Probing the Limits of Stylistic Alignment in Vision-Language Models

Open in new window