Enforcing LLM Safety through DMD-based Classification of Prompt-Response Embedding Dynamics

Abstract

arXiv:2608.19579v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly deployed in high-stakes applications, yet their tendency to generate toxic, harmful, or policy-violating content poses significant risks. Detecting these unsafe outputs efficiently in a black-box manner remains an open challenge. In this paper, we extend a recently proposed dynamical systems framework designed for hallucination detection to LLM safety classification. By projecting both prompts and respo