Research

I build probabilistic models that enable large-scale hypothesis testing with a minimum of human oversight. By working towards automated interventions, my tools will make new experiment designs possible from what once were observational studies.

Broadly, I work on interpretable machine learning and vision with focus on behavior analysis, multi-object tracking, Bayesian nonparametrics applied to time-series, distributions on manifolds, and using uncertainty to guide decision making, analysis, and experiment design.

News

• I joined the AI Research Team at Cruise as a Senior Research Scientist in Nov, 2021.
• I presented my parts modeling and multi-object tracking with uncertainty work at the CV4Animals workshop at CVPR 2021.
• I defended my doctoral thesis Uncertainty Quantification and Structure Discovery for Scalable Behavior Science in May, 2021!
• Our paper on Sequential Bayesian Experimental Design with Variable Cost Structure was accepted at Neurips, 2020!
• I'm giving the following talk at MIT, Stanford, Harvard, G. Tech, MPI, JHU and others for Summer/Fall, 2020. Please reach out if interested.
• ATHack 2020 was a resounding success. Great work to all involved and thanks to my fellow organizers for bringing me on!
• My paper Nonparametric Object and Parts Modeling with Lie Group Dynamics was accepted for oral presentation at CVPR 2020.
• Our paper Atypical Behaviour and Connectivity in SHANK3-Mutant Macaques was published in Nature, June 2019.