Tianyue Yang
I work at the intersection of AI, physics, and chemistry, with a focus on generative modelling for scientific machine learning and physics-inspired methods.
Research Interests
I work on generative modelling and AI for science, especially applications in computational fluid dynamics, chemistry, and physics-informed learning.
Current Role
I am a Research Assistant at CCS@UCL and CSE@CUHK, working with Prof. Peter Coveney, Dr. Xiao Xue, and Dr. Shengchao Liu on generative methods for CFD and related scientific problems.
★ Selected Publications ★
Towards Scalable One-Step Generative Modeling for Autoregressive Dynamical System Forecasting
First autoregressive one-step generative model for dynamical systems.
MENO: MeanFlow-Enhanced Neural Operators for Dynamical Systems
One-step generative modelling for multiscale fluid dynamics with neural operators.
Project Links
Past projects I've been working on!
PsiFlax
Deep QMC framework built with JAX and Flax.
AstroGen
Generative Modeling for Redshift-conditioned galaxy image synthesis.
See the project page for more!
More experiments, research code, and side projects.
✦ Personal Interests hidden archive
A few things I enjoy outside research:
- Sichuan Cuisine is my favourite!
- As a former ChOer, my passion for theoretical chemistry persists...
- Interested in modern political philosophy, my favourite modern philosopher is Giorgio Agamben
✦ Per aspera ad astra ✦