Structure and Generalization in Multimodal Language Understanding (SAGE-MLU)

The first SAGE-MLU workshop brings together speakers and participants to explore whether and how current (vision and) language models can use abstract conceptual and linguistic structure to generalize in human-like ways. 

The program is split over two days and features four keynote speakers, interactive discussion panels, and a hackathon. 

We welcome researchers and students from various fields, such as (computational) linguistics, conversation analysis, computer science, AI, psychology, neuroscience, and philosophy. We especially encourage Master's & PhD students to attend the workshop, as there will be opportunities for mentorships and establishing research collaborations.