Yi-Chung (Andrew) Chen
PhD Student in Electrical and Computer Engineering at Purdue University
Purdue University
West Lafayette, IN
chen5262@purdue.edu
I am a third-year PhD student in Electrical and Computer Engineering at Purdue University (started 2024), co-advised by Prof. Jing Gao and Prof. David Inouye. My research focuses on trustworthy machine learning and physical AI.
Prior to my PhD, I earned my M.S. in Communication Engineering from National Taiwan University (2021-2023), where I was advised by Prof. Ming-Syan Chen. I completed my B.S. in Electronic Engineering at National Yang Ming Chiao Tung University (2017-2021), where I was mentored by Prof. Hong-Han Shuai and Prof. Wen-Huang Cheng.
My primary research interest lies at the intersection of trustworthy machine learning and physical AI, with a focus on robustness, interpretability, and uncertainty. I am currently interested in uncertainty-aware planning with world models. Previously, I have explored generative classification, multi-agent behavior prediction, diffusion model quantization, and federated learning.
news
| Sep 2026 | Our paper StarCraft Motion: A Dataset for Agent Simulation in Adversarial and Partially Observable Scenarios was accepted to the NeurIPS 2026 Evaluations & Datasets (E&D) Track. |
|---|---|
| May 2026 | Joined Uber AV Labs in Sunnyvale, CA as a PhD Software Engineer Intern for the summer. |
| Jan 2026 | Our paper Your VAR Model is Secretly an Efficient and Explainable Generative Classifier was accepted to ICLR 2026. |
| Sep 2024 | Our paper StepbaQ: Stepping backward as Correction for Quantized Diffusion Models was accepted to NeurIPS 2024. |
| Aug 2024 | Started my PhD in ECE at Purdue University, co-advised by Prof. Jing Gao and Prof. David Inouye. |
experience
selected publications
- NeurIPS E&DStarCraft Motion: A Dataset for Agent Simulation in Adversarial and Partially Observable ScenariosIn Advances in Neural Information Processing Systems, Evaluations & Datasets Track, 2026
- NeurIPSStepbaQ: Stepping backward as Correction for Quantized Diffusion ModelsAdvances in Neural Information Processing Systems, 2024