Abstract
In motor learning, the slow development of implicit learning, following explicit components of learning is well established. While much is known about behaviour during adaptation to a perturbation in reaches, saccades and locomotion, little is known about implicit processes during adaptation. Implicit learning is characterized by both changes in internal models and state estimates of limb position, which we measure as reach aftereffects and shifts in hand localization, after every training trial. This allows trial-by-trial mapping of implicit learning. Participants reached to targets with aligned, then 30° rotated, counter-rotated and finally error-clamped cursor feedback. This paradigm allows fitting a common state space model to the reach performance. The slow process of the model did not match the time course of either of our implicit measures. The observed implicit changes were near asymptote after only one perturbed training trial and thus occur much faster than conventionally believed.
