However, the best designs are a product of tremendous work of high-skilled domain experts. I will show that we are on the verge of a transition where computational methods start beating humans at design. I will describe a series of questions that need to be addressed to move the field of computational design forward: how to represent a design, how to represent design space, how to find designs with optimal performance, and how to bridge the gap between simulation and reality.
His research interests are in computer graphics, computational design and fabrication, computer vision, robotics and human-computer interaction. Improving Trustworthiness in Foundation Models: Assessing, Mitigating, and Analyzing ML Risks Chulin Xie University of Illinois Urbana-Champaign hosted by Jana Hofmann , - β¦ Virtual talk Abstract: As machine learning ML models continue to scale in size and capability, they expand the surface area for safety and privacy risks, raising concerns about model trustworthiness and responsible data use.
My research uncovers and mitigates these risks. In this presentation, I will focus on the three cornerstones of trustworthy foundation models and agents: safety, privacy, and generalization. For privacy, I will present a solution for protecting data privacy with a synthetic text generation algorithm under differential privacy guarantees. For generalization, I will introduce our study on the interplay between the memorization and generalization of LLMs in logical reasoning during the supervised fine-tuning SFT stage.
Finally, I will conclude with my future research plan for assessing and improving trustworthiness in foundation model-powered ML systems. Her research focuses on the principles and practices of trustworthy machine learning, addressing the safety, privacy, and generalization risks of Foundation Models, agents, and federated distributed learning.
Focusing on applications in criminal justice and social services, this talk will examine the significant gaps between current AI capabilities and the demands of real-world high-stakes decision-making. I will demonstrate critical shortcomings of current approaches to AI transparency, fairness, and effective human oversight, and discuss my work on addressing these issues, and its impact on policy and UK public services to date.
