On Monday, March 31, 2025, the Learning Informatics Lab hosted Karin de Langis from the University of Minnesota. In this talk, she discussed the current research landscape around cognition in Large Language Models (LLMs), as well as the methodological challenges involved. She also discussed her work with Minnesota NLP and highlighted several of their recent findings on LLM performance on tasks across multiple domains, including memory, executive function, and narrative comprehension.