杨佳澍的照片

杨佳澍 (Jiashu Yang)

北京, 中国

Email

Google Scholar

Zhihu Icon Zhihu

Hello, I'm Jiashu Yang (杨佳澍), My research interests include open-world visual understanding and large models (language, vision, and action). I am committed to the research of practical and deployable intelligent algorithms. In the field of large language models, I led and open-sourced Wenyuan Pavilion, a community focused on developing Chinese-culture-centered open-source LLMs.

If you are interested in discussing or collaborating with me, please feel free to contact me via email.

2023.11 - 2025.09: Developed an eyeball system for embodied vision.

2024.05 - 2025.01: Research on multimodal document understanding with Prof. Huchuan Lu & Prof. Xu Jia.

📝 Publications

  • EvaGaussians: Event Stream Assisted Gaussian Splatting from Blurry Images
    W Yu, C Feng, J Tang, Jiashu Yang, X Jia, L Yuan, Y Tian
    ICCV 2025 | Paper
  • Kongzi: A Historical Large Language Model with Fact Enhancement
    J Yang, N Wang, Y Zhao, C Feng, J Du, H Pang, Z Fang, X Cheng
    COLM 2026 (Under Review) | Paper
  • Tune-Your-Style: Intensity-tunable 3D Style Transfer with Gaussian Splatting
    Y Zhao, R Ye, R Zheng, Z Cheng, C Feng, J Yang, P Qiao, C Liu, J Chen
    ICCV 2025 | Paper
  • Breaking the Vicious Cycle: Coherent 3D-GS from Sparse & Blurred Views
    Z Xu, C Feng, Y Li, J Zhao, J Yang, W Yu, L Yuan, Y Tian
    TCSVT 2025 (Under Review) | Paper
  • Look, Zoom, Understand: The Robotic Eyeball for Embodied Perception
    J Yang, Y Han, Y Xie, N Guo, W Lian
    ACMMM 2026 (Under Review) | Paper
  • AutothinkRAG: Complexity-Aware Control of Retrieval-Augmented Reasoning for Image-Text Interaction
    J Yang, C Zhang, A Wuerkaixi, X Cheng, C Liu, K Zeng, X Jia, X Cai
    ACL 2026 (Under Review) | Paper
  • Prospective and intelligent antiviral drug design against potential viral variants
    Bioinformatics (Under Review)
  • A survey of document understanding
    Project Leader
  • A survey of activate perception
    Project Leader

📖 Education

💻 Projects

  • 📂 Our multi-modal PDF interaction tool "Squirrel" will be launched on January 15, 2025!
  • Wenyuan Pavilion Logo
    Wenyuan Pavilion: Ancient Chinese Language Community
    I lead a research community focused on domain-specific models for Ancient Chinese and Literature. We have open-sourced various models and datasets to support scientific research and project implementation.

💼 Work Experience

  • Institute of Automation Logo
    2023.11 - 2024.08
    Institute of Automation, Chinese Academy of Sciences (CAS)
  • Dalian University of Technology Logo
    2024.05 - 2025.01
    IIAU lab, Research Assistant
  • ByteDance Logo
    2025.04 - 2025.07
    ByteDance, Beijing, Applications of Large Language Models
  • Shanghai Jiao Tong University Logo
    2025.07 - 2025.09
    IIAU lab, School of Artificial Intelligence, Shanghai Jiao Tong University,Research Assistant
  • Meituan Logo
    2025.12 - now
    Meituan, Beijing, Longcat Interaction

🎖 Honors and Awards

About Life

🚧 This section is under construction... 🚧