Date: Friday, September 26, 2025
Time: 12:00 pm – 1:00 pm
Location: Seminar Room 3310

Title: Computational and Neural Mechanisms of Behavioral Flexibility 

Abstract:
Survival in a complex world depends on an organism’s ability to adapt to unpredictable changes in its environment. Adapting to these changes is challenging because they vary widely in both their characteristics, such as sensory modalities and reward feedback, and the timescales over which they occur. For instance, fluctuations in reward feedback may arise from the inherent randomness of outcomes, sudden shifts in reward probabilities, or changes in the relationships between cues, actions, and outcomes—each posing a unique form of uncertainty and demanding a different adaptive strategy. In this talk, I will explore how computational modeling can illuminate the neural mechanisms that support behavioral flexibility in the face of such uncertainty across species (rodents, monkeys, and humans). Our work to date has revealed multiple layers of adaptation: at the synaptic level, mechanisms such as reward-dependent metaplasticity and dynamic learning rates modulated by higher cognitive functions; at the cellular level, adjustments in neural response timescales and range normalization; and at the circuit level, arbitration between competing learning systems along with flexible changes in the geometry of neural representations. Along the way, I will introduce new computational methods for quantifying behavioral and neural adjustments, and demonstrate how these tools can uncover the core computational and neural mechanisms underlying flexibility in learning and decision making across mammals.