RESEARCHInvestigateNEXT 12 MONTHS
Injection-Execution Dissociation: A Mechanistic Evaluation of Persistent Memory Attacks and Defenses in Stateful LLM Agents
arXiv cs.LG — Machine Learning
Factual evidence
What the source reports
Researchers demonstrate that LLM agents store malicious prompt injections 97.5% of the time, but downstream execution varies independently.
Open source