Strategy choice: Individual difference and temporal dynamics

In computational cognitive neuroscience, the standard approach to individual differences is to fit every participant to the same model - a single "algorithm" - and compare parameters across participants. We might ask if one person has a faster learning rate or a higher risk attitude than another. While using the same model across individuals puts everyone on the same scale, we might completely miss a more fundamental difference - in the cognitive strategy itself.

In this line of research, we move beyond parameter-fitting to identify the diverse repertoire of strategies people use to solve problems. We focus on two core dimensions:

• Diversity across individuals: How do people differ in their fundamental approach to a task? We seek to identify reliable "strategic phenotypes" that distinguish one person from another.
• Dynamics over time: Strategies are not static. We investigate the temporal dynamics of cognition - how a person’s strategy emerges, evolves through practice, or shifts due to internal states like mood or fatigue.

Our goal is to uncover the computational principles that govern these choices. By developing new inference frameworks, we aim to transform "individual variability" from a source of noise into a structured, predictable signal that explains the rich diversity of human behavior.