State of Security in LLMs

3 papers · avg viability 5.7

View topic page

Top papers

AttriGuard: Defeating Indirect Prompt Injection in LLM Agents via Causal Attribution of Tool Invocations(7.0)
VidDoS: Universal Denial-of-Service Attack on Video-based Large Language Models(5.0)
Automating Agent Hijacking via Structural Template Injection(5.0)