See LionPATH to schedule.
Course Topic: Advances in Generative AI and Applications
This graduate special topics course surveys recent advances in generative AI and focuses on reading, presentation, and project-based learning rather than traditional lectures. Students will study contemporary methods and systems for building generative AI applications, such as large language models, diffusion models, multimodal models, and agents. In this course, students will analyze whether these systems are reliable and trustworthy, and how they are evaluated and deployed in real-world settings. Students will examine common failure modes (e.g., hallucinations, robustness and security issues, privacy risks) and practice strategies for testing, measuring, and improving system performance. The course is designed to support both students aiming for research (e.g., reading papers, developing novel ideas) and students focused on applications (e.g., building and evaluating end-to-end GenAI systems). The course culminates in a substantial team project where students build and evaluate a GenAI system for real problems or conduct a research-oriented study.
