Postdoctoral Fellowship: SPRF: Exploring the Role of Episodic Memory in Adaptive Decision Making

Nicholas, Jonathan, New York NY

Investigators

Abstract

Under the sponsorship of Dr. Marcelo Mattar at New York University, this postdoctoral fellowship award supports an early career scientist investigating how humans use detailed memories of past experiences to make smart decisions when the future is uncertain. Effective decision making often depends on referencing details from our past experiences. Yet we rarely know what aspects of the present will be important for decisions in the future. The research addresses a fundamental challenge in artificial intelligence and human cognition: how to store information effectively when its future usefulness is unknown. The findings and computational models that result from this research will advance the development of more human-like artificial intelligence systems that can adapt to changing circumstances and make better decisions based on past experiences. The project investigates whether humans solve the challenge of storing information for flexible use in the future by maintaining detailed memories of individual experiences that can be accessed when needed. The research combines behavioral experiments with brain imaging techniques to determine when and how people rely on detailed memories during decision-making. Participants complete tasks where they must encode experiences with multiple features and later make decisions based on these features to maximize rewards. The project also develops computational models using artificial neural networks to understand what memory strategies emerge as optimal solutions when future task demands are uncertain. By comparing human behavior to these computational models, the research will reveal whether humans adopt strategies that are theoretically optimal given the brain's biological constraints. Additionally, the project examines how the organization of memories - such as their timing and relationships to each other - influences decision-making. This work bridges gaps between research on memory, decision-making, and artificial intelligence, providing insights that could improve both our understanding of human cognition and the development of more adaptive AI systems. 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.