Neurocomputational maturation of explore-exploit decision-making in adolescence and emerging adulthood.

Poster Session F - Tuesday, April 16, 2024, 8:00 – 10:00 am EDT, Sheraton Hall ABC

Dr. Jeremy Hogeveen¹ (jhogeveen@unm.edu), Teagan Mullins², Ethan Campbell³, Caitlin Enders⁴, Cidney Robertson-Benta⁵, Margaret Austin⁶, Yiliang Zhu⁷, Vincent Costa⁸, Katie Witkiewitz⁹; ¹The University of New Mexico, ²Oregon National Primate Research Center

To discover our favorites, we start by trying something new. Typically, such exploration of novel choice alternatives comes at the cost of foregoing options that have benefited us in the past. This highlights the canonical tradeoff between exploration (i.e., sampling novel options with an uncertain value) and exploitation (i.e., sticking with familiar favorites with a known value) in human reinforcement learning. During adolescence—ever-increasing motivations to explore novel options are adaptive and help promote self-discovery, but may also elevate risk-taking and psychiatric vulnerability. We merged behavior, computational modeling, and task fMRI to probe the neurocomputational maturation of explore-exploit decision-making across adolescence (N=135 13-21 year-old participants). We also examined explore-exploit phenotypes across ‘reward’ and ‘loss’ environments, to determine whether distinct neurocomputational mechanisms underpin exploration to maximize gains, versus exploration to avoid losses. Our data reveal a clear age-related shift in exploration based on valence: younger participants demonstrate a marked shift in choice computations causing them to engage in hyper-exploratory behavior in loss relative to win contexts. In contrast, young adults demonstrate consistent goal-directed exploration across win and loss. Analysis of fMRI data is ongoing, but will determine how responses to feedback in cortico-subcortical circuits anchored in the dorsal and ventral striatum, amygdala, and the bed nucleus of the stria terminalis shape the probability of exploration versus exploitation in win versus loss contexts as a function of age. These data represent an initial cross-sectional preview of a 5-year longitudinal protocol designed to chart the definitive neurodevelopmental trajectory of explore-exploit decision-making across adolescence.

Topic Area: EXECUTIVE PROCESSES: Development &aging