Research Projects
alabi
Active Learning Accelerated Bayesian Inference (ALABI) —
An open-source Python package for performing Bayesian inference with computationally expensive forward models.
Given a likelihood function and priors, ALABI trains a Gaussian Process surrogate model to predict posterior
probability and uses active learning to iteratively improve predictions in high-likelihood regions. Supports
both affine-invariant MCMC (emcee) and nested sampling (dynesty), reducing expensive
model evaluations by factors of thousands.
Python Bayesian Inference Gaussian Processes MCMC Active Learning
spotgp
Physically motivated Gaussian Process kernels for modeling stellar starspot variability, implemented in
JAX. Provides differentiable GP kernels designed to capture the quasi-periodic signals
produced by rotating starspots on stellar surfaces.
Python JAX Gaussian Processes Stellar Variability
vplanet_inference
Python tools for doing inference with VPLanet, a framework for simulating planetary system evolution. Provides utilities for setting up parameter sweeps, running VPLanet simulations, and performing Bayesian parameter estimation on planetary and stellar evolution models.
Python VPLanet Planetary Evolution Bayesian Inference
Other Open Source Projects
revealjs_gui
A self-hostable WYSIWYG presentation editor powered by reveal.js. Build and present slides in the browser — no account, no cloud, no tracking. Features rich formatting, shape tools, code blocks with syntax highlighting, $\LaTeX$ math support, HTML embeds, speaker notes, themes, and export to HTML/PDF.
JavaScript reveal.js WYSIWYG Editor Presentations
tikz_editor
A graphical interface for creating and editing LaTeX TikZ diagrams. Try it out here! https://jessicabirky.com/tikz_editor/
HTML LaTeX TikZ