Large Language Models as Falsifiers for Cyber-Physical Systems
A research paper explores using LLMs to falsify cyber-physical system models, which could improve safety verification but also lower the barrier to finding exploitable flaws in critical infrastructure.
A research paper explores using LLMs to falsify cyber-physical system models, which could improve safety verification but also lower the barrier to finding exploitable flaws in critical infrastructure.
Auto-scored: Dual-use capability research that could aid both safety verification and adversarial discovery of CPS vulnerabilities.