Ph.D. Candidate in Electrical & Computer Engineering
About
I am a post-candidacy Ph.D. candidate in Electrical and Computer Engineering at The Ohio State University, advised by Prof. Aylin Yener in the INSPIRE@OhioState Lab.
My research develops learning methods for institutions and edge devices that need to collaborate while keeping local data and model information private. I work across algorithms, theory, and systems evaluation.
Xue Zheng · The Ohio State University
Research interests
Collaborative learning across heterogeneous clients without centralizing their data.
Secure aggregation and peer-to-peer methods that protect local information.
Robust learning when deployed models influence future data distributions.
Resource-aware federated fine-tuning and sparse expert assignment.
Foundation models · Federated learning · Edge AI
My ongoing work explores how pretrained foundation models, including large language models and mixture-of-experts architectures, can be collaboratively trained, adapted, and deployed across distributed edge clients without centralizing their local data.
Distribution shifts · Robust optimization
Deployed models can alter the data they later observe. This line of work studies robust federated optimization and decentralized learning over changing interaction graphs when model deployment, network evolution, and data shifts are tightly coupled.
Personalization · Secure aggregation
Personalized agents learn whom to collaborate with while keeping individual models private. The work combines directed collaboration graphs, clustering, and Shamir secret sharing for practical secure aggregation in peer-to-peer networks.
Teaching & Mentoring
Lecturer and graduate student mentor, 2024–2026. Topics include probability, research problem formulation, and probabilistic generation in modern language models.
News
Presentations