Research Interests
My research seeks to unveil and leverage the hidden world of observational data, aiming to uncover the truthful process underlying the data and thus go beyond the ceiling of correlation fitting. I develop foundations for causal representation learning, tackling open problems in trustworthy machine learning, and translate these insights into foundation models (e.g., world models) for both efficiency and reliability.
Education
PhD, Carnegie Mellon University, 2026
Additional Campus Affiliations
Siebel School of Computing and Data Science
Recent Publications
DSReg: Provably Recovering Individual World Latents without Reconstruction
Yujia Zheng, David Klindt, Randall Balestriero, Bernhard Schölkopf
Preprint
How Causality Bridges the Semantic Gap
Shuhao Zhang, Xuran Zhou, Han Guo, Pengtao Xie, Yujia Zheng
Preprint
SplitJEPA: Learning Invariant and Variant Latent Worlds without Reconstruction
Ruijin Hua, Zichuan Liu, Zhuokai Zhao, Yujia Zheng
Preprint
From Generalist to Specialist Representation
Yujia Zheng, Fan Feng, Yuke Li, Shaoan Xie, Kevin Murphy, Kun Zhang
ICML 2026
Diverse Dictionary Learning
Yujia Zheng, Zijian Li, Shunxing Fan, Andrew Gordon Wilson, Kun Zhang
ICLR 2026
Thought Communication in Multiagent Collaboration
Yujia Zheng, Zhuokai Zhao, Zijian Li, Yaqi Xie, Mingze Gao, Lizhu Zhang, Kun Zhang
NeurIPS 2025
View more on Google Scholar.