Owen Marschall

This paper develops a theory of multi-tasking in neural circuits, distinguishing between sequential and simultaneous task performance and introduces a model of task barcodes to parameterize task selectivity.

The paper develops a theoretical model to explain multi-task computation in neural networks, showing how task selection can overcome interference between different task dynamics.

The paper develops a computational framework to analyze how dynamical systems in recurrent circuits evolve during learning across different tasks and algorithms, finding a common path to task mastery.