Min Zhang’s interest in developing trustworthy artificial intelligence grew from observing real-world problems while using AI.

“I noticed that it sometimes generates harmful or overly confident but incorrect responses and I became concerned about how privacy can be better protected as we share more of our information with AI,” said Zhang, who is advised by Chang-Tien Lu.

“I also started to think about how human reasoning patterns could inspire the way AI systems learn and improve. These experiences motivated me to explore how to optimize models to address such issues,” she said.

In her research as a Ph.D. student at the Sanghani Center, Zhang is developing trustworthy AI agents by enhancing their reasoning capabilities and reliability via integrating techniques such as tool use, data augmentation, and uncertainty estimation. She is trying to build AI systems that are not only intelligent but also fair, privacy-preserving, reliable, and generalizable for real-world deployment. 

“I hope future AI can act like smart and responsible humans,” said Zhang.

Zhang’s first author publications include:

·      “Don’t Go To Extremes: Revealing the Excessive Sensitivity and Calibration Limitations of LLMs in Implicit Hate Speech Detection,” in main proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) 

·      “Can LLM Find the Green Circle? Investigation and Human-Guided Tool Manipulation for Compositional Generalization,” at the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2024)

Her collaborative papers include “Large Language Model Cascades with Mixture of Thoughts Representations for Cost-efficient Reasoning,” at the International Conference on Learning Representations (ICLR 2024).

Zhang obtained a bachelor's degree from Beijing Jiaotong University and a master's degree from the University of Chinese Academy of Sciences, both in electrical and electronics engineering.

In searching for a Ph.D. program in computer science, Zhang said she was drawn to Virginia Tech and the Sanghani Center because they offered a strong research environment in artificial intelligence and data science, the opportunity to collaborate with leading researchers, and the center’s emphasis on both fundamental and applied research.” 

She is projected to graduate in Spring 2027 and her career goal is a research role in either academia or industry.