Polos: Multimodal Metric Learning from Human Feedback for Image Captioning Paper • 2402.18091 • Published Feb 28, 2024
