Information-processing principles for sensorimotor learning and decision-making

Recent advances in movement neuroscience suggest that sensorimotor control can be considered as a continuous decision-making process in complex environments in which uncertainty and task variability play a key role. Leading theories of motor control assume that the motor system learns probabilistic models and that motor behavior can be explained as the optimization of payoff or cost criteria under the expectation of these models. Here we discuss how the motor system exploits task variability to build up efficient models and then discuss evidence that humans deviate from Bayes optimal behavior in their movements, because they exhibit effects of model uncertainty. Furthermore, we discuss in how far model uncertainty can be considered as a special case of a general information-processing and decision-making framework inspired by statistical physics and thermodynamics