Figure 6: A sophisticated algorithm for adaptive regulation of motor variability.
Proposed algorithm implemented by the brain to actively regulate motor variability. This algorithm comprises two processes that operate over different timescales. A fast reward-dependent process (red) integrates the outcome of recent trials to generate an estimate of the reward rate that in turn determines levels of exploratory variability via the variability control function. A slower process (green) tunes the gain of the reward-dependent process over several sessions based on the uncertainty of the reward landscape. See also Figure S6.
