Zexue He

I am a fourth-year Ph.D. candidate advised by Prof. Julian McAuley in the Computer Science department at the University of California, San Diego. Prior to joining UCSD in 2020, I earned my B.S. degree in Computer Science from Beijing Normal University.

My research interests covers a wide range of topics related to language models and multimodal learning:

  • Trustworthy LLM: fairness, robustness, interpretability, and controllability;
  • Data Efficiency: learning with synthetic data;
  • LLM Applications: domain-specific LLMs, e.g., medical or high-fashion LLMs.
  • Memory-augmented LLM : novel architectures for long-context language/video tasks (my new research direction!).
I am honored to be awarded the IBM PhD Fellowship !

Email  /  Google Scholar  /  DBLP  


Ph.D.          2020 - 2024 (expected)
                       Computer Science and Engineering, University of California San Diego (UCSD), U.S.
                       Ph.D. student in Computer Science
                       Advisor: Prof. Julian McAuley

B.S.              2015 - 2019
                       College of Information Science and Technology, Beijing Normal University (BNU), China
                       B.S. in Computer Science and Technology

Trustworthy LLM
Zexue He, Marco Tulio Ribeiro, Fereshte Khani
ACL 2023 .
Bodhisattwa Majumder* Zexue He*, Julian McAuley
EMNLP 2023. *Equal.
Zexue He, Yu Wang, Julian McAuley, Bodhisattwa Majumder
EMNLP2022 Findings
Canwen Xu, Zexue He, Zhankui He, Julian McAuley
AAAI 2022 .
Zexue He, Bodhisattwa Majumder, Julian McAuley
EMNLP 2021 Findings.
Haohan Wang, Zexue He, Prof. Zachary C. Lipton, Eric P. Xing.
ICLR 2019, Oral Presentation.
Fuli Luo, Tianyu Liu, Zexue He, Qiaolin Xia, Zhifang Sui Baobao Chang.
EMNLP 2018
Zexue He*, Graeme Blackwood*, Rameswar Panda, Julian McAuley, Rogerio Feris
ACL2023 Findings.
*Equal Contribution.
Noveen Sachdeva, Zexue He, Wang-Cheng Kang, Jianmo Ni, Derek Zhiyuan Cheng, Julian McAuley
Preprint 2023
LLM Applications
Yu Wang, Zexue He, Zhankui He, Hao Xu, Julian McAuley
AAAI 2024 .
Zexue He, Yu Wang, An Yan, Yao Liu, Eric Y Chang, Amilcare Gentili, Julian McAuley, Chun-Nan Hsu
EMNLP 2023.
Zexue He, An Yan, Amilcare Gentili, Julian McAuley, Chun-Nan Hsu
AAAI 2023.
An Yan, Yu Wang, Yiwu Zhong, Chengyu Dong, Zexue He, Yujie Lu, William Wang, Jingbo Shang, Julian McAuley
ICCV 2023.
An Yan, Yu Wang, Petros Karypis, Zexue He, Amilcare Gentili, Chun-Nan Hsu, Julian McAuley
NeurIPS 2023 Medical Imaging Workshop.
An Yan, Zexue He, Xing Lu. Jiang Du, Eric Chang, Amilcare Gentili, Julian McAuley and Chun-Nan Hsu
EMNLP 2021 Findings.
Zexue He, Li Zhu, Minjie Li, Jinyao Li, Yiran Chen, Yanlin Luo.
SCI Journal: SCIENCE CHINA Information Sciences. Coappear at International Conference of Virtual Reality and Visualization (ICVRV) 2018.
Human in The Loop
Xiaochuan Wang, Ning Su, Zexue He, Yiqun Liu, Shaoping Ma.
SIGIR 2018
Jiaxin Mao, Yiqun Liu, Noriko Kando, Zeuxe He, Min Zhang, Shaoping Ma.
CHIIR 2018
Xiangsheng Li, Yiqun Liu, Jiaxin Mao, Zexue He, Min Zhang, Shaoping Ma.
CIKM 2018
Work Experience

MIT-IBM AI Lab, Cambridge, MA
Research Intern • summer 2023
Memory-augmented Large Langauge Models for Efficient Long-Context Modeling
Collaborators: Dr. Dmitry Krotov and Dr. Donghyun Kim at MIT-IBM
Advisors: Prof. Yoon Kim at MIT; Dr. Leonid Karlinsky and Dr. Rogerio Feris at MIT-IBM

Microsoft, Redmond, Washington
Research Intern • summer 2022
Targeted Data Generation
Advisor: Dr. Fereshte Khani Prof. Marco Tulio Ribeiro
Office of Applied Research

NEC Labs, Princeton, NJ
Research Intern • June. 2021 to Sept. 2021
Multimodality Data Representation Learning
Advisor: Dr. Yuncong Chen
Data Science & System Security Group

Microsoft Research Asia, Beijing, China
Research Intern • Oct. 2019 to Dec. 2019
Algorithmic Trading: High-Frequency Time Series Machine Learning and Data Mining
Advisor: Dr. Kan Ren
Machine Learning Group

Google, Beijing, China
Engineering Practicum Intern • Jul. 2017 to Sept. 2017
Knowledge Graph Source Discovery: Wikipedia-like Sites Discovery and Analysis
Advisor: Jiang Bian, team manager
Dataz Group

Machine Learning Department, Carnegie Mellon University, Pittsburgh, U.S.
Research Intern • Apr. 2018 to Oct. 2018
Robust Learning for Domain Generalization (DG) without Domain Information
Mentor: Haohan Wang, Ph.D. candidate at LTI, CMU
Advisors: Prof. Zachary C. Lipton.
Information Retrieval Group, Tsinghua University, Beijing, China
Research Assistant • Jun. 2017 to May 2018
Investigating Human Examination Behavior on Mobile Search
Advisor: Prof. Yiqun Liu
Key Laboratory of Computational Linguistic, Peking University, Beijing, China
Research Assistant • Dec. 2017 to May 2018
Leveraging Gloss Knowledge in Neural Word Sense Disambiguation (WSD)
Chinese Word Segmentation (CWS) with Character Glyph Embedding
Advisor: Prof. Baobao Chang
Engineering Research Center of Virtual Reality and Applications, Beijing Normal University, Beijing, China
Research Assistant • Jul. 2017 to Feb. 2017
Human Brain's CT-MRI Heterogeneous Data Fusion and Visualization
Advisor: Prof. Yanlin Luo
School of Life Science, Beijing Normal University, Beijing, China
Research Assistant • Nov. 2015 to Sept. 2017
Genetic Biological Parallel Computing System for NP-hard Problems
Advisor: Prof. Xudong Zhu
Honors & Awards

Gold Medal in International Genetically Engineered Machine Competition (iGEM) at Boston, Massachusetts, U.S. • 2016
Silver Medal in International Collegiate Programming Contest at Beijing regional site (ACM/ICPC, Beijing) • 2016
Bronze Medal in International Collegiate Programming Contest at Dalian regional site (ACM/ICPC, Dalian) • 2016
Best Female Team in China Collegiate Programming Final Contest (CCPC Final) • 2016

IBM Ph.D. Fellowship • 2022-2024 • IBM
TwoSigma Ph.D. Fellowship Final Nomination • 2022-2024 • TwoSigma
Jacobs School of Engineering Fellowship • 2020-2021 • University of California San Diego
The First-class Scholarship for Academic Excellence • 2018 • Beijing Normal University
The First-class Scholarship for Competition Excellence • 2016, 2017, 2018 • Beijing Normal University
Google Intern Scholarship • 2017 • Google Inc.

Gallery of My Previous Fun Projects

With our designed system, we made a movie about Daiyu Fengao, a chapter of Dream of the Red Chamber, one of China's Four Great Classical Novels

With our data fusion algorithm, we upgraded the VisAll visulaization platform, and made it into practium by deploying it at one hospital in Beijing.

An education App with which teachers can share notes in class in real time, arrange slides, broadcast, and display.