Researchers have demonstrated that self-organization of neurons as they ‘learn’ follows a mathematical theory called the free energy principle. The principle accurately predicted how real neural networks spontaneously reorganize during learning, as well as how altering neural excitability can disrupt the process.
