Research

My research lies at the intersection of machine learning and scientific modeling. I develop methods for simulation-based inference and apply them to problems in neuroscience and beyond.

Simulation-Based Inference

Many scientific models are defined through simulators that are easy to run forward but difficult to invert. I develop neural network-based methods to perform Bayesian inference for such models, enabling researchers to identify parameters and even select between competing models from observed data.

Computational Neuroscience

During my PhD, I built biophysically detailed models of retinal circuits — from individual synapses to networks of neurons. This work combined electrophysiological recordings with mechanistic modeling to understand how the retina processes visual information.

Applications of SBI

Simulation-based inference is broadly applicable across the sciences. I have contributed to applications in glaciology (inferring ice shelf properties), water quality mapping in coral reef environments, and diffusion MRI tractography.

For a full list of publications, see the Publications page.