Memory decay and generalization following distinct motor learning mechanisms

Abstract

Motor skill learning is considered to arise out of contributions from multiple learning mechanisms, including error-based learning (EBL), use-dependent learning (UDL), and reinforcement learning (RL). These learning mechanisms exhibit dissociable roles and engage different neural circuits during skill acquisition. However, it remains largely unknown how a newly formed motor memory acquired through each learning mechanism decays over time and whether distinct learning mechanisms produce different generalization patterns. Here, we used variants of reaching paradigms that dissociated these learning mechanisms to examine the time course of memory decay following each learning and the generalization patterns of each learning. We found that motor memories acquired through these learning mechanisms decayed as a function of time. Notably, 15 minutes, 6 hours, and 24 hours after acquisition, the memory of EBL decayed much greater than that of RL. The memory acquired through UDL faded away within a few minutes. Motor memories formed through EBL and RL for given movement directions generalized to untrained movement directions, with the generalization of EBL being greater than that of RL. In contrast, motor memory of UDL could not generalize to untrained movement directions. These results suggest that distinct learning mechanisms exhibit different patterns of memory decay and generalization.

Reinforcement Learning during Locomotion

“…Reinforcement learning also impacts motor memories. Implicit aftereffects, defined as motor memories not under conscious control ( Krakauer et al, 2019 ), are strengthened when reinforcement is combined with sensorimotor adaptation ( Huang et al, 2011 ; Shmuelof et al, 2012 ; Galea et al, 2015 ) or use-dependent learning ( Mawase et al, 2017 ; Bao and Lei, 2022 ; c.f., Tsay et al, 2022 ). Reinforcement learning may also strengthen explicit retention, defined as the ability to consciously remember and reproduce a previously learned movement ( Schmidt and Lee, 2005 ), potentially because individuals benefit from determining what the successful movement is themselves through exploration, improving engagement in the task, unlike when receiving full visual feedback of movements ( Winstein and Schmidt, 1990 ; Winstein et al, 1994 ; Hasson et al, 2015 ).…”

Reinforcement learning during locomotion

“…Reinforcement learning also impacts motor memories. Implicit aftereffects, defined as motor memories not under conscious control (Krakauer et al, 2019), are strengthened when reinforcement is combined with sensorimotor adaptation (Huang et al, 2011; Shmuelof et al, 2012; Galea et al, 2015), or use-dependent learning (Mawase et al, 2017; Bao and Lei, 2022; c.f., Tsay et al, 2022). Reinforcement learning may also strengthen explicit retention, defined as the ability to consciously remember and reproduce a previously learned movement (Schmidt and Lee, 2005).…”

“…Reinforcement learning also impacts motor memories. Implicit aftereffects, defined as motor memories not under conscious control (Krakauer et al, 2019), are strengthened when reinforcement is combined with sensorimotor adaptation (Galea et al, 2015; Huang et al, 2011; Shmuelof et al, 2012), or use-dependent learning (Mawase et al, 2017; Bao and Lei, 2022; c.f., Tsay et al, 2022). Reinforcement learning may also strengthen explicit retention, defined as the ability to consciously remember and reproduce a previously learned movement (Schmidt and Lee, 2005), potentially because individuals benefit from determining what the successful movement is themselves through exploration, improving engagement in the task, unlike when receiving full visual feedback of movements (Hasson et al, 2015; Winstein et al, 1994; Winstein and Schmidt, 1990).…”

Exploration in reward-based motor learning

“…To save time and facilitate measuring performance, experimental tasks often involve movements of fewer body parts in a setting with an instructed movement goal and in which movement success can easily be defined and measured. Most experiments involve target-directed movements with the arm (Bao & Lei, 2022;Bernardi et al, 2015;Cashaback et al, 2019;Holland et al, 2019;Holland et al, 2018;Ikegami et al, 2022;Izawa & Shadmehr, 2011;Kuling et al, 2019;Manley et al, 2014;Pekny et al, 2015;Roth et al, 2023;Shmuelof et al, 2012;Sidarta et al, 2018Sidarta et al, , 2022Therrien et al, 2016Therrien et al, , 2018van der Kooij & Overvliet, 2016; or finger (Uehara et al, 2019).…”

“…Alternatively, feedback about limb movement is not perturbed but feedback about movement success is: external feedback can be manipulated in such a way that not all movements have an equal probability of reward, so that participants learn to move in ways that yield the most reward. Reaches that are within a visual target may for example be more frequently rewarded the more they are on the left side of the target (Cashaback et al, 2019) or only when they are in a specific zone within the visual target (Bao & Lei, 2022;Bernardi et al, 2015;Manley et al, 2014;Roth et al, 2023;Sidarta et al, 2018Sidarta et al, , 2022. In other studies, participants had to learn to correct for their own natural biases in reaching to unseen targets (Kuling et al, 2019) or the relation between their joint motion and reward (Vassiliadis et al, 2021(Vassiliadis et al, , 2022Wiegel, 2021).…”