Biostatistics and Medical Informatics Department Seminar with Benjamin Lengerich (University of Wisconsin-Madison

AI models, especially foundation model–based approaches, have become remarkably effective across a wide range of biomedical & scientific tasks. Their success raises a natural question: should we still care about interpretability? In this talk, Dr. Lengerich will present a system-level argument for interpretability in biomedicine being valuable for at least three reasons, and will illustrate the roles with examples from maternal health, treatment effect heterogeneity, & personalized gene networks