Free energy principle

<p>The free energy principle is a mathematical principle of information physics. Its application to fMRI brain imaging data as a theoretical framework suggests that the brain reduces <a href="/facts/Surprisal/EkUdL669">surprise</a> or <a href="/facts/Uncertainty/VM9pnGvk">uncertainty</a> by making predictions based on <a href="/facts/Mental_model/ErV8WY6h">internal models</a> and uses <a href="/facts/Stimulus_(physiology)/aBcpBfYQ">sensory input</a> to update its models so as to improve the <a href="/facts/Accuracy_and_precision/JQ47xXDq">accuracy of its predictions</a>. This principle approximates an integration of <a href="/facts/Bayesian_inference/tEpK3zLx">Bayesian inference</a> with <a href="/facts/Active_inference/Qgdkg134">active inference</a>, where actions are guided by predictions and <a href="/facts/Sensory_feedback/XnuWH3zp">sensory feedback</a> refines them. From it, wide-ranging inferences have been made about <a href="/facts/Brain_function/EKla9NkJ">brain function</a>, <a href="/facts/Perception/xLK0n2q4">perception</a>, and <a href="/facts/Action_(philosophy)/AHXKoq1V">action</a>. Its applicability to living systems has been questioned.
</p>

Free energy principle open-in-new

Free energy principle