Neural Networks as Surrogate Solvers for Time-Dependent Accretion Disk Dynamics
Mao, S., et al. (2025). "Neural Networks as Surrogate Solvers for Time-Dependent Accretion Disk Dynamics." The Astrophysical Journal Letters, 992, L20.
CV
Education
- Ph.D. in Astronomy, University of Victoria, 2022–2024
- M.Sc. in Astronomy, University of Victoria, 2019–2022
- B.Sc. in Astronomy, Wang Shouguan Honors Program in Astronomy, University of Science and Technology of China, 2015–2019
Appointments
- Cosmic AI Postdoctoral Fellow, National Radio Astronomy Observatory, July 2026–present
- Postdoctoral Researcher, Department of Physics and Astronomy, Rice University, March 2025–July 2026
- Affiliate, Ken Kennedy Institute for AI and Computing
- Postdoctoral Researcher, Los Alamos National Laboratory, September 2024–February 2025
Research interests
- AI for astronomy
- Radio interferometry
- Physics-informed neural networks and scientific machine learning
- Protoplanetary disks and planet formation
- Astrophysical fluid dynamics
Publications
Disk2Planet: A Robust and Automated Machine Learning Tool for Parameter Inference in Disk–Planet Systems
Mao, S., et al. (2024). "Disk2Planet: A Robust and Automated Machine Learning Tool for Parameter Inference in Disk–Planet Systems." The Astrophysical Journal, 976, 200.
PPDONet: Deep Operator Networks for Fast Prediction of Steady-state Solutions in Disk–Planet Systems
Mao, S., et al. (2023). "PPDONet: Deep Operator Networks for Fast Prediction of Steady-state Solutions in Disk–Planet Systems." The Astrophysical Journal Letters, 950, L12.
DRAGraces: A pipeline for the GRACES high-resolution spectrograph at Gemini
Chené, A.-N., Mao, S., et al. (2021). "DRAGraces: A pipeline for the GRACES high-resolution spectrograph at Gemini." The Astronomical Journal, 161, 109.
Selected talks
Cosmic Horizons Conference
Contributed conference talk at Cosmic Horizons Conference, Charlottesville, Virginia, United States
2026 AstroAI Workshop
Contributed workshop talk at 2026 AstroAI Workshop, Boston, Massachusetts, United States
ML4Disks Workshop
Contributed workshop talk at ML4Disks Workshop, University of Milan, Milan, Italy
2025 Star and Planet Formation in the Southwest Conference
Contributed conference talk at Star and Planet Formation in the Southwest Conference, Austin, Texas, United States
Open SkAI 2025
Contributed conference talk at Open SkAI 2025, Chicago, Illinois, United States
Cosmic Horizons Conference
Contributed conference talk at Cosmic Horizons Conference, Austin, Texas, United States
ET Seminar
Invited seminar at Shanghai Astronomical Observatory, Shanghai, China
13th National Conference on Fluid Mechanics
Contributed conference talk at 13th National Conference on Fluid Mechanics, Harbin, China
Crunch Machine Learning + X Seminars
Invited seminar at Brown University, Providence, Rhode Island, United States
Scientific Machine Learning Workshop
Invited workshop talk at Banff International Research Station for Mathematical Innovation and Discovery (BIRS), Banff, Alberta, Canada
Chinese Congress of Theoretical and Applied Mechanics
Contributed conference talk at Chinese Congress of Theoretical and Applied Mechanics, Chengdu, China
Awards
- National Energy Research Scientific Computing Center DOE Mission Science Allocation Award, 2026–2027 — Accelerating Fluid Dynamics Simulations: A Data-Free, Physics-Informed Deep Learning Framework for Solving Partial Differential Equations (20,000 CPU node hours)
Open-source software
- PPDONet: Deep Operator Networks for fast prediction of solutions in disk–planet systems
- Disk2Planet: automated machine-learning parameter inference for disk–planet systems
- fargo_utils: Python interface for FARGO3D hydrodynamic simulations
Media
- A Rapid Disk Predictor, AAS Nova, October 2023