|
Zedong Wang I am currently a CSE PhD student at HKUST advised by Prof. Dan Xu. Previously, I obtained my B.Eng. in Electronics and Information Engineering from Huazhong University of Science and Technology, where I was fortunately to be advised by Prof. Xinggang Wang. I am honored to be a recipient of outstanding reviewer awards at multiple learning venues, including BMVC'25, ICLR'25 (rate: 2.6%), UniReps'25, ECCV'24 (rate: 2.7%), ACM MM'24, and BMVC'24. Email / CV / Google Scholar / Twitter (X) / LinkedIn |
|
News
|
ResearchMy overarching goal is to build scalable multimodal foundation models that understand, reason, and interact with the world as holistically as humans do. Specifically, I worked on (i) multi-task learning: Rep-MTL (ICCV'25 Highlight, top 2.3%), (ii) multi-modal understanding and generation: CARE-Edit (CVPR'26), (iii) efficient model architectures: MogaNet (ICLR'24). |
|
CARE-Edit: Condition-Aware Routing of Experts for Contextual Image Editing
Yucheng Wang*, Zedong Wang*, Yuetong Wu, Yue Ma, Dan Xu IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026 arXiv / project / Hugging Face / video A condition-aware mixture-of-experts framework for contextual image editing. |
|
Rep-MTL: Unleashing the Power of Representation-level Task Saliency for Multi-Task Learning
Zedong Wang, Siyuan Li, Dan Xu IEEE/CVF International Conference on Computer Vision (ICCV), 2025 (Highlight, Top 2.3%) arXiv / project / Hugging Face / video Exploiting representation-level task saliency for effective multi-task learning. |
|
MergeVQ: A Unified Framework for Visual Generation and Representation with Disentangled Token Merging and Quantization
Siyuan Li*, Luyuan Zhang*, Zedong Wang, Juanxi Tian, Cheng Tan, Zicheng Liu, Chang Yu, Qingsong Xie, Haonan Lu, Haoqian Wang, Zhen Lei IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025 arXiv / project / Hugging Face (Top-1 paper of the day) A unified framework for visual generation and representation with disentangled token merging and quantization. Ranked top 1 at Hugging Face Daily Papers on Apr 3rd, 2025. |
|
Taming LLMs by Scaling Learning Rates with Gradient Grouping
Siyuan Li*, Juanxi Tian*, Zedong Wang*, Xin Jin, Zicheng Liu, Wentao Zhang, Dan Xu Annual Meeting of the Association for Computational Linguistics (ACL), 2025 arXiv / project / Hugging Face Taming LLMs by scaling learning rates with gradient grouping. |
|
MogaNet: Multi-order Gated Aggregation Network
Siyuan Li*, Zedong Wang*, Zicheng Liu, Cheng Tan, Haitao Lin, Di Wu, Zhiyuan Chen, Jiangbin Zheng, Stan Z. Li International Conference on Learning Representations (ICLR), 2024 arXiv / project Multi-order gated aggregation network for efficient vision backbone. |
|
VQDNA: Unleashing the Power of Vector Quantization for Multi-Species Genomic Sequence Modeling
Siyuan Li*, Zedong Wang*, Zicheng Liu, Di Wu, Cheng Tan, Jiangbin Zheng, Yufei Huang, Stan Z. Li International Conference on Machine Learning (ICML), 2024 arXiv / OpenReview Unleashing the power of vector quantization for multi-species genomic sequence modeling. |
|
Short-Long Convolutions Help Hardware-Efficient Linear Attention to Focus on Long Sequences
Zicheng Liu, Siyuan Li, Li Wang, Zedong Wang, Yunfan Liu, Stan Z. Li International Conference on Machine Learning (ICML), 2024 arXiv / project Short-long convolutions with linear attention for long sequence modeling. |
|
|
OpenSTL: A Comprehensive Benchmark of Spatio-Temporal Predictive Learning
Cheng Tan, Siyuan Li, Zhangyang Gao, Wenfei Guan, Zedong Wang, Zicheng Liu, Lirong Wu, Stan Z. Li Annual Conference on Neural Information Processing Systems (NeurIPS), 2023 arXiv / project A comprehensive benchmark of spatio-temporal predictive learning. |
|
OpenMixup: Open Mixup Toolbox and Benchmark for Visual Representation Learning
Siyuan Li*, Zedong Wang*, Zicheng Liu, Di Wu, Stan Z. Li arXiv preprint, 2022 arXiv / project Open source mixup augmentation benchmark for visual representation learning. |
Awards and RecognitionsAwards:
Recognitions: |
ServicesConference Reviewer / PC Member:
Journal Reviewer:
Membership:
Teaching Assistant:
|
AcknowledgementMy research cannot be possible without the support from my awesome mentors, collaborators, and friends:
Beyond academia, I feel incredibly fortunate to have met wonderful friends along the way (particularly during my middle and high school years in Shenzhen and the two years visiting in Hangzhou). I am deeply grateful for our shared moments and wish you all the best. Whether we are still in touch or not, my door is always open for a chat — coffee's on me! |
MiscellaneaBeyond research, I enjoy going on runs. Feel free to follow me on Strava. |
|
Last updated Apr. 2026. Template from Jon Barron. |