Scientific Software

I build performance-minded research software for scientific machine learning, uncertainty quantification, dynamical systems, and physical simulation. Selected public work includes:

Machine Learning and Generative Modeling

JaxDiffusion
Conditional diffusion models implemented in JAX for scientific applications.

SymbolicMemory
Transformer-oriented experimentation for structured and symbolic sequence modeling.

Dynamical Systems and Reduced-Order Modeling

MarkovChainHammer.jl
Julia tools for transfer operators, state-space partitions, and reduced Markov representations of dynamical systems.

StateSpacePartitions.jl
State-space partitioning methods for representing turbulent statistics and complex dynamics.

Scientific Computing Contributions

Oceananigans.jl
Contributor and coauthor on GPU-native geophysical fluid dynamics and high-resolution ocean modeling work.

CalibrateEmulateSample.jl
Contributor and coauthor on accelerated parametric uncertainty quantification workflows in Julia.