Donglin Yang

My name is Donglin Yang (杨东霖). I am an MPhil student in the Department of Electrical and Electronic Engineering at The University of Hong Kong, advised by Prof. Xiaojuan Qi. I am currently a research intern at Tencent ARC Lab, working on long video generation. My research interests broadly span machine learning and deep generative modeling.

Previously, I was an intern with the KLING Team at Kuaishou Technology, and at the Shanghai AI Laboratory. I was also a Ph.D. student in the Department of Electrical and Computer Engineering at the University of British Columbia, under the supervision of Prof. Renjie Liao, and earlier a Ph.D. student in the Department of Computer Science at Northwestern University during 2022–2023, advised by Prof. Han Liu. I left both programs due to visa-related constraints. I received my B.Eng. in Electronic Information Science and Technology from the Department of Electronic Engineering at Tsinghua University in 2022, where I worked closely with Prof. Yang Gao on imitation learning.

News

Mar 18, 2026	One paper QDM was accepted to CVPR 2026 Findings!
Mar 10, 2026	Joined Tencent ARC Lab as a research intern, working on long video generation.
Jul 08, 2025	Joined KLING Team, Kuaishou Technology as a Research Intern.
Apr 22, 2025	Received AISTATS 2025 Best Reviewer Award! 🎉
Mar 08, 2025	One paper DiffuGTS was accepted to CVPR 2025!
Jan 22, 2025	One paper SymmetricDiffusers was accepted to ICLR 2025 as Oral (213/11672, Top 1.8%)!

Selected Publications [full list]

arXiv

Stable Velocity: A Variance Perspective on Flow Matching

Donglin Yang, Yongxing Zhang, Xin Yu, and 5 more authors

arXiv preprint arXiv:2602.05435 2026

arXiv Bib Code Website

@article{yang2026stablevelocity,
  title = {Stable Velocity: A Variance Perspective on Flow Matching},
  author = {Yang, Donglin and Zhang, Yongxing and Yu, Xin and Hou, Liang and Tao, Xin and Wan, Pengfei and Qi, Xiaojuan and Liao, Renjie},
  journal = {arXiv preprint arXiv:2602.05435},
  year = {2026},
}

CVPR Findings

QDM: Quadtree-Based Region-Adaptive Sparse Diffusion Models for Efficient Image Super-Resolution

Donglin Yang, Paul Vicol, Xiaojuan Qi, and 2 more authors

CVPR Findings 2026

arXiv Bib Code

@inproceedings{yang2026qdm,
  title = {QDM: Quadtree-Based Region-Adaptive Sparse Diffusion Models for Efficient Image Super-Resolution},
  author = {Yang, Donglin and Vicol, Paul and Qi, Xiaojuan and Liao, Renjie and Zhang, Xiaofan},
  booktitle = {CVPR Findings},
  year = {2026},
}

ICLR

SymmetricDiffusers: Learning Discrete Diffusion on Finite Symmetric Groups

Yongxing Zhang, Donglin Yang, and Renjie Liao

ICLR 2025 (Oral Presentation, 213/11672, Top 1.8%)

arXiv Bib Code

Oral Presentation, 213/11672, Top 1.8%

@article{zhang2024symmetricdiffusers,
  title = {SymmetricDiffusers: Learning Discrete Diffusion on Finite Symmetric Groups},
  author = {Zhang, Yongxing and Yang, Donglin and Liao, Renjie},
  journal = {ICLR},
  year = {2025},
}

NeurIPS

On Sparse Modern Hopfield Model

Jerry Yao-Chieh Hu, Donglin Yang, Dennis Wu, and 3 more authors

NeurIPS 2023

arXiv Bib Code

@article{hu2023sparse,
  title = {On Sparse Modern Hopfield Model},
  author = {Hu, Jerry Yao-Chieh and Yang, Donglin and Wu, Dennis and Xu, Chenwei and Chen, Bo-Yu and Liu, Han},
  journal = {NeurIPS},
  volume = {36},
  pages = {27594--27608},
  year = {2023},
}

ICML

Adaptive Model Design for Markov Decision Process

Siyu Chen^*, Donglin Yang^*, Jiayang Li, and 3 more authors

ICML 2022 (Spotlight)

Bib Website

Spotlight

@inproceedings{chen2022adaptive,
  title = {Adaptive Model Design for Markov Decision Process},
  author = {Chen, Siyu and Yang, Donglin and Li, Jiayang and Wang, Senmiao and Yang, Zhuoran and Wang, Zhaoran},
  booktitle = {ICML},
  pages = {3679--3700},
  year = {2022},
  organization = {PMLR},
}