Projects :)

The overarching goal of this project is to understand how cognitive fatigue relates to effort, and how switching between tasks may impact that relationship. So far, I have designed an online behavioral task to analyze how self-selected rest times (as a measure of fatigue) change depending on performance levels and task switches; results suggest that switching tasks when performing poorly may be beneficial to performance. Currently, I am developing a computational model based on this data to better understand the impact of task switches on the rejuvenating aspects of rest.

This project aims to understand changes in brain activity when people get more experience with a specific task. Currently, I am using an fMRI dataset where subjects had numerous experiences with the same cognitive control tasks to understand how the distribution of activated networks changes over time from practice.

Advised by Tom Griffiths as well as Bill Thompson and Robert Hawkins, this project was the result of a semester-long independent work research project at Princeton. We utilized, coded, and adapted multiple versions of a Hierarchical Bayesian Model to simulate language emergence in various population sizes.

Advised by Olga Russakovsky at Princeton, this project was the result of a semester-long independent work research project during my undergraduate degree. I tested the CLIP image recognition system on a dataset of images taken by the Blind or Visually Impaired (the VizWiz-Captions dataset), to identify whether this novel image recognition system—not explicitly trained on images from sighted individuals—would be able to fulfill the needs of a large part of the population. In doing so, I developed a system to statistically evaluate CLIP’s accuracy on this dataset, and attempted to improve it’s accuracy by fine-tuning a neural network classifier and adding it to the testing pipeline.