I am a Ph.D. candidate at Huazhong University of Science and Technology (HUST), supervised by Prof. Xinggang Wang and Prof. Wenyu Liu.

My research interests include foundation models, and visual representation learning.

Currently I am interning at ByteDance Doubao Team. Previously, I worked at Vivo AI Laboratory, ByteDance Doubao Vision Group (mentored by Dr. Jiashi Feng & Dr. Zilong Huang), and Beijing Academy of Artificial Intelligence (mentored by Dr. Xinlong Wang).

[Email] [Blog] [Google Scholar] [GitHub] [X]

Education

• Ph.D. Candidate in Computer Vision & Deep Learning, HUST, Sep. 2023 – Dec. 2027
• M.S. in Computer Vision & Deep Learning, HUST, Sep. 2021 – Jun. 2023
• B.Eng. in Information Engineering, HUST, Sep. 2016 – Jun. 2021 (Ranking: 4/28 in Key Class)

Latest Posts

• The Second Half of Model Architecture (Apr 2026)
  We spent a decade scaling computation inside layers. We forgot to scale communication between them. That’s about to change.

Selected Publications

• MoDA: Mixture-of-Depths Attention
  Lianghui Zhu, Yuxin Fang, Bencheng Liao, Shijie Wang, Tianheng Cheng, Zilong Huang, Chen Chen, Lai Wei, Yutao Zeng, Ya Wang, Yi Lin, Yu Li, Xinggang Wang
  arXiv 2026 · Paper & Code

• Vision Mamba: Efficient Visual Representation Learning with Bidirectional State Space Model
  Lianghui Zhu*, Bencheng Liao*, Qian Zhang, Xinlong Wang, Wenyu Liu, Xinggang Wang
  ICML 2024 Most Influential Paper (Rank 2nd, Citation 1st) · 3144 citations · 3.8k GitHub stars · Paper & Code

• JudgeLM: Fine-tuned Large Language Models are Scalable Judges
  Lianghui Zhu, Xinggang Wang, Xinlong Wang
  ICLR 2025 Spotlight (Top 3%) · 321 citations · 425 GitHub stars · Paper & Code

• LENS: Learning to Segment Anything with Unified Reinforced Reasoning
  Lianghui Zhu*, Bin Ouyang*, Yuxuan Zhang, Tianheng Cheng, Rui Hu, Haocheng Shen, Longjin Ran, Xiaoxin Chen, Li Yu, Wenyu Liu, Xinggang Wang
  AAAI 2026 Oral (Top 3.5%) · Paper & Code

• DiG: Scalable and Efficient Diffusion Models with Gated Linear Attention
  Lianghui Zhu, Zilong Huang, Bencheng Liao, Jun Hao Liew, Hanshu Yan, Jiashi Feng, Xinggang Wang
  CVPR 2025 · 48 citations · Paper & Code

• WeakTr: Exploring Plain Vision Transformer for Weakly-supervised Semantic Segmentation
  Lianghui Zhu, Yingyue Li, Jiemin Fang, Yan Liu, Hao Xin, Wenyu Liu, Xinggang Wang
  TIP 2026 · 76 citations · 140 GitHub stars · Paper & Code

• WeakCLIP: Weakly-supervised Semantic Segmentation with Prompt Learning
  Lianghui Zhu, Xinggang Wang, Jiapei Feng, Yingyue Li, Dingwen Zhang, Junwei Han
  IJCV 2024 · 45 citations · Paper & Code

• GroundingSuite: Measuring Complex Multi-Granular Pixel Grounding
  Rui Hu*, Lianghui Zhu*, Yuxuan Zhang, Tianheng Cheng, Longjin Liu, Hao Liu, Longjin Ran, Xiaoxin Chen, Wenyu Liu, Xinggang Wang
  ICCV 2025 · Paper & Code

• WeakSAM: Segment Anything Meets Weakly-supervised Instance-level Recognition
  Lianghui Zhu*, Junwei Zhou*, Yan Liu, Xin Hao, Wenyu Liu, Xinggang Wang
  ACM MM 2024 · 28 citations · Paper & Code

• ViG: Linear-complexity Visual Sequence Learning with Gated Linear Attention
  Bencheng Liao, Xinggang Wang, Lianghui Zhu, Qian Zhang, Chang Huang
  AAAI 2025 · 15 citations · Paper & Code

Awards & Honors

• Basic Research Program for Young Students (Ph.D. Candidates), NSFC, China, 2025
• Young Talent Support Project Special Program for Doctoral Students, CAST, China, 2025
• National Scholarships, Ministry of Education, China, 2024 & 2025
• Academic Rising Star, HUST (rank 1st, only 10 in the whole school), 2025
• Science, Technology and Innovation Scholarship, Ministry of Education, China, 2025

Service

Reviewer for TPAMI, TIP, CVPR, ICCV, ICML, NeurIPS, ICLR

(last updated: Apr 2025)