be
distinguished from human-written texts. This is a serious problem for the
credibility of the text. Easymark is a simple yet effective solution to this
problem. Easymark can inject a watermark without changing the meaning of the
text at all while a validator can detect if a text was …

family human important language language models large llm llms problem serious simple solution text texts watermarking watermarks written

SOC 2 Manager, Audit and Certification

@ Deloitte | US and CA Multiple Locations

Open-Source Intelligence (OSINT) Policy Analyst (TS/SCI)

@ WWC Global | Reston, Virginia, United States

Security Architect (DevSecOps)

@ EUROPEAN DYNAMICS | Brussels, Brussels, Belgium

Infrastructure Security Architect

@ Ørsted | Kuala Lumpur, MY

Contract Penetration Tester

@ Evolve Security | United States - Remote

Senior Penetration Tester

@ DigitalOcean | Canada