Bilkent University
Department of Computer Engineering
SEMINAR
Exploring Reliability Issues in Generative AI
Dr.Soheil Feizi
Abstract: Generative AI, such as text-to-image synthesis and large-scale language models, have achieved extensive practical application across numerous domains. In this presentation, I will discuss our most recent insights into a range of reliability challenges associated with these models. This includes their susceptibility to jailbreaking, issues with knowledge localization, and the effectiveness of detection and alignment strategies.
Biography: Dr. Feizi is an Associate Professor and the director of Reliable AI Lab in the Computer Science Department at the University of Maryland, College Park. He holds a Ph.D. from MIT and was a post-doctoral research scholar at Stanford University prior to his appointment at UMD. Dr. Feizi has received multiple awards for his work on developing reliable and trustworthy Artificial Intelligence including the ONR's Young Investigator Award, the NSF CAREER award, the ARO's Early Career program Award, two best paper awards, the Ernst Guillemin thesis award, a teaching award, and more than fifteen research awards from national agencies and industry.
DATE: July 30, Tuesday @ 11:00
Place: EA 409