My research focuses on building a deeper understanding of exactly what a robot or agent has learned. For instance, engineers should have an idea of which situations their robot may act incorrectly in, and end-users should be able to anticipate how a robot they are interacting with will behave in various situations. This is essential for building trust, enabling seamless human-robot collaboration, verifying what a robot has learned, and deploying robots in safety-critical situations.
I previously worked on teaching robots how to manipulate deformable objects (knot tying and towel folding) through learning from demonstrations. While at Stanford, I applied machine learning to problems in clinical decision making and information flow tracking.
Outside of academics, I am involved in STEM outreach and mentoring. I like running, yoga, tea, baking, and board game nights.