Researchers from City University of Hong Kong and The Chinese University of Hong Kong have described how the brain reorganizes itself while learning to handle two tasks, in a study published September 3.
The team tracked the same individual neurons in mice as the animals learned to keep moving a lever while also responding to sounds. At first, neurons shared the work and coordinated the competing demands.
With practice, the brain recruited more task-specific neurons and separated the two patterns of activity. The mice then became better at doing both tasks with less interference.
The researchers also trained recurrent neural networks on a similar challenge. The networks learned faster when they kept some early coordination before gradually separating the tasks, suggesting a possible idea for future artificial intelligence systems.
