A New AI Research from Tel Aviv and the University of Copenhagen Introduces a ‘Plug-and-Play’ Approach for Rapidly Fine-Tuning Text-to-Image Diffusion Models by Using a Discriminative Signal