Feb. 8, 2024, 5:10 a.m. | Baihe Huang Hanlin Zhu Banghua Zhu Kannan Ramchandran Michael I. Jordan Jason D. Lee Jiantao Jiao

cs.CR updates on arXiv.org arxiv.org

We study statistical watermarking by formulating it as a hypothesis testing problem, a general framework which subsumes all previous statistical watermarking methods. Key to our formulation is a coupling of the output tokens and the rejection region, realized by pseudo-random generators in practice, that allows non-trivial trade-offs between the Type I error and Type II error. We characterize the Uniformly Most Powerful (UMP) watermark in the general hypothesis testing setting and the minimax Type II error in the …

cs.cl cs.cr cs.it cs.lg error framework general key math.it non practice problem random stat.ml study testing tokens trade trade-offs watermarking

More from arxiv.org / cs.CR updates on arXiv.org

Cloud Support Engineer

@ General Dynamics Information Technology | USA UT Roy - 5770 Missile Way, Roy, UT 84067 (UTC018)

Senior SIEM Developer (Cortex)

@ Palo Alto Networks | Tel Aviv-Yafo, Israel

Director, Product Management (Cloud Application Security)

@ Palo Alto Networks | Tel Aviv-Yafo, Israel

Cyber Security Specialist, Cyber Awareness Training & Strategic Projects

@ Grab | Petaling Jaya, Malaysia

Cyber Security Analyst (m/f/d)

@ Project A | Berlin

Cyber Security Analyst (m/w/d)

@ Project A | Berlin