From cs-speakerseries 9/16/2024
Abstract:
Data plays an increasingly crucial role in both the performance and the safety of AI models. In this talk, I will advocate for data attribution--a family of techniques aimed at quantifying the impact of individual training data points on a model trained on them--as a principled data-centric approach to ensuring trustworthy development and deployment of AI. I will begin by discussing our recent studies on the applications of data attribution, including its role in addressing the copyright challenges posed by generative AI and in supporting the development of robust counterfactual explanation methods for AI models under dataset shift. Additionally, I will cover our recent progress in overcoming the technical challenges of data attribution, particularly in terms of robustness and efficiency.
Bio:
Jiaqi Ma is an Assistant Professor in the School of Information Sciences at University of Illinois Urbana-Champaign (UIUC). His research interests lie in the broad area of trustworthy AI, with recent focuses including data attribution, machine unlearning, explainable machine learning, and training data curation. Jiaqi's work has been recognized with the Gary M. Olson Outstanding Student Award from University of Michigan and a Best Paper Award form the DPFM Workshop at ICLR 2024. Prior to joining UIUC, Jiaqi earned his PhD from the University of Michigan and worked as a postdoctoral researcher at Harvard University.