Role of Lateral Inhibition on Visual Number Sense

Abstract

Lateral inhibition is a basic principle of information processing and widely exists in the human and animal nervous systems. Lateral inhibition is also involved in processing visual information because it travels through the retina, primary visual cortex, and visual nervous system. This finding suggests that lateral inhibition is associated with visual number sense in humans and animals. Here, we show a number-sensing neural network model based on lateral inhibition. The model can reproduce the size and distance effects of the output response of human and animal number-sensing neurons when the network connection weights are set randomly without adjustment. The number sense of the model disappears when lateral inhibition is removed. Our study shows that the first effect of lateral inhibition is to strengthen the linear correlation between the total response intensity of the input layer and the number of objects. The second one is to allow the output cells to prefer different numbers. Results indicate that lateral inhibition plays an indispensable role in untrained spontaneous number sense.