AARON BAKER

aaron@aajbaker.com

Home
Interests
Positions
Research
and CV
  CV  
Github
The Virtue Project
In this research program based out of Boston College, our team has charted the variation of virtuous decision-making in children across cultures. Now, our goal is to test how children in each culture respond to cultural norms delivered by peers and adults differentially.

My role upon entering was to finish data collection, develop a coding mechanism and write an appendix for the first cross-cultural study. Since then I have taken on authorship responsibilities in three subsequent papers in prep and in data collection. Across all these projects, I have participated in each stage of the reseach process from creating testing materials and writing preregistrations to analysis and manuscript drafting.

Publications:
Apparatus Animations
SPSP 2023 Poster
Bayesian Norm Updating
Day to day, in many situations, we look to others for cues about what to do. What we see them do may align with what we expect, contradict it, or we may not have any expectations to begin with.

I built this computational model which seeks to use bayesian reasoning to capture how people update their beliefs about what is acceptable in different scenarios after observing others' behavior. The model is intended to complement survey data collected by Paul Deutchman with the Cooperation Lab at Boston College. Built in python, this model can show us where respondents align with and deviate from statistical learning in norm learning domains.

Follow the link to my github at the top of the page to view the code.

Publications:
Norm Graphs
Exploration
Working for Eliza Kosoy in the Gopnik lab at UC Berkeley, I contributed to this project exploring how children explore. Do children behave like the reinforcement learning agents we've built to navigate virtual environments? Can we use how they differ to inform how we design RL agents in the future? This project in collaboration with DeepMind Labs wants to find out.

I was primarily involved in data collection for this study. Participants ages 4-6 were recruited and tested at local children's museums. Our roles as experimenters involved running the DeepMind maze from terminal and facilitating children's navigation through the virtual environment using an Arduino-based controller.

Publications:
Mazes
Zebrafish
During an undergraduate research fellowship at USC, I worked with a lab innovating on bioimaging technologies. One of their microscopes allowed for beautiful neuroimaging of zebrafish, leading to a series of studies investigating the neural patterns of zebrafish during sleeping states compared to wake states.

Having access to such rich 4D neuroimaging data, my project was to map correlated neural activity across the zebrafish brain during sleep and wake bouts. I cleaned, processed, analyzed, and visualized lagged correlations of neural activities in zebrafish across sleep and wake bouts. This process involved accessing and understanding data and processing pipelines that members of the lab had already developed, as well as familiarizing myself with several data management and analysis softwares.

Click here for my slides on the Zebrafish project.
Zebrafish