avatar

Jiachen Yao

Ph.D. Student
Caltech CMS
firstname (at) caltech.edu


About Me

Hello! This is Jiachen. ▶ More about my name… Jiachen is pronounced “chia-ch’en”, and my name is written as Chinese in Chinese. Feel free to call me James!

I’m a Ph.D. student in Computing and Mathematical Sciences at Caltech, where I am fortunate to be advised by Professor Anima Anandkumar. My research lies at the intersection of generative modeling and scientific computing.

Prior to my PhD, I completed my undergraduate studies at Tsinghua University, where I had the privilege of working with Professor Jun Zhu and Professor Hang Su on physics-informed machine learning. I am also grateful for the opportunity to conduct research at Stanford University, where I worked with Professor James Landay and Professor Monica Lam on human-AI interaction.

Research Interests

My long term goal is to develop a unified framework that learns and uses efficient representations of data, domain knowledge, and available observations to model complex systems. I am particularly interested in generative approaches to achieve this objective.

I am interested in the intersection of generative modeling and physical sciences. Scientific problems are often ill-posed, and the strong priors offered by generative models can significantly enhance robustness and efficiency. However, these models typically require vast amounts of data and lack physics guarantee. Therefore, I focus on developing scalable, uncertainty-aware methods for generative models + scientific inverse problems.

News

Publications

  1. NeurIPSW
    Flow-Guided Neural Operator For Self- Supervised Learning On Time Series Data
    Duy Nguyen*, Jiachen Yao*, Jiayun Wang*, Julius Berner, Anima Anandkumar.
    NeurIPS Workshop on Learning from Time Series for Health, 2025.
    PDF

  2. ICMLW
    EquiReg: Equivariance Regularized Diffusion for Inverse Problems
    Bahareh Tolooshams*, Aditi Chandrashekar*, Rayhan Zirvi*, Abbas Mammadov, Jiachen Yao, Chuwei Wang, Anima Anandkumar.
    ICML Workshop on World Models, 2025.
    PDF BibTex

  3. NeurIPS
    FunDPS: Guided Diffusion Sampling on Function Spaces with Applications to PDEs
    Jiachen Yao*, Abbas Mammadov*, Julius Berner, Gavin Kerrigan, Jong Chul Ye, Kamyar Azizzadenesheli, Anima Anandkumar.
    Neural Information Processing Systems (NeurIPS), 2025.
    PDF Code BibTex

  4. NeurIPS
    PINNacle: A Comprehensive Benchmark of PINNs for Solving PDEs
    Zhongkai Hao*, Jiachen Yao*, Chang Su*, Hang Su, Ziao Wang, Fanzhi Lu, Zeyu Xia, Yichi Zhang, Songming Liu, Lu Lu, and Jun Zhu.
    Neural Information Processing Systems (NeurIPS), 2024.
    PDF Code Project Page BibTex

  5. WhisperBuds: Whispered Speech Input with Earbuds
    Jackie Yang*, Jiachen Yao*, Yingtian Shi, Monica Lam, and James Landay.
    PDF Video

  6. Preconditioning for Physics-Informed Neural Networks
    Songming Liu, Chang Su*, Jiachen Yao*, Zhongkai Hao, Hang Su, Youjia Wu, and Jun Zhu.
    PDF BibTex

  7. ICML
    MultiAdam: Parameter-wise Scale-invariant Optimizer for Multiscale Training of Physics-informed Neural Network
    Jiachen Yao*, Chang Su*, Zhongkai Hao, Songming Liu, Hang Su, and Jun Zhu.
    International Conference on Machine Learning (ICML), 2023.
    PDF Code BibTex

  8. CHI
    Selecting Real-World Objects via User-Perspective Phone Occlusion
    Yue Qin, Chun Yu, Wentao Yao, Jiachen Yao, Chen Liang, Yueting Weng, Yukang Yan,and Yuanchun Shi.
    Conference on Human Factors in Computing Systems (CHI), 2023.
    PDF Video BibTex

    9. * denotes equal contribution

Services

As a reviewer

Invited talks

Honors and Awards

Graduate Honors

Undergrad Honors

Personal Interests

Hobbies

Technical Skills


Powered by Jekyll and Minimal Light theme.