Marco Arazzi

The paper introduces SecureBreak, a manually annotated, safety-oriented dataset designed to help detect harmful outputs from large language models (LLMs) that bypass existing security alignments.

This paper provides the first comprehensive Systematization of Knowledge (SoK) on the security aspects of LLM-as-a-Judge (LaaJ) systems, identifying key vulnerabilities and proposing a taxonomy for future research.

The paper introduces NeWTral, a framework that restores safety alignment to specialized LLM adapters without sacrificing their domain-specific knowledge, achieving a significant reduction in attack success rates while maintaining high fidelity.