Chenyu Huang <br> (黄辰宇)

Chenyu Huang
(黄辰宇)

Fudan University

About Me

I am currently a Ph.D. student (Sep. 2024 - Jun. 2028, expected) in the School of Information Science and Technology, Fudan University, supervised by Prof. Tao Chen. I am also fortunate to work closely with Dr. Peng Ye from Shanghai AI Lab. Before this, I obtained my Master's and Bachelor's degrees from the School of Electronic Science and Engineering, Southeast University, supervised by Prof. Yuning Zhang. I’m currently working in the fields of vision-language models, on-policy distillation, multi-modal reasoning, and model merging.
Research Interests
🔗 Model Merging 👁️ Multimodal LLM 📏 On-Policy Distillation 🌐 Vision-Language Model

🔥 News

  • 2026.8:  🎉🎉 DO-Merging is accepted by EMNLP 2026 Findings.

  • 2026.5:  🎉🎉 FRISM is accepted by ICML 2026.

  • 2026.1:  🎉🎉 Sci-Verifier is accepted by ICLR 2026.

  • 2025.9:  🎉🎉 UltraDelta is accepted by NeurIPS 2025.

  • 2025.2:  🎉🎉 DeRS is accepted by CVPR 2025.

  • 2025.2:  🎉🎉 MoW-Merging is accepted by IEEE TCSVT.

  • 2024.9:  🎉🎉 EMR-Merging is accepted by NeurIPS 2024 Spotlight.

📝 Publications

  1. EMR-Merging: Tuning-free high-performance model merging
    C Huang, P Ye, T Chen, T He, X Yue, W Ouyang
    NeurIPS 2024 Spotlight Advances in Neural Information Processing Systems, 2024
  2. FRISM: Fine-Grained Reasoning Injection via Subspace-Level Model Merging for Vision-Language Models
    C Huang, P Ye, X Tan, J Mu, S Zheng, L Shen, T Chen
    ICML 2026 arXiv preprint arXiv:2601.21187
  3. Dynamic model merging with mixture of weights
    P Ye, C Huang, M Shen, T Chen, Y Huang, W Ouyang
    TCSVT 2025 IEEE Transactions on Circuits and Systems for Video Technology, 2025
  4. DeRS: Towards extremely efficient upcycled mixture-of-experts models
    Y Huang, P Ye, C Huang, J Cao, L Zhang, B Li, G Yu, T Chen
    CVPR 2025 Proceedings of the Computer Vision and Pattern Recognition, 2025
  5. Sci-verifier: Scientific verifier with thinking
    S Zheng, C Huang, F Yu, J Yao, J Ye, T Chen, Y Luo, N Ding, L Bai, G Cui, ...
    ICLR 2026 arXiv preprint arXiv:2509.24285
  6. Breaking the compression ceiling: Data-free pipeline for ultra-efficient delta compression
    X Wang, P Ye, C Huang, S Zheng, B Zhang, L Bai, W Ouyang, T Chen
    NeurIPS 2025 Advances in Neural Information Processing Systems 38, 23167-23195
  7. Decouple and Orthogonalize: A Data-Free Framework for LoRA Merging
    Shenghe Zheng, Hongzhi Wang, C Huang, Xiaohui Wang, Tao Chen, Jiayuan Fan, Shuyue Hu, Peng Ye
    EMNLP 2026 Findings arxiv preprint 2505.15875

📖 Educations

  • 2024.09 — Present
    Ph.D. Candidate · College of Future Information and Technology
    Fudan University
    Supervised by: Prof.Tao Chen
  • 2021.09 — 2024.06
    Master of Electronic Science and Technology · School of Electronic Science and Engineering
    Southeast University
    Supervised by: Prof.Yuning Zhang
  • 2017.09 — 2021.06
    Bachelor of Electronic Science and Technology · School of Electronic Science and Engineering
    Southeast University

💬 Invited Talks

  • 2025.04
    EMR-Merging: Tuning-Free High-Performance Model Merging
    ▶ Watch Video

🏆 Honors and Awards

National Scholarship
2023
Scholarship of Suzhou Industrial Park
2022
Merit Master Student
2022
Merit Bachelor Student
2018, 2019

📝 Academic Services

Serving as a reviewer for:

NeurIPS ECCV CVPR ICML IEEE T-CSVT