I am an incoming Ph.D. student in Machine Learning at Mohamed bin Zayed University of Artificial Intelligence (MBZUAI). My research centers on data mining, multimodal large language models, and representation learning, with a focus on how models extract, organize, and use signals from user behavior, multimodal content, and large-scale interaction data.

I am especially interested in building representations that remain useful beyond clean benchmark settings. For example, in recommender systems and value prediction, the hardest problems often come from noisy modalities, delayed feedback, shifting user cohorts, and rare but important high-value cases. My goal is to connect representation-level understanding with decision-facing AI systems: models should not only predict well, but also support reliable reasoning, adaptation, and intervention in realistic environments.

Before starting my Ph.D., I studied Artificial Intelligence at Lanzhou University and worked across academic and industrial research settings, including Alibaba Group, MBZUAI, University College Dublin, Westlake University, and UNC Chapel Hill. These experiences shaped my interest in methods that are technically grounded, empirically useful, and aware of the gap between model behavior and real-world deployment.