Yueqi Xie

Yueqi Xie 谢悦琪

I am a final-year Ph.D student in Department of Computer Science and Engineering at HKUST. I am also a visiting student researcher in Department of Sociology at Princeton University. My supervisors are Prof. Sunghun KIM and Prof. Qifeng Chen. I graduated from Peking University in 2020 with a bachelor's degree.

Email / Google Scholar

Research

I work on interdisciplinary study of AI and Social Science, including AI's social impact, socially responsible AI (AI safety, security, and privacy), and computational social science. My long-term goal is to understand and advance the relationship between AI and society, exploring both its technological and social dimensions.

Topics I currently focus on:

Socially Responsible Generative AI: [Nature Machine Intelligence'23], [ACL'24], [EMNLP'24], [KDD'25]

AI's Social Impact / Computational Social Science: Human-Contribution-Evaluation

Secure and Robust Federated Learning: [CVPR'23], [ICML'24], PoisonedFL, SmartFL

Others:

Recommender Systems: [SIGIR'22], [WSDM'23], [RecSys'23a], [RecSys'23b], [WWW'24]

Image Compression: [ACMMM'21], [ICCV'21], [ECCV'22]

Education / Experience

2020-Now: Ph.D. Student in Computer Science, HKUST
2024/4-Now: Visiting Student Researcher in Sociology at Princeton University, Princeton
2023/3-2024/2: Visiting Student Researcher at Duke University, Remote
2022/3-2022/8: Research Intern at Microsoft Research Asia, Beijing
2016-2020: B.S. in Intelligence Science and Technology, Peking University
2019/6-2019/9: Visiting Student at HKUST, Hong Kong

Selected Work

(* joint first authors)

Full Publication List on Google Scholar.

Socially Responsible Generative AI

Defending ChatGPT against Jailbreak Attack via Self-Reminder
Yueqi Xie*, Jingwei Yi*, Jiawei Shao, Justin Curl, Lingjuan Lyu, Qifeng Chen, Xing Xie, Fangzhao Wu
Nature Machine Intelligence, 2023
/ / / Media Coverage: TechXplore, Tech Times

GradSafe: Detecting Jailbreak Prompts for LLMs via Safety-Critical Gradient Analysis
Yueqi Xie, Minghong Fang, Renjie Pi, Neil Gong
Annual Meeting of the Association for Computational Linguistics (ACL), 2024
/ /

MLLM-Protector: Ensuring MLLM's Safety without Hurting Performance
Renjie Pi*, Tianyang Han*, Jianshu Zhang*, Yueqi Xie, Rui Pan, Qing Lian, Hanze Dong, Jipeng Zhang, Tong Zhang
The Conference on Empirical Methods in Natural Language Processing (EMNLP), 2024
/ /

Benchmarking and Defending Against Indirect Prompt Injection Attacks on Large Language Models
Jingwei Yi*, Yueqi Xie*, Bin Zhu, Emre Kiciman, Guangzhong Sun, Xing Xie, Fangzhao Wu
ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD), 2025
/ /

AI's Societal Impact / Computational Social Science

Measuring Human Contribution in AI-Assisted Content Generation
Yueqi Xie*, Tao Qi*, Jingwei Yi*, Ryan Whalen, Junming Huang, Qian Ding, Yu Xie, Xing Xie, Fangzhao Wu
Preprint, 2024
/ /

Secure and Robust Federated Learning

Defending against Model Poisoning Attacks for Federated Learning using Model Update Reconstruction Error
Yueqi Xie, Minghong Fang, Neil Gong
International Conference on Machine Learning (ICML), 2024
/ /

DYNAFED: Tackling Client Data Heterogeneity with Global Dynamics
Renjie Pi*, Weizhong Zhang*, Yueqi Xie*, Jiahui Gao, Xiaoyu Wang, Sunghun Kim, Qifeng Chen
The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023
/ /

PoisonedFL: Model Poisoning Attacks to Federated Learning via Multi-Round Consistency
Yueqi Xie, Minghong Fang, Neil Gong
Preprint, 2024
/

Source code from Jon Barron, Aljaž Božič, XINTAO and Ekun.