Undergraduate Research & Scholarships

Jennifer Yang

Comparing Social Reasoning in Human Infants and Large Language Models

Infants have an understanding of others’ goals and intentions that emerges very early in development, but it is unclear if AI has the same abilities. This project compares social reasoning in human infants and large language models (LLMs) by converting classic developmental psychology studies into text prompts to evaluate LLMs. This project will provide insight into whether LLMs and humans have comparable social reasoning abilities, and, if not, illuminate possible deficits in LLMs’ social reasoning capabilities. This summer, I will assist with designing and running prompting experiments with LLMs and build my theoretical foundation in both cognitive development and AI.

Message To Sponsor

Thank you so much for this opportunity! I am truly grateful to be selected and excited to dive into the work I am passionate about. This funding will allow me to develop real technical experience with large language models while contributing to research I find genuinely meaningful.