ARGUS cuts context-aware prompt-injection success from 28.8% to 3.8% on AgentLure while keeping 87.5% clean utility, beating prior defenses on the security-utility tradeoff.
Findings of the Association for Computational Linguistics: ACL 2025 , pages=
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CR 2years
2026 2verdicts
CONDITIONAL 2representative citing papers
Activation-space divergence detects and corrects compromised LLM agents in multi-agent systems without interaction graphs or synchronized rounds, outperforming graph baselines especially under async stealthy attacks.
citing papers explorer
-
ARGUS: Defending LLM Agents Against Context-Aware Prompt Injection
ARGUS cuts context-aware prompt-injection success from 28.8% to 3.8% on AgentLure while keeping 87.5% clean utility, beating prior defenses on the security-utility tradeoff.
-
When Agents Go Rogue: Activation-Based Detection of Malicious Behaviors in Multi-Agent Systems
Activation-space divergence detects and corrects compromised LLM agents in multi-agent systems without interaction graphs or synchronized rounds, outperforming graph baselines especially under async stealthy attacks.