Neural circuit mechanisms for multi-tasking

Neural activity during cognitive or motor tasks can be described as existing on a low-dimensional manifold within the space of all possible activity patterns. When multiple tasks are performed simultaneously, a brain region's activity may generate multiple such manifolds. Researchers have proposed various models to explain this multi-tasking capability, each with distinct assumptions about architecture and inputs.

This article presents a theory of multi-tasking that categorizes models and experimental paradigms based on sequential and simultaneous task performance.

Sequential multi-tasking involves performing one task at a time, while simultaneous multi-tasking entails executing multiple tasks in parallel. The study demonstrates that interference between tasks often hinders multi-tasking performance. To address this issue, the authors introduce a model called "task barcodes" that overcomes the problem and parameterizes the neuronal population's task selectivity.

The theory outlined in the research reveals categorical distinctions between neural circuit mechanisms supporting multi-tasking. These distinctions can be investigated using physiological recordings, providing a framework for understanding the underlying neural mechanisms of multi-tasking.