Post by Taggart :ifin:

@laguiri Indirect prompt injection occurs when the LLM references an external source—a website or file, usually—that contains text intended to provide further instructions to the model. Everything th…

By Taggart :ifin:

View on We The Citizens