1 paper · 1 filter
Hongcheng Gao, Tianyu Pang, Chao Du +3
With the rapid progress of diffusion-based content generation, significant efforts are being made to unlearn harmful or copyrighted concepts from pretrained diffusion models (DMs)…