Unified Safety Evaluation Benchmark (380 prompts)

Combined benchmark of ~380 prompts across 8 safety domains sourced from HarmBench, JailbreakBench, StrongREJECT, SALAD-Bench, and domain-specific suites

Neighborhood — ranked by edge-count

Papers (1)

paper

Persona Non Grata: Single-Method Safety Evaluation Is Incomplete for Persona-Imbued LLMs
mentions