I am an undergraduate student in Computer Science and Technology at Shanghai Jiao Tong University. I will join the Department of Computer Science and Engineering at The Chinese University of Hong Kong as a Ph.D. student in Fall 2026, advised by Prof. Yu Cheng.

My research interests include large language models, vision-language models, multimodal agent systems, and self-evolving agents.

🔥 News

• 2026.08: I will join The Chinese University of Hong Kong as a Ph.D. student in Computer Science and Engineering, advised by Prof. Yu Cheng.
• 2026.05: Our paper Persistent Visual Memory was released as an arXiv preprint.
• 2026.05:  🎉🎉 Our paper DiffThinker was accepted to ICML 2026.
• 2026.03: Our paper GEMS was released as an arXiv preprint.
• 2026.02:  🎉🎉 Our paper VideoSSR was accepted to CVPR 2026.
• 2026.02: Our paper LatentMem was released as an arXiv preprint.
• 2026.01:  🎉🎉 Our paper on token perception for multimodal reinforcement learning was accepted to ICLR 2026.
• 2026.01:  🎉🎉 Our paper FrameThinker was accepted to ICLR 2026.
• 2025.09:  🎉🎉 Our paper on sampling-efficient test-time scaling was accepted to NeurIPS 2025 as a Spotlight.

📝 Publications

• Persistent Visual Memory: Sustaining Perception for Deep Generation in LVLMs
  Siyuan Huang, Xiaoye Qu, Yafu Li, Tong Zhu, Zefeng He, Muxin Fu, Daizong Liu, Wei-Long Zheng, and Yu Cheng. arXiv preprint, 2026.
  Code

• GEMS: Agent-Native Multimodal Generation with Memory and Skills
  Zefeng He*, Siyuan Huang*, Xiaoye Qu, Yafu Li, Tong Zhu, Yu Cheng, and Yang Yang. arXiv preprint, 2026. *Equal contribution.
  Project | Code

• LatentMem: Customizing Latent Memory for Multi-Agent Systems
  Muxin Fu, Xiangyuan Xue, Yafu Li, Zefeng He, Siyuan Huang, Xiaoye Qu, Yu Cheng, and Yang Yang. arXiv preprint, 2026.
  Code

• DiffThinker: Towards Generative Multimodal Reasoning with Diffusion Models
  Zefeng He, Xiaoye Qu, Yafu Li, Tong Zhu, Siyuan Huang, and Yu Cheng. ICML 2026.
  Project | Code

• VideoSSR: Video Self-Supervised Reinforcement Learning
  Zefeng He, Xiaoye Qu, Yafu Li, Siyuan Huang, Daizong Liu, and Yu Cheng. CVPR 2026.
  Code

• Spotlight on Token Perception for Multimodal Reinforcement Learning
  Siyuan Huang, Xiaoye Qu, Yafu Li, Yun Luo, Zefeng He, Daizong Liu, and Yu Cheng. ICLR 2026.
  Code

• FrameThinker: Learning to Think with Long Videos via Multi-Turn Frame Spotlighting
  Zefeng He, Xiaoye Qu, Yafu Li, Siyuan Huang, Daizong Liu, and Yu Cheng. ICLR 2026.
  Code

• Sampling-Efficient Test-Time Scaling: Self-Estimating the Best-of-N Sampling in Early Decoding
  Yiming Wang, Pei Zhang, Siyuan Huang, Baosong Yang, Zhuosheng Zhang, Fei Huang, and Rui Wang. NeurIPS 2025 Spotlight.
  Code

🎖 Honors and Awards

• 2026 Outstanding Graduates of Shanghai
• 2024 - 2025 National Scholarship
• 2023 - 2024 National Scholarship

📖 Educations

• 2026 - Present Ph.D. student, Computer Science and Engineering, The Chinese University of Hong Kong. Advisor: Prof. Yu Cheng.
• 2022 - 2026 B.S., Computer Science and Technology, Shanghai Jiao Tong University.

💻 Internships

• Apr. 2026 - Present Research Intern, Qwen Business Unit, Alibaba.
• Jun. 2025 - Apr. 2026 Research Intern, Center for Advanced Exploration, Shanghai Artificial Intelligence Laboratory.