Special BMI Seminar // Rikki Rabinovich & Jack Bowler

Speaker 1: Rikki Rabinovich (University of Utah, Department of Neurosurgery)
Title: Human brain network control of creativity

The ability to think creatively is fundamental to human cognition, shaping how we interact with the world and navigate complex problems. In this talk, I discuss our recent findings that reveal how the human brain generates creative ideas. Using multiscale intracranial electrophysiology—from single-unit recordings to local field potentials across distributed brain networks—we demonstrated that unique brain states underlie creative vs. mathematical cognitive states. Further, we uncovered nonlinear, high-dimensional representations of moment-by-moment fluctuations in creative performance. These representations define a creativity axis shared between network-level and single-neuron computations. Our findings reveal widespread neural representations of cognitive state and suggest distinct roles of specific brain networks in controlling creativity, with the default mode network gating access to creative states and the dorsal attention network regulating the quality of creative output.

Speaker 2: Jack Bowler (University of Utah, Department of Neurobiology)
Title: Learning to generalize: Artificial networks reveal how the structure of prior experience shapes abstract thinking in real brains

Animals can solve new, complex tasks by reusing and adapting what they’ve learned before. This kind of flexibility depends not just on having prior experience, but on how that experience was structured in the first place. To investigate how early training shapes learning, we first trained recurrent neural networks (RNNs) on an odor-timing task previously used to study complex timing behavior in mice. We then tested the RNN predictions regarding the effect of experience on generalization by comparing simulated RNN results to behavioral and electrophysiological recordings in mice trained on the same task through a progression of staged training sequences. Both RNNs and mice trained without well-structured early experience developed rigid strategies and made repeated errors. In contrast, properly structured early training improved generalization in RNNs and mice. Using dynamical systems approaches, we revealed a mechanism for this effect: networks trained with appropriately structured curricula developed distinct dynamical motifs that support the correct abstractions when complexity was increased. Networks that lacked early training or received improperly structured curricula developed single fixed-point solutions that failed to generalize about the task structure. Together, these findings demonstrate how the structure of prior experience governs flexible and generalizable knowledge emergence in both biological systems and computational models.