Overview
I’m a PhD student interested in researching embodied computational models of human behavior, especially perceptual, linguistic, and social skills.
Using knowledge of how humans perceive and operate on the world and the social beings around them can inform the creation of more intelligent, adaptive, and communicative systems.
I am particularly interested in the potential application of these skills in socially assistive robotics and robots for education.
My other interests include human-robot interaction, natural language processing, artificial intelligence, and social learning across cultures.
Education
I majored in Cognitive Science with a concentration on Mind & Computation.
My thesis, Grounded Natural Language for Collaborative Robotics, was advised by Prof. Brian Scassellati.
- GPA: 3.92/4.00
- Graduated Magna Cum Laude
Research experience
Graduate research assistant and co-investigator as part of the Human Interaction & Robotics Group
and the Collaborative AI & Robotics Lab.
Projects
These are some of my most recent projects.
Towards equitable robot tutoring -- an intersectional analysis of human-robot interaction in racially diverse classrooms, with Dr. Alessandro Roncone and Dr. Tiera Tanksley
- 2021-present (ongoing). We investigate the potential of autonomous pedagogical interventions for students from racially minoritized backgrounds. More specifically, this work will evaluate the performance of AI partners with early middle school Students of Color from low-income urban districts within the context of small group tutoring, with the goal of providing foundational insights on the specific needs of underserved students. Our long-term goal is to design and deploy socially and intellectually engaging AI partners that support both content learning and the development of constructive learning behaviors that will prepare students for their continuing educational careers.
Teaching Grounded Language Skills via an Interactive Robot Tutor, with Dr. Bradley Hayes
-
2021-present (ongoing). We propose and develop a classroom-based social robot tutoring system that targets reading comprehension skills using methods inspired by successful embodied reading pedagogical practices. This system will incorporate skill tracing of comprehension skills in order to identify problem areas, and to help students ground their reading in concrete experiences.
Grounded Natural Language for Collaborative Robotics, with Jake Brawer and Dr. Brian Scassellati
-
2019-2020. Social and communicative barriers between humans and robots limit human-robot collaboration from reaching the fluency and intuitiveness of human-human interactions. I helped to address this limitation by creating a cognitively-inspired semantic grounding model for language generation, based largely on recent work in psychophysics and psycholinguistics. I collected a corpus of original referring expression data from human participants and analyzed it to ensure the model was based on sound and consistent principles of human speech. This model allowed the robot to dynamically generate natural and informative requests for tools and other task-relevant objects from a human partner.
Teaching & Mentorship
Free college advising for low-income high school students (ASC-certified)
Academic and residential advisor to first-year students
Residential advisor to high school students in summer course program
Artificial Intelligence (Yale CPSC470) and Data Structures & Programming Techniques (Yale CPSC223)
Teaching college math fundamentals for incoming Yale students from low-income backgrounds
Technical Skills & Proficiency
Python and its many ML libraries (PyTorch, Tensorflow, OpenCV, etc.)
ROS (Robot Operating System) package development
Reinforcement Learning for real-world problem-solving
Natural Language Processing for robotics applications
Systems programming & software engineering
