Yunzhe Zhang | Brandeis University

About Me

I am now a fourth-year Ph.D. student in Computer Science at Brandeis University, advised by Professor Pengyu Hong. Before joining Brandeis, I received my B.S. in Computer Science and Statistics from the University of Illinois Urbana–Champaign (UIUC) and my M.S. in Data Science from Columbia University.

My research focuses on efficient image generation with diffusion and flow-based models. I study training-time distillation and inference acceleration, as well as how inference-time choices, such as seed selection, affect generation quality and alignment with text prompts.

I am also interested in video generation and synthesis, particularly maintaining long-term temporal consistency and coherence across frames. If you share similar interests or have ideas to discuss, please feel free to reach out—I would love to exchange ideas and explore potential collaborations.

News

[2026-10] 🔥 Our paper CMCM-DLM: Cross-Modality Controlled Molecule Generation with Diffusion Language Model has been accepted to IEEE Big Data 2026!
[2026-09] 🔥 Our paper ABSS has been accepted to NeurIPS 2026! I am grateful to Professor Hongfu Liu and Professor Pengyu Hong for their guidance and support.
[2026-09] 🔥 I completed my research internship at HiDream.ai (May–September 2026), working closely with Chengmin Gao, Dr. Qi Cai, and Dr. Tao Mei on efficient generative modeling and distillation. I am grateful for their guidance, support, and inspiring discussions!
[2026-05] 🔥 I joined HiDream.ai as a research intern.

Publications & Preprints

arXiv

SpurFlow: Spatially-Adaptive Residual Steering for Few-Step Image Editing

Yunzhe Zhang, Chengmin Gao, Qi Cai, Yingwei Pan, Pengyu Hong^†, Ting Yao, and Tao Mei^†

Submitted to ICLR 2027

Work completed during my research internship at HiDream.ai.

PDF will be released soon.
NeurIPS 2026

Boosting Text-to-Image Diffusion Models via Core Token Attention-Based Seed Selection

Yunzhe Zhang, Hongfu Liu, and Pengyu Hong

Accepted to NeurIPS 2026

PDF CODE
IEEE Big Data 2026

Cross-Modality Controlled Molecule Generation with Diffusion Language Model

Yunzhe Zhang, Yifei Wang, Khanh Vinh Nguyen, and Pengyu Hong

Accepted to IEEE Big Data 2026

PDF

Education

Brandeis University
Ph.D. in Computer Science
Sep. 2023 – May 2028 (expected) · Waltham, MA
Advisor: Prof. Pengyu Hong

Columbia University
M.S. in Data Science
Aug. 2021 – Dec. 2022 · New York, NY

University of Illinois Urbana–Champaign
B.S. in Computer Science and Statistics
Aug. 2018 – May 2021 · Urbana, IL
Dean's List and Highest Graduation Distinction

Teaching

COSI 159a-1: Computer Vision

[2026] Teaching Assistant (TA) with Prof. Hongfu Liu.

COSI 126a-1: Unsupervised Learning and Data Mining

[SP 2025] Teaching Assistant (TA) with Prof. Hongfu Liu.

COSI 104a-1: Introduction to Machine Learning

[SP 2024] Teaching Assistant (TA) with Prof. Pengyu Hong.

COSI 29a-1/2: Discrete Structures

[FA 2023] Teaching Assistant (TA) with Prof. Tim Hickey.

Work Experience

Machine Learning Researcher · HiDream.ai
Internship · May 2026 – Sep. 2026
Shanghai, China · On-site

Machine Learning Researcher · Accenture
Part-time · Sep. 2022 – Dec. 2022
New York, United States · Remote

Data Scientist Intern · WuXi AppTec
Apr. 2020 – Sep. 2020
Shanghai, China