Research

Google Research

At Google I work on machine learning for science, with a focus on scientific computing and applications in climate/weather modeling.

I believe that the combination of automatic differentiation software (e.g., JAX), accelerator hardware (GPU, TPU, etc.) and deep learning are posed to transform traditional scientific computing, by vastly accelerating and improving existing numerical methods.

Most of my current work (as of May 2021) focuses on applications involving simulating Partial Differential Equations (PDEs), and in particular on improving simulation methods with machine learning.

Some recent highlights:

• Machine learning accelerated computational fluid dynamics (PNAS 2021)
• Variational Data Assimilation with a Learned Inverse Observation Operator (ICML 2021)
• Kohn-Sham equations as regularizer: Building prior knowledge into machine-learned physics (PRL 2021)
• Lagrangian neural networks (ICLR 2020 workshop)
• Neural reparameterization improves structural optimization (NeurIPS 2019 workshop)
• Learning data-driven discretizations for partial differential equations (PNAS 2019)

For my full publications, see my Google Scholar profile.

UC Berkeley

I graduated from the University of Calfornia, Berkeley in 2013 with a Ph.D. in theoretical physics. My advisor was Birgitta Whaley.

The topic of dissertaion was the role of electronic quantum coherence in photosynthetic energy transfer. My main model system was the Fenna-Matthews-Olson complex of green sulfur bacteria (shown above).

1. S Jang, S Hoyer, G Fleming, KB Whaley. Generalized Master Equation with Non-Markovian Multichromophoric Förster Resonance Energy Transfer for Modular Exciton Densities, Phys Rev Lett 113, 188102 (2014). arXiv:1311.2091
2. S Hoyer, F Caruso, S Montangero, M Sarovar, T Calarco, MB Plenio, KB Whaley, Realistic and verifiable coherent control of excitonic states in a light harvesting complex, New J Physics 16, 045007 (2014). arXiv:1307.4807
3. S Hoyer and KB Whaley, Inverting pump-probe spectroscopy for state tomography of excitonic systems, J Chem Phys 138, 164102 (2013), arXiv:1209.6625. Mathematica source for dimer model in Sec. IV(A).
4. S Hoyer, A Ishizaki and KB Whaley, Spatial propagation of excitonic coherence enables ratcheted energy transfer, Phys Rev E 86, 041911 (2012), arXiv:1106.2911
5. S Hoyer, M Sarovar and KB Whaley, Limits of quantum speedup in photosynthetic light harvesting, New J Phys 12, 065041 (2010), arXiv:0910.1847
6. S Hoyer and DA Meyer, Faster transport with a directed quantum walk. Phys Rev A 79, 024307 (2009), arXiv:0901.1007. Mathematica source for Fig. 2.

Swarthmore College

At Swarthmore, I majored in physics and minored in mathematics. Here are my final projects for each department:

• Undergraduate thesis: Quantum random walk search on satisfiability problems
• Math senior conference paper: Classification of completely positive maps