BRAINSTORM workshop trains researchers in next-generation brain modeling

ARIA co-sponsored program equips 30 researchers with hands-on skills in reinforcement learning, Bayesian inference and AI-driven model discovery

Brown University’s Carney Institute for Brain Science hosted its 2026 BRAINSTORM Computational Modeling Workshop, an intensive five-day program co-sponsored by ARIA that brought together approximately 30 researchers, including participants from Brown and ARIA partner institutions, to learn how to build and test computational models of the mind and brain.

Combining lectures with hands-on coding tutorials, the workshop took participants from foundational principles of computational modeling through reinforcement learning, Bayesian models, sequential sampling models and hierarchical Bayesian inference. It also introduced emerging AI methods that make it possible to fit more realistic, neurally inspired models to behavioral data and to automate the discovery of new models.

Rather than stopping at theory, the program focused on practice: participants learned to implement, fit, compare and evaluate models themselves through a series of guided coding exercises.

From the classroom to an international challenge

Following the program, several participants formed teams to compete in the MindRL Modeling Challenge, an international competition with public leaderboards that aims to redefine how models are evaluated in the AI era. Results will be announced later, but organizers count the teams’ participation — applying the methods they learned at Brown — as an early marker of the workshop’s success.