About me
I am a third-year PhD student at UC San Diego Department of CSE, advised by Prof. Rose Yu. My current research interest lies in generative AI, physics-guided deep learning, and geometric deep learning. I am particularly excited about machine learning methods for scientific discovery, such as symmetry discovery and governing equation discovery. I was a research intern at Bytedance during 2024 summer and, prior to that, a research intern at NEC Laboratories during 2023 summer. For more details, see my cv.
News
- Mar 2025: Invited talk at Boston Symmetry Day
- Nov 2024: Invited talk about symmetry and equation discovery at AAAI Fall Symposium
- September 2024: Symmetry-Informed Governing Equation Discovery accepted at NeurIPS 2024!
- April 2024: Latent Space Symmetry Discovery accepted at ICML 2024!
- October 2023: Invited talk about symmetry discovery at UMich SciML Webinar
- June 2023: Interning at NEC Laboratories Princeton
- April 2023: LieGAN accepted at ICML 2023!
Publications
Discovering Symbolic Differential Equations with Symmetry Invariants
Jianke Yang, Manu Bhat, Bryan Hu, Yadi Cao, Nima Dehmamy, Robin Walters, and Rose Yu.
arxiv, 2025
Symmetry-Informed Governing Equation Discovery
Jianke Yang, Wang Rao, Nima Dehmamy, Robin Walters, and Rose Yu.
NeurIPS, 2024
Latent Space Symmetry Discovery
Jianke Yang, Nima Dehmamy, Robin Walters, and Rose Yu.
International Conference on Machine Learning (ICML), 2023
Paper | Workshop Paper | Code | BibTeX
Generative Adversarial Symmetry Discovery
Jianke Yang, Robin Walters, Nima Dehmamy, Rose Yu.
International Conference on Machine Learning (ICML), 2023
Experiences
Bytedance | San Jose, CA
- (06/2024 - 09/2024) Research Intern
- Latent diffusion model for protein dynamics generation
NEC Laboratories | Princeton, NJ
- (06/2023 - 09/2023) Research Intern
- Diffusion model for 3D human motion generation
- Incorporate physics simulator into training to perform physically valid generation
USC Melady Lab | Los Angeles, CA
- (06/2021 - 09/2021) Research Assistant
- Federated learning under distribution shift