Categories

what is prompt injection in large language models

Prompt injection

Prompt injection is an attack or failure mode where external content tries to manipulate the system's instructions. It happens when a large language model treats untrusted text, documents, webpages, emails ...

Prompt injection is an attack or failure mode where external content tries to manipulate the system's instructions. It happens when a large language model treats untrusted text, documents, webpages, emails Read article