Watermark for LLM-Generated Text

Researchers at Google have developed a watermark for LLM-generated text. The basics are pretty obvious: the LLM chooses between tokens partly based on a cryptographic key, and someone with knowledge of the key can detect those choices. What makes this hard is (1) how much text is required for the watermark to work, and (2) how robust the watermark is to post-generation editing. Google’s version looks pretty good: it’s detectable in text as small as 200 tokens.

Tags: , , , , ,

Posted on October 25, 2024 at 9:56 AM • 5 Comments

<

p id=”powered”>Sidebar photo of Bruce Schneier by Joe MacInnis.

Discover more from Cyber Elias Academy

Subscribe to get the latest posts sent to your email.

Leave a Reply

Explore More

Navigating Compliance: A Guide to the U.S. Government Configuration Baseline

For cybersecurity professionals tasked with defending the public sector, tackling the U.S. Government Configuration Baseline (USGCB) is just another hurdle to a safer federal tomorrow. Part of a wide collection

Read More

North Texas-Based StrikeReady Raises $12M for AI Security Command Platform

StrikeReady Founder and CEO Yasir Khalid [File Photo] Richardson-based security solutions provider StrikeReady has received $12 million in Series A funding. The funding round was led by 33N Ventures with

Read More

Hurricane Season Scams: What you need to know

Cybercriminals are notorious for their opportunism. No situation is off limits: whether they exploit conflict and human suffering, blackmail vulnerable individuals by threatening to leak therapy notes, or even bring

Read More