VMix: Improving Text-to-Image Diffusion Model with Cross-Attention Mixing Control Paper • 2412.20800 • Published 8 days ago • 8