What happens when neighbouring neurons have less overlap in their responses after learning?

When neighbouring neurons have less overlap in their responses following perceptual learning, it indicates a sharpening of their "tuning functions." This process, often referred to as decorrelation, means that each neuron becomes more specialized by narrowing the range of stimulus attributes to which it is sensitive.1

By reducing the overlap between the responses of neighbouring neurons, the brain effectively increases the distance between them along the trained stimulus dimension. This change allows each neuron to produce a response that is as distinct as possible from its neighbours. Consequently, this sharpening can lead to a reduction in the total number of neurons that respond to a given stimulus, as the neural representation becomes more precise and efficient.1