I am a Postdoc in the Machine Learning & Analytics Group at Lawrence Berkeley National Laboratory. Here, my research explores visualization for scientific ML (e.g., developing new ways to visualize loss landscapes and characterize learning) with diverse scientific applications (e.g., physics-informed neural networks, weather forecasting).
I earned my PhD in Biophysics from Stanford University. For my dissertation, I worked with Manish Saggar in the Brain Dynamics Lab to develop new computational methods for capturing and quantifying fluctuations in brain dynamics using ideas from machine learning and topological data analysis (TDA). My initial work in the lab focused on applying the Mapper algorithm from TDA to neuroimaging data and developing DyNeuSR. Other projects included optimizing Mapper for neuroimaging data and exploring how Mapper (and other unsupervised techniques from TDA) can be combined with supervised learning approaches, for example, to develop more robust brain decoding models.
Before starting graduate school, I worked with Rhiju Das as a research programmer in the Biochemistry department at Stanford. My research there focused on developing computational methods for studying biophysical systems, but my role and specific projects ranged from the in-house Rosetta software developer implementing new algorithms for RNA structure modeling, to one of the main developers for the Eterna project leading a re-design of the popular citizen science game.
I earned my BS in Molecular and Cellular Biology from the University of Illinois, Urbana-Champaign.
Email (Personal): calebgeniesse [at] gmail [dot] com
Email (Stanford): geniesse [at] stanford [dot] edu
Email (Berkeley Lab): cgeniesse [at] lbl [dot] gov
Linkedin: Caleb Geniesse
Twitter: @calebgeniesse
GitHub: calebgeniesse