World Models

Catalog description

This course introduces world models—learned representations and predictors of environment dynamics for perception, planning, and decision making. We cover key approaches for learning predictive models and representations, and how they support control and reasoning in settings such as reinforcement learning, video/3D understanding and generation, LLM-based multimodal agents and robotics. Students complete a course project exploring a world-modeling method in one application domain. To prepare for this course students are encouraged to have taken CIS 5190 or CIS 5200 or equivalent background.