Zhenyuan Dong

I am a second-year Ph.D. student in the Elmore Family School of Electrical and Computer Engineering at Purdue University, advised by Prof. Qiang Qiu.

Previously, I received my M.S. in Computer Engineering from New York University and my B.Eng. in Computer Science from Shanghai University.

News

Sept. 2026One paper is accepted to NeurIPS 2026.
Nov. 2025One paper is accepted to AAAI 2026.
Aug. 2025I started my Ph.D. in the Elmore Family School of Electrical and Computer Engineering at Purdue University, advised by Prof. Qiang Qiu.
Aug. 2025One paper is accepted to Findings of EMNLP 2025.
May 2025I received the Myron M. Rosenthal Award for Best MS Academic Achievement at New York University.
Oct. 2024One paper is accepted to WACV 2025.

Research Interests

My research interests include machine learning and computer vision, with a focus on generative models, self-supervised learning, and representation learning. In particular, I am interested in representation learning in both feature space and parameter space for generative models and world modeling, especially on post-training and test-time learning.

Publications

* Equal contribution

Model embedding space aligns task-specific updates in a shared subspace.

NeurIPS 2026

Approaching Effective Merging in Model Embedding Space

Wei Chen, Zhenyuan Dong, Qiang Qiu

Advances in Neural Information Processing Systems (NeurIPS 2026).

AD-L-JEPA architecture with context and target encoders for LiDAR representation learning.

AAAI 2026

Self-Supervised Representation Learning with Joint Embedding Predictive Architecture for Automotive LiDAR Object Detection

Haoran Zhu, Zhenyuan Dong, Kristi Topollai, Beiyao Sha, Anna Choromanska

Proceedings of the AAAI Conference on Artificial Intelligence (AAAI 2026).

Project / Paper / Code

Diffusion Transformer activations before and after enhanced activation smoothing.

WACV 2025

DiTAS: Quantizing Diffusion Transformers via Enhanced Activation Smoothing

Zhenyuan Dong, Sai Qian Zhang

IEEE/CVF Winter Conference on Applications of Computer Vision (WACV 2025).

Draft and target language models in a generation-refinement decoding framework.

Findings of EMNLP 2025

Mitigating Sequential Dependencies: A Survey of Algorithms and Systems for Generation-Refinement Frameworks in Autoregressive Models

Yunhai Hu*, Zining Liu*, Zhenyuan Dong*, Tianfan Peng*, Bradley McDanel, Sai Qian Zhang

Findings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2025 Findings).

Teaching & Service

Academic Service

Conference Reviewer: CVPR 2026; NeurIPS 2026; AAAI 2026, 2027.

Teaching

ECE 20875: Python for Data Science, Purdue University
ECE-GY 9483: Efficient AI and Hardware Accelerator Design, New York University

Honors

Myron M. Rosenthal Award for Best MS Academic Achievement, New York University
Graduate Scholarship, New York University
Outstanding Graduates, Shanghai University
Outstanding Students, Shanghai University