Our new preprint is online!! Super excited about this one :) "Unsupervised Discovery of Multimodal Links in Multi-Image, Multi-Sentence Documents" Code and data: (joint with Lillian Lee and @dmimno )