Postdoctoral Fellowship: SPRF: Behavioral, Computational, and Developmental Approaches to Understanding Knowledge Inference

Under the sponsorship of Dr. Julian Jara-Ettinger at Yale University, this postdoctoral fellowship award supports an early career scientist investigating knowledge inference. How do we figure out what someone else knows, especially when we’ve just met them? This is a foundational capacity for effective communication, as it helps avoid over-explaining and under-informing. This project brings a computational approach to understand how humans rapidly infer others’ knowledge from sparse cues, like how someone answers a question or how long they take to answer. Understanding this process is essential for building artificial intelligence (AI) that can fluently interact with people. Such systems must be able to infer users’ knowledge and adjust their communication strategies without relying on explicit feedback. This research advances our understanding of how such adaptive, socially aware behavior can be achieved. This research investigates how people infer what others know based on minimal evidence. The project combines computational modeling and empirical studies with children and adults to study this capacity across ages. Participants will watch a stranger’s response to just one or two queries (e.g., how they answer a question, including length of pause) and be asked to estimate what else the person knows in that domain. By aligning behavioral data with model predictions, the research will formalize how humans estimate others’ knowledge using subtle aspects of what someone says (like hesitation) and graph-based expectations of how knowledge is organized (like which topics are widely known). This work will deepen our understanding of human reasoning about knowledge and provide a formal foundation for designing AI systems that better track what users know in real time. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.