Universal and Context-Independent Triggers for Precise Control of LLM Outputs
Introduces a novel gradient-based prompt-injection technique that can generate universal triggers to manipulate open-source Large Language Model (LLM) outputs.
Summary
No summary yet — watch the video above, or open it on YouTube.
Related talks
CyberRisk TV·16 min
Ramin Farassat
Ramin Farassat investigates How Menlo Security Is Securing AI Agents from Prompt Injection.
aiai-agentsinjection+1
App Sec Village·27 min
Javan Rasokat · Andra Lezza
Practical insights and defensive strategies from real-world development and hardening of production LLM applications.
llmaiprompt-injection+2
Main Stage·32 min
Alexandre Grattafiori · Ivan Evtimov · Ron Bitton
Methodologies, automated multi-turn adversarial agents, and benchmark tooling used by Meta's AI Red Team to test Llama 3.
llama-3aired-teaming+2