NeuroSelect

The brain has the remarkable ability to imagine different solutions to a problem and to adjust to new situations more rapidly than any artificial system facing similar problems. Yet, the difficulty of probing and manipulating brain-wide activity at cellular resolution during decision-making has limited our exploration of the computational and dynamical neural properties that makes the brain so flexible. By leveraging the latest advances in genetical and neurophysiological techniques in mice, which can be trained to perform simple forms of flexible reasoning, the overarching goal of the project is to identify fundamental neural principles of flexible behaviour. We posit that the brain can acquire a powerful form of cognitive flexibility by parallelizing operations of decision processing, allowing animals to focus on one decision strategy, while their brains maintain several alternative solutions readily available in the background. We will test this hypothesis through a series of experiments that aim at understanding: (1) how the brain selects a strategy among multiple alternatives, (2) what neural circuits govern which decision strategy to use and when, (3) what neuronal mechanism triggers a change in strategy. We will address these questions in the context of foraging, a fundamental survival behaviour shared across all vertebrates and that consists of finding resources in volatile environments. The project will deliver a new conceptual framework for thinking about behavioural flexibility as an adaptive selection mechanism of parallel decision computations and elucidate its neural bases. Ultimately, this research will address several aspects of the decision-making conundrum, and perhaps lay the groundwork for explaining diversity in behaviour, including pathological conditions characterized by deficits in behavioural flexibility.