About Me
lead you to my latest work!Hello! I am Zhenglin Cheng, a third-year Ph.D. student of 
LINs lab, Westlake University (through joint program with ZJU), advised by Prof. Tao LIN.
Before that, I received my bachelor’s degree in Software Engineering from Zhejiang University (ZJU).
📄 Find my CV here (Jan 2026 Update).
News
- 2026/09, We released the LLaDA-Image Base and Turbo checkpoints, have fun!
- 2026/01, 🥳 TwinFlow is accepted to ICLR’26, see you in Rio de Janeiro, Brazil 🇧🇷 !
- 2025/12, 🚀 We release TwinFlow, a simple and effective framework for one-step generation!
- 2025/01, 🥳 Dynamic Mixture of Experts (DynMoE) is accepted to ICLR’25, see you in Singapore 🇸🇬 !
Research Interests
My long-term research goal is to build efficient multimodal agents that can understand the physical world, reason on real-world problems, and generate novel ideas, which could also learn from experience and evolve themselves in the constantly changing environment.
Looking at the present, I put my focus on:
- Interactive video/world models: how can we make video diffusion models strong and realtime interactive entrypoints for real-world users?
- Few-step generation: how can we effectively train/distill continuous diffusion generators into 1-step or few-step generators?
Publications/Manuscripts (* denotes equal contribution; denotes core contributors)
📖 LLaDA-Image: Building Strong Image Generators with Fully Open Training Recipes
Chuyan Chen, Haoxing Chen, Kun Chen, Zhenglin Cheng, Long Cui, Ruishan Fang, Zhangxuan Gu, Zhicheng Huang, Zhenzhong Lan, Yuanting Lei, Haoquan Li, Jianguo Li, Rongchuan Li, Sidu Li, Tao Lin, Deyuan Liu, Jiacheng Liu, Lin Liu, Yuxuan Lou, Zhisheng Lu, Yuxin Ma, Shuheng Shen, Peng Sun, Chaoyang Wang, Hongjun Wang, Xiaomei Wang, Yongxin Wang, Chengzhang Wu, Hongru Wu, Jun Xie
👉 LLaDA-Image is a 6B unified diffusion model family for high-quality image generation and editing. It supports text-to-image and VQ-conditioned generation, instruction-guided editing, and Chinese–English text rendering; its Turbo variant enables fast generation and editing in only 2–4 sampling steps.
📖 TwinFlow: Realizing One-step Generation on Large Models with Self-adversarial Flows
Zhenglin Cheng*, Peng Sun*, Jianguo Li, Tao Lin
👉 TwinFlow tames large-scale few-step training through self-adversarial flows, eliminating the need for any auxiliary networks (discriminators, teachers, fake scores) by one-model design. This scalable approach transforms Qwen-Image-20B into a high-quality few-step generator.
📖 Dynamic Mixture of Experts: An Auto-Tuning Approach for Efficient Transformer Models
Yongxin Guo*, Zhenglin Cheng*, Xiaoying Tang, Zhaopeng Tu, Tao Lin
👉 DynMoE frees the burden of pivotal hyper-parameter selection for MoE training by enabling each token to activate different number of experts, and adjusting the number of experts automatically, acheiving stronger sparsity well maintaining performance.
Experiences
- 2026/09 - Present, Tencent.
- 2025/07 - 2026/09, inclusionAI, Ant Group (Tech Leader: Dr. Jianguo Li).
Academic Services
- Conference Reviewer: ICLR, ICML, NeurIPS.
Educations
- 2024/09 - 2029/06, Westlake University, College of Engineering.
- 2020/09 - 2024/06, Zhejiang University, College of Computer Science and Technology.