Gone in 60 seconds: BEAST AI model attack needs just a minute of GPU time to breach LLM guardails Dátum: 2024-02-29 [theregister.co.uk]
Scientists reveal automated adversarial prompt generation too powered Nvidia visualization accelerator
Computer scientists from the University of Maryland have developed an efficient way to generate adversarial attack phrases that elicit harmful responses from large language models (LLMs). A TELJES HÍR ELOLVASÁSA >>
|