My Research

I'm currently working as a Research Scientist at GE Research. Previously, I worked as a Graduate Research Assistant at UCLA in the Computational Vision and Learning Lab working with Dr. Hongjing Lu. I was also a member of the Reasoning Lab working with Dr. Keith Holyoak. I received my PhD in Computational Cognition in 2018 at UCLA, and I received my BS in Physics in 2013 while working in the Learning and Decision-Making Lab at UT Austin.

Direction

My research focuses on how people extract physical knowledge (both low-level and abstract) from ambiguous perceptual inputs in the environment. Although the physical principles governing "how the world works" are oftentimes complex and computationally intractable, people can form predictions and judgments that are consistent with ground-truth constraints with relative ease. Moreover, they can plan and execute motor movements to interact with the world in meaningful and goal-directed ways. How does the human cognitive system achieve this feat?

I am particularly interested in how people perceive and reason about complex physical entities (e.g., non-solid substances) and how they form judgments that extend beyond observable physical properties (e.g., causality). A fundamental example is that of phenomenal causality--the immediate and irresistible impression that an initially moving object causes an initially stationary one to move forwards following a collision. What are the perceptual underpinnings guiding this impression, and how can those inputs be manipulated to enhance or attenuate causal judgments?

My research methods have primarily utilized realistic dynamic imagery from modern simulation methods in computer graphics--as well as virtual reality environments--to study the perceptual and reasoning components in intuitive physics. In sum, my research has shown that people's mental simulations about novel situations are consistent with ground-truth principles given noisy approximation of perceptual inputs. Moreover, different physical reasoning tasks may not always rely on the same simulation mechanism: e.g., a baseball player might utilize different forms of simulation to predict where a fly ball will land vs. how long it will stay in the air. My overall goal is to further explore how the brain emulates classical mechanical principles to build expectations about how entities behave in the physical world.