S
Susanoo
NEWS // IA & TECH
LIVE
RECHERCHEIL Y A 1SEM — 1 MIN DE LECTURE

LLMs respond differently to harmful prompts when AI watermarking is used

SynthID can cause models to follow harmful instructions they would otherwise refuse.

PAR SUSANOO NEWSSOURCE : ARS TECHNICA
SynthIDwatermarkingsafety-alignmentrefus-de-generationjailbreakGoogle-DeepMind
SynthID can cause models to follow harmful instructions they would otherwise refuse.
← RETOUR À L'ACCUEIL
LLMs respond differently to harmful prompts when AI watermarking is used — SUSANOO NEWS | SUSANOO NEWS