Jie Cai

I'm a first-year PhD student at the Melady Lab in the Thomas Lord Computer Science Department at the University of Southern California. I am fortunate to be advised by Professor Yan Liu. Previously, I earned my Master's Degree at Tsinghua University where I was advised by Professor Wenwu Zhu and Professor Xin Wang. I obtained my Bachelor degree from South China University of Technology in Mathematics and applied Mathematics.

Email  /  Scholar

profile photo

Research

I am deeply interested in the fields of multimodal learning, large language models (LLMs), interpretability, and graph learning.

Specifically, my research focuses on exploring how multimodal systems can effectively integrate diverse data sources, exploring the capabilities and limitations of LLMs in complex tasks, and developing Explainable Artificial Intelligence (XAI) methods that provide meaningful insights into their decision-making processes.

Additionally, I am fascinated by graph learning techniques and their applications in representing and analyzing structured data, particularly in scenarios that require a combination of multimodal inputs and interpretability.

Publications

Beyond Forecasting: Compositional Time Series Reasoning for End-to-End Task Execution
Wen Ye, Yizhou Zhang, Wei Yang, Lumingyuan Tang, Defu Cao, Jie Cai, Yan Liu
arXiv, 2024
AutoGL: A Library for Automated Graph Learning [Project page]
Ziwei Zhang, Yijian Qin, Zeyang Zhang, Chaoyu Guan, Jie Cai, Heng Chang, Jiyan Jiang, Haoyang Li, Zixin Sun, Beini Xie, Yang Yao, Yipeng Zhang, Xin Wang, Wenwu Zhu
arXiv, 2024
Multimodal Graph Neural Architecture Search under Distribution Shifts
Jie Cai, Xin Wang, Haoyang Li, Ziwei Zhang, Wenwu Zhu
AAAI, 2024
Knowledge Graph Completion with Counterfactual Augmentation
Heng Chang, Jie Cai, Jia Li
WWW, 2023
Multimodal Continual Graph Learning with Neural Architecture Search
Jie Cai, Xin Wang, Chaoyu Guan, Yateng Tang, Jin Xu, Bin Zhong, Wenwu Zhu
WWW, 2022