Yuchen Zhang

profile.png

Hi there!👋 I am currently a third-year Ph.D. student in Computer Science at Rensselaer Polytechnic Institute (RPI), advised by Dr. Mohammad Mohammadi Amiri. I obtained a Bachelor’s degree from Nanjing University of Information Science & Technology in 2021 and a Master’s degree from Washington University in St.Louis (WashU) in 2023.

My research interest lies broadly in the filed of Machine Learning and Artificial Intelligence. I aim to make AI be more intelligent by focusing on enhancing model efficiency and generalizability through both data-centric and model-centric approaches. My current work centers on the fundamental challenge of foundation model generalization, particularly investigating how these models can reliably scale their capabilities. For instance, generalizing to longer sequence lengths or novel compositional structures beyond their training data. Furthermore, I am also interested in investigating Agentic AI for achieving scalable coordination and collective intelligence in complex environments.

The journey is still unfolding, and I’m always eager to connect, collaborate, and learn along the way. Please feel free to reach out to discuss potential research collaborations or shared interests!

News

Sep 01, 2025 I will be joining AWS as an Applied Scientist Intern.
Aug 22, 2024 First launch of my website.

Industry Experience

AWS Logo
Amazon Web Services (AWS)
Applied Scientist Intern
Sep 2025 - Dec 2025

Service

Reviewer

Neurips, ICLR, ICML, IEEE Transactions on Big Data

Teaching Experience

Teaching Assistant

CSCI 4963 & CSCI 6963 Reinforcement Learning (RPI)
CSCI 2200 Foundations of Computer Science (RPI)
ESE 524 Estimation and Detection Theory (WashU)
ESE 520 Probability Theory and Stochastic Process (WashU)

Selected Publications

  1. influenceFunction.png
    Toward Efficient Influence Function: Dropout as a Compression Tool
    Yuchen Zhang and Mohammad Mohammadi Amiri
    arXiv preprint arXiv:2509.15651, 2025
  2. teaser-new.jpg
    Self-control of llm behaviors by compressing suffix gradient into prefix controller
    Min Cai, Yuchen Zhang, Shichang Zhang, and 5 more authors
    arXiv preprint arXiv:2406.02721, 2024