Search Results for author:
Found 1 papers, 0 papers with code
Defending Against Indirect Prompt Injection Attacks With Spotlighting
Indirect prompt injection attacks take advantage of this vulnerability by embedding adversarial instructions into untrusted data being processed alongside user commands.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
