Fu-Yun Wang
Fu-Yun Wang
Ph.D. Candidate @ MMLab, CUHK

Fu-Yun Wang (IPA: [fu˧˥ yn˧˥ wɑŋ]) is a third-year Ph.D. candidate at MMLab@CUHK. He works on scalable post-training. Currently vibing with world model and agentic stuff.

Job market 2027. Looking for opportunities (industry or postdoc) to work on interesting projects. Always happy to chat early.

Recent Work New

PromptRL: Prompt Matters in RL for Flow-Based Image Generation
Fu-Yun Wang, Han Zhang, Michael Gharbi, Hongsheng Li, Taesung Park
arXiv 2026
Key Insight: Achieves 2× sample efficiency over flow-only RL while obtaining an adaptive prompt refinement agent that improves test-time performance.

Method

PromptRL jointly trains language models and flow-matching models in a unified RL loop for text-to-image generation, using LMs as adaptive prompt refiners.

  • Unified RL Training: Joint optimization of both LM and FM in a single loop
  • Exploration Collapse Solution: Overcomes insufficient diversity through adaptive prompting
  • Test-time Adaptation: Adaptive prompt refinement at inference
Project arXiv GitHub
Image Diffusion Preview with Consistency Solver
Fu-Yun Wang¹², Hao Zhou¹, Liangzhe Yuan¹, Sanghyun Woo¹, Boqing Gong¹, Bohyung Han¹, Ming-Hsuan Yang¹, Han Zhang¹, Yukun Zhu¹, Ting Liu¹, Long Zhao¹ · ¹Google DeepMind ²CUHK
arXiv 2026
Key Insight: Proposes a preview-and-refine paradigm for diffusion models: users first get fast low-step previews, then only run expensive full-step sampling on satisfactory ones—reducing overall interaction time by nearly 50%.

Method

ConsistencySolver is a lightweight, learnable high-order ODE solver derived from general linear multistep methods, optimized via Reinforcement Learning (PPO). It adapts integration coefficients to each sampling step, improving both preview fidelity and consistency with the full-step output—without modifying the base diffusion model.

  • Diffusion Preview Paradigm: Fast few-step preview → user evaluation → full-step refinement only when satisfied
  • Learnable ODE Solver: Adaptive multistep coefficients predicted by a lightweight MLP, trained end-to-end via RL
  • Preview Consistency: Preserves the deterministic PF-ODE mapping so preview closely matches the final output
  • FID on-par with 47% fewer steps vs. Multistep DPM-Solver; outperforms distillation baselines
arXiv GitHub

Research Summary

Interactive tree diagram of my research. Click nodes to expand/collapse; click paper titles to visit links.

Research Directions

Internship Experience

ByteDance Seed Current

Research Intern · 2025.10 - Present

Video Generation, Multimodal Models

Mentor: Haoqi Fan

Reve Inc

Research Intern · 2025.6 - 2025.11

Multimodal LMs, Diffusion Models, RL

Supervised by: Dr. Han Zhang

Google DeepMind

Research Intern · 2025.2 - 2025.5

Diffusion Distillation, RL

Supervised by: Dr. Long Zhao, Dr. Ting Liu, Dr. Hao Zhou, Dr. LiangZhe Yuan

Collaborated with Prof. Bohyung Han, Prof. Boqing Gong

Avolution AI Acquired by MiniMax

Research Collaboration · 2023.10 - 2024.10

Video Diffusion, Distillation

Collaborated with: Dr. Zhaoyang Huang, Dr. Xiaoyu Shi, Weikang Bian

Tencent AI Lab

Research Intern · 2022.6 - 2022.12

Class-Incremental Learning

Supervised by: Dr. Liu Liu, Prof. Yatao Bian

Education

The Chinese University of Hong Kong

Ph.D. in Engineering · 2023 - Present

Supervisor: Prof. Hongsheng Li & Prof. Xiaogang Wang

Nanjing University

B.Eng. in AI (Rank 2/88) · 2019 - 2023

Supervisor: Prof. Han-Jia Ye & Prof. Da-Wei Zhou

Selected Publications

Categorized by theme. Full list on Google Scholar.

Diffusion Post-Training: Acceleration & Reinforcement Learning

Rectified Diffusion
Rectified Diffusion: Straightness Is Not Your Need in Rectified Flow
Fu-Yun Wang, Ling Yang, Zhaoyang Huang, Mengdi Wang, Hongsheng Li
ICLR 2025
In-depth theoretical analysis and empirical validation of flow matching, rectified flow, and the rectification operation. ZHIHU blog garnered 10k+ views and ~400 likes.
arXiv GitHub Poster
PCM
Phased Consistency Model
Fu-Yun Wang, Zhaoyang Huang, Alexander William Bergman, Dazhong Shen, Peng Gao, Michael Lingelbach, Keqiang Sun, Weikang Bian, Guanglu Song, Yu Liu, Xiaogang Wang, Hongsheng Li
NeurIPS 2024
Validate the feasibility of one-step video generation; Adopted by Qwen-Image-2512 for first-stage initialization; Adopted by Early version of FastVideo, accelerating HunyuanVideo and WAN.
Project GitHub Paper Poster
Diffusion-NPO
Diffusion-NPO: Negative Preference Optimization for Diffusion Models
Fu-Yun Wang, Yunhao Shui, Jingtan Piao, Keqiang Sun, Hongsheng Li
ICLR 2025
A general, simple yet effective method for strengthened diffusion preference optimization.
Paper Poster GitHub

Generative Vision Applications

Motion-I2V: Consistent and Controllable Image-to-Video Generation
Xiaoyu Shi*, Zhaoyang Huang*, Fu-Yun Wang*, Weikang Bian*, et al.
SIGGRAPH 2024 Technical Papers Trailer
Project GitHub arXiv
ZoLA: Zero-Shot Creative Long Animation Generation
Fu-Yun Wang, Zhaoyang Huang, Qiang Ma, Xudong Lu, Weikang Bian, Yijin Li, Yu Liu, Hongsheng Li
ECCV 2024 Oral
Project Paper

Class-Incremental Learning

PyCIL
PyCIL: A Python Toolbox for Class-Incremental Learning
Da-Wei Zhou*, Fu-Yun Wang*, Han-Jia Ye, De-Chuan Zhan
SCIENCE CHINA Information Sciences
Nearly 1000 GitHub stars — the most widely collected CIL toolkit.
GitHub arXiv
FOSTER
FOSTER: Feature Boosting and Compression for Class-Incremental Learning
Fu-Yun Wang, Da-Wei Zhou, Han-Jia Ye, De-Chuan Zhan
ECCV 2022
GitHub arXiv

Talks

Awards & Honor

  • 2025 CVPR 2025 Outstanding Reviewer
  • 2023 HKPFS (Hong Kong PhD Fellowship Scheme)
  • 2023 Outstanding Graduate of Nanjing University
  • 2023 Outstanding Undergraduate Thesis of NJU
  • 2022 Sensetime Scholarship
  • 2022 Huawei Scholarship
  • 2021 National Scholarship

Academic Services

Reviewer for:

VenueYears
TPAMI—
TCSVT / PRL—
CVPR2023–2025
NeurIPS2023, 2025
ICLR2024, 2025
ICML2024, 2025
ECCV / ICCV2024 / 2025
BMVC / SIGGRAPH Asia2024 / 2025