About Me
Hi, I’m Yifan!
I’m a postdoc with the EconCS group at Microsoft Research, New England. I obtained my PhD from Northwestern University, where I was advised by Prof. Jason Hartline. I received my B.S. in Computer Science from Turing Class at Peking University in 2020, where I was advised by Prof. Yuqing Kong.
I have a broad interest in the intersections between theoretical computer science, AI, and economics. Recently, my research focuses on the theory for AI trustworthiness.
Dropbox link to my [CV].
My Research
My research develops theoretical foundations for understanding the trustworthiness of AI systems, especially when AI systems make predictions under uncertainty. Drawing on ideas from game theory and statistical decision theory, I study how to evaluate probabilistic predictions, how such predictions support good decisions, and how to design AI systems with provable guarantees.
Across these questions, I view AI systems through the lens of decision-making under uncertainty. A common theme in my work is to model components of an AI system as rational agents optimizing expected payoffs. This perspective helps characterize optimal decision-making and identify sources of suboptimality in real systems. It applies to predictors, human decision-makers, and downstream decision processes.
More specifically, my research includes:
- Optimal information elicitation, designing mechanisms to elicit high-quality information, which provides a theoretical foundation for my broader framework.
- Calibration, studying when predictive model outputs can be reliably interpreted as probabilities. I focus on the evaluation of miscalibration and how evaluation guides algorithm design.
- Obtaining provable guarantees from AI systems, such as strategic robustness and no-regret.
- Theoretical benchmark for human-computer interaction and human-AI interaction, aimed at understanding human behavior in algorithm- and AI-assisted decision-making.
Mathematically, many of these questions are studied using proper scoring rules, a tool for evaluating probabilistic predictions and incentivizing truthful reports. Proper scoring rules are fundamental in game theory, statistical decision theory, and machine learning.