Jiazi Bu 卜家梓

  PhD (NTU MMLab & SJTU, in CS), BEng (SJTU, in AI)

I am a first-year joint Ph.D. student at Nanyang Technological University and Shanghai Jiao Tong University, co-advised by Prof. Dahua Lin and Prof. Xingang Pan. Prior to that, I received the B.Eng. degree with honors in Artificial Intelligence (the 1st session of GuoZhi Class Programme, held by Prof. Xiaoou Tang and Prof. Cewu Lu), Shanghai Jiao Tong University. Moreover, I have also worked with Prof. Junchi Yan at SJTU-ReThinkLab.

My current research focuses on Generative AI and VLM. I work closely with Dr. Yuhang Zang, Dr. Tong Wu and Dr. Jiaqi Wang.

Looking for collaborations. Please contact me via E-mail.

bujiazi001@sjtu.edu.cn  /  CV (26.06)  /  Scholar  /  GitHub

profile photo
🔥 Recent News

2026-04: I accept the offer of Joint Ph.D. Degree Programme at Nanyang Technological University (NTU).

2026-03: 1 paper, Uni-Classifier, is accepted by ICME 2026 (as Co-First Author).

2026-02: 1 paper, Granular-GRPO, is accepted by CVPR 2026 (as Co-First Author).

2026-01: 1 paper, DiCache, is accepted by ICLR 2026 (as First Author).

2025-09: 1 paper, HiFlow, is accepted by NeurIPS 2025 (as First Author).

2025-09: I transition directly into the CS Ph.D. Programme at SJTU, advised by Prof. Dahua Lin.

2025-06: My Bachelor's Thesis is awarded the Best Bachelor's Thesis of SJTU (TOP 1%).

2025-06: I receive an additional Zhiyuan Honor Bachelor's Degree from SJTU.

2025-06: I receive my B.Eng. Degree from SJTU (as one of TOP 30 Outstanding Graduates).

2025-06: 1 paper, Light-A-Video, is accepted by ICCV 2025 (as Co-First Author).

2025-02: 1 paper, ByTheWay, is accepted by CVPR 2025 (as First Author).

2025-01: 1 paper, MotionClone, is accepted by ICLR 2025 (as Co-First Author).

2025-01: 1 paper, HierDDI, is accepted by SCIENCE CHINA Information Sciences.

2024-09: 1 paper, Unveiling The Matthew Effect Across Channels, is accepted by NeurIPS 2024.

2023-11: I begin the internship at Shanghai AI Laboratory, working with Dr. Jiaqi Wang.

2023-11: 1 paper, ViTree, is accepted by AAAI 2024.

2023-11: I receive National First Prize & First Prize in Shanghai Chapter in CUMCM 2023.

2022-11: I become a member of SJTU-RethinkLab, advised by Prof. Junchi Yan.

2021-12: I am enrolled in the first session of Artificial Intelligence GuoZhi Class Programme of SJTU.

2021-09: I am enrolled in the Zhiyuan Honors Programme of SJTU.

2021-08: I enter Shanghai Jiao Tong University (SJTU) as an Undergraduate student, majoring in AI.
🥇 (Co-) First Author Publications
dicache
DiCache: Let Diffusion Model Determine Its Own Cache
Jiazi Bu*, Pengyang Ling*, Yujie Zhou*, Yibin Wang, Yuhang Zang, Dahua Lin, Jiaqi Wang
ICLR 2026
Page / Paper / Code
hiflow
HiFlow: Training-free High-Resolution Image Generation with Flow-Aligned Guidance
Jiazi Bu*, Pengyang Ling*, Yujie Zhou*, Pan Zhang, Tong Wu, Xiaoyi Dong, Yuhang Zang, Yuhang Cao, Dahua Lin, Jiaqi Wang
NeurIPS 2025
Page / Paper / Code / Proceedings
bytheway
ByTheWay: Boost Your Text-to-Video Generation Model to Higher Quality in a Training-free Way
Jiazi Bu*, Pengyang Ling*, Pan Zhang, Tong Wu, Xiaoyi Dong, Yuhang Zang, Yuhang Cao, Dahua Lin, Jiaqi Wang
CVPR 2025
Page / Paper / Code / Proceedings
mvgrpo
From Sparse to Dense: Multi-View GRPO for Flow Models via Augmented Condition Space
Jiazi Bu*, Pengyang Ling*, Yujie Zhou*, Yibin Wang, Yuhang Zang, Tianyi Wei, Xiaohang Zhan, Jiaqi Wang, Tong Wu, Xingang Pan, Dahua Lin
arXiv 2026
Page / Paper / Code (Under Review)
adagrpo
AdaGRPO: A Capability-Aware Adaptive Enhancement for Flow-based GRPO
Jiazi Bu*, Pengyang Ling*, Yujie Zhou*, Yibin Wang, Yuhang Zang, Tianyi Wei, Xiaohang Zhan, Jiaqi Wang, Tong Wu, Xingang Pan, Dahua Lin
arXiv 2026
Page / Paper / Code (Under Review)
motionclone
MotionClone: Training-Free Motion Cloning for Controllable Video Generation
Pengyang Ling*, Jiazi Bu*, Pan Zhang, Xiaoyi Dong, Yuhang Zang, Tong Wu, Huaian Chen, Jiaqi Wang, Yi Jin
ICLR 2025
Page / Paper / Code / Proceedings
light-a-video
Light-A-Video: Training-free Video Relighting via Progressive Light Fusion
Yujie Zhou*, Jiazi Bu*, Pengyang Ling*, Pan Zhang, Tong Wu, Qidong Huang, Jinsong Li, Xiaoyi Dong, Yuhang Zang, Yuhang Cao, Anyi Rao, Jiaqi Wang, Li Niu
ICCV 2025
Page / Paper / Code / Proceedings
g2rpo
Fine-Grained GRPO for Precise Preference Alignment in Flow Models
Yujie Zhou*, Pengyang Ling*, Jiazi Bu*, Yibin Wang, Yuhang Zang, Jiaqi Wang, Li Niu, Guangtao Zhai
CVPR 2026
Page / Paper / Code / Proceedings
uni-c
Uni-Classifier: Leveraging Video Diffusion Priors for Universal Guidance Classifier
Yujie Zhou*, Pengyang Ling*, Jiazi Bu*, Bingjie Gao, Li Niu
ICME 2026
Paper / Code (Under Review)
pavegrpo
Pave-GRPO: Beyond Instantaneous Guidance through Principled Average Velocity Decomposition
Pengyang Ling*, Jiazi Bu*, Yujie Zhou*, Yibin Wang, Zhenyu Hu, Zihan Zhang, Yi Jin, Huaian Chen, Yuhang Zang
arXiv 2026
Paper / Code (Under Review)
🥈 Co-Author Publications
mathew
Unveiling The Matthew Effect Across Channels: Assessing Layer Width Sufficiency via Weight Norm Variance
Yiting Chen, Jiazi Bu, Junchi Yan
NeurIPS 2024
Paper / Code / Proceedings
vitree
ViTree: Single-Path Neural Tree for Step-Wise Interpretable Fine-Grained Visual Categorization
Danning Lao, Qi Liu, Jiazi Bu, Junchi Yan, Wei Shen
AAAI 2024
Paper / Code / Proceedings
hierddi
Drug-drug Interaction Prediction via Hierarchical Structure Modeling
Huaijin Wu, Jiazi Bu, Nianzu Yang, Yao Sun, Haitao Song, Ning Liu, Junchi Yan
SCIENCE CHINA Information Sciences
Paper / Code / Proceedings
unigenbench++
Unified Personalized Reward Model for Vision Generation
Yibin Wang, Yuhang Zang, Feng Han, Jiazi Bu, Yujie Zhou, Cheng Jin, Jiaqi Wang
arXiv 2026
Page / Paper / Code
unigenbench++
UniGenBench++: A Unified Semantic Evaluation Benchmark for Text-to-Image Generation
Yibin Wang*, Zhimin Li*, Yuhang Zang*, Jiazi Bu, Yujie Zhou, Chunyu Wang, Qinglin Lu, Cheng Jin, Jiaqi Wang
arXiv 2025
Page / Paper / Code
pref-grpo
Pref-GRPO: Pairwise Preference Reward-based GRPO for Stable Text-to-Image Reinforcement Learning
Yibin Wang*, Zhimin Li*, Yuhang Zang*, Yujie Zhou, Jiazi Bu, Chunyu Wang, Qinglin Lu, Cheng Jin, Jiaqi Wang
arXiv 2025
Page / Paper / Code
endocot
EndoCoT: Scaling Endogenous Chain-of-Thought Reasoning in Diffusion Models
Xuanlang Dai, Yujie Zhou, Long Xing, Jiazi Bu, Xilin Wei, Yuhong Liu, Beichen Zhang, Kai Chen, Yuhang Zang
arXiv 2026
Page / Paper / Code
caprl++
CapRL++: Unified Reinforcement Learning with Verifiable Rewards for Dense Image and Video Captioning
Penghui Yang, Long Xing, Xiaoyi Dong, Yuhang Zang, Yuhang Cao, Yibin Wang, Yujie Zhou, Jiazi Bu, Jianze Liang, Qidong Huang, Jiaqi Wang, Feng Wu, Dahua Lin
arXiv 2026
Paper / Code
🥉 Technical Reports
metascript
MetaScript: Few-Shot Handwritten Chinese Content Generation via Generative Adversarial Networks
Jiazi Bu*, Qirui Li*, Kailing Wang*, Xiangyuan Xue*, Zhiyuan Zhang*
arXiv 2023, Project of AI3604 Computer Vision, 2023 Fall, SJTU
Report / Code
🏆 Honors and Awards
  • 2025: Ye Jun and Shen Nanpeng Outstanding Scholarship (Top 30 Graduates of SJTU)
  • 2022: Guozhi Scholarship
  • 2021~2024: Zhiyuan Scholarship × 4
  • 2022~2025: SJTU Undergraduate Scholarship × 4
  • 2026: SJTU Postgraduate Scholarship for PhD student
  • 2023: CUMCM National First Prize (Top 0.55%) & First Prize in Shanghai Chapter
  • 2025: SJTU Zhiyuan Honor Bachelor's Degree
  • 2023: SJTU President Award
  • 2021: SJTU Outstanding Freshman Training Camp Member
  • 2025: SJTU Outstanding Graduate
  • 2025: SJTU Outstanding Bachelor's Thesis of School of Computer Science
  • 2025: SJTU Best Bachelor's Thesis Award (Top 1%)
🖥️ Selected Services

Academic Reviewer:

  • Conference Reviewer: ECCV (2026), CVPR (2026), ICLR (2026), COLM (2026).
  • Journal Reviewer: T-PAMI, TMLR.

Teaching Assistant:

  • TBD

Volunteer:

  • Student volunteer (for 55 hours) during the COVID-19 pandemic.
  • Student volunteer for CUMCM 2024.