Multitasking Capability Versus Learning Efficiency in Neural Network Architectures

One of the most salient and well-recognized features of human goal-directed behavior is our limited ability to conduct multiple demanding tasks at once. Previous work has identified overlap between task processing pathways as a limiting factor for multitasking performance in neural architectures. This raises an important question: insofar as shared representation between tasks introduces the risk of cross-talk and thereby limitations in multitasking, why would the brain prefer shared task representations over separate representations across tasks? We seek to answer this question by introducing formal considerations and neural network simulations in which we contrast the multitasking limitations that shared task representations incur with their benefits for task learning. Our results suggest that neural network architectures face a fundamental tradeoff between learning efficiency and multitasking performance in environments with shared structure between tasks.