How does the reweighting model explain visual perceptual learning?

In the context of visual perceptual learning, the reweighting model—also referred to as read-out modification—proposes that learning occurs through changes in the strength of connections between cortical sensory representations and mid- or high-level brain areas.1

Rather than modifying the properties of neurons in the early stages of visual processing (as in representation modification models), the reweighting model suggests that the brain improves performance by adjusting how it "reads out" or interprets the information already provided by those early sensory representations. By altering the weights of these connections, the system becomes more efficient at integrating sensory input to inform decision-making processes.1