InGuard: Towards Generalized Inner Guardrail for Safe Text-to-Image Generation
The Inner Guardrail (InGuard) is proposed, a safety framework that works inside the pipeline on the model's own representations, leaving base-model parameters untouched and the RevGen Safety Benchmark to evaluate T2I safety under realistic conditions.
Ze-Yu Wang, Xiao-Dan Li, Zhi-Wen Li et al.
· 0 citations