Matt and Sebastian discuss the challenges in representing relational knowledge in language models like Elmo and Bert. They explore how these models handle facts and common sense knowledge, showcasing the nuances in understanding specific details versus general concepts.
