Eric Nalisnick

Eric Nalisnick is an Assistant Professor in the Department of Computer Science at Johns Hopkins University. He holds affiliations with the Institute for Assured Autonomy, Mathematical Institute for Data Science, and Data Science and AI Institute. His research focuses on developing safe and robust intelligent systems through probabilistic modeling and computational statistics, with applications in healthcare, online content moderation, and sign language processing. His work emphasizes human-centered design, exploring how to incorporate prior knowledge, detect system failures, and integrate human-machine decision-making.

Research Trends: Recent publications highlight advancements in uncertainty quantification (e.g., Lightning UQ Box tool), early-exiting neural networks for risk control, and generative models for symmetry transformations. He also investigates calibration in multi-distribution learning and ethical considerations in hate speech detection systems.

Advising & Contributions: Eric mentors a diverse cohort of PhD students and has developed influential open-source tools like Lightning UQ Box and Learning to Defer frameworks. His teaching includes courses on Deep Learning, Human-in-the-Loop Machine Learning, and Bayesian Methods.