Xin Yu

Ph.D. student

I am currently a Ph.D. Candidate in Statistics at The Pennsylvania State University and am honored to be advised by Prof. Lingzhou Xue. Before that, I obtained my master's degree at Penn State in 2025 and obtained my bachelor's degree in Statistics from Nankai University in 2023. I was also fortunate to be supervised by Prof. Changliang Zou for my undergraduate thesis.

Email: xinlucasyu1026@gmail.com
Email LinkedIn Google Scholar X
Research Interests
  • Statistical Machine Learning: Interpretability, Generalization, and Robustness.
  • Post-Training in Foundation Models: Infrastructure Efficiency, Reasoning, and Human Alignment.
News
  • 2026-05-01 Our paper has been accepted to ICML 2026. 🧑‍💻
  • 2026-04-23 Received a full travel award to attend MLSS NYC 2026 at Columbia University. See you in NYC this summer! 🗽
  • 2026-02-06 Our paper has been accepted to ICLR 2026. 🧑‍💻
  • 2026-01-19 Started my internship at TikTok, focusing on multimodal understanding and exploring continual learning + agent-based automated error correction. ✨
  • 2025-12-01 Presented a poster at Neurips 2025, San Diego Convention Center.☀️
  • 2025-07-22 Presented a poster at ICML 2025, Vancouver Convention Center.☀️
  • 2025-05-22 Summer internship in Seattle at ByteDance Seed Infra🧑‍💻 ; worked with Cong Xie and Zhi Zhang.
  • 2023-05-01 Thrilled to share that I’ve completed my five-month internship at Microsoft Research Asia (MSRA); grateful to work with Yan Lu and Xiaoyi Zhang; honored with the Stars of Tomorrow award! 🧑‍💻
Some representative works

See Google Scholar for an updated list of publications.

[Towards Better Optimization For Listwise Preference in Diffusion Models]
Towards Better Optimization For Listwise Preference in Diffusion Models
Jiamu Bai, Xin Yu, Meilong Xu, Weitao Lu, Xin Pan, Kiwan Maeng, Daniel Kifer, Jian Wang, Yu Wang
International Conference on Learning Representations ICLR, 2026
Archive Code

[AltLoRA: Towards Better Gradient Approximation in Low-Rank Adaptation with Alternating Projections]
AltLoRA: Towards Better Gradient Approximation in Low-Rank Adaptation with Alternating Projections
Xin Yu, Yujia Wang, Jinghui Chen, Lingzhou Xue
Neural Information Processing Systems NeurIPS, 2025
Paper Code

[Understanding the Statistical Accuracy-Communication Trade-off in Personalized Federated Learning with Minimax Guarantees]
Understanding the Statistical Accuracy-Communication Trade-off in Personalized Federated Learning with Minimax Guarantees
Xin Yu, Zelin He, Ying Sun, Lingzhou Xue, Runze Li
International Conference on Machine Learning ICML, 2025
Paper Code

Service
Reviewer
  • The Thirteenth International Conference on Learning Representations (ICLR 2025): 5.
  • 2026 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2026): 2.
  • 2025 IEEE Transactions on Circuits and Systems for Video Technology: 1.
  • 2026 Pattern Recognition: 1.
  • 2026 IEEE Transactions on Neural Networks and Learning: 3.
  • International Journal of Data Science and Analytics: 1.
  • KDD 2026 AI4Sciences Track: 3.
  • Forty-Third International Conference on Machine Learning (ICML 2026): 6.
  • ECCV 2026: 2.
Conference / Seminar Organizer
  • 2025-11-01 🎤 Co-organizer of the session “Low-Rank Adaptation for Efficient Fine-Tuning in Foundation Models” at 2025 SIAM NNP Conference, Penn State.
Teaching Assistent/ Consultant
  • Statistical Consultant at Statistical Consulting Center
  • TA: STAT 184 : Introduction to R; STAT 400 : Statistical Modeling II; STAT 440: STAT 440: Computational Statistics