Active inference unifies intentional and conflict-resolution imperatives of motor control

Abstract

The field of motor control has long focused on the achievement of external goals through action (e.g., reaching and grasping objects). However, recent studies in conditions of multisensory conflict, such as when a subject experiences the rubber hand illusion or embodies an avatar in virtual reality, reveal the presence of unconscious movements that are not goal-directed, but rather aim at resolving multisensory conflicts; for example, by aligning the position of a person’s arm with that of an embodied avatar. This second, conflict-resolution imperative of movement control did not emerge in classical studies of motor adaptation and online corrections, which did not allow movements to reduce the conflicts; and has been largely ignored so far in formal theories. Here, we propose a model of movement control grounded in the theory of active inference that integrates intentional and conflict-resolution imperatives. We present three simulations showing that the active inference model is able to characterize movements guided by the intention to achieve an external goal, by the necessity to resolve multisensory conflict, or both. Furthermore, our simulations reveal a fundamental difference between the (active) inference underlying intentional and conflict-resolution imperatives by showing that it is driven by two different (model and sensory) kinds of prediction errors. Finally, our simulations show that when movement is only guided by conflict resolution, the model incorrectly infers that is velocity is zero, as if it was not moving. This result suggests a novel speculative explanation for the fact that people are unaware of their subtle compensatory movements to avoid multisensory conflict. Furthermore, it can potentially help shed light on deficits of motor awareness that arise in psychopathological conditions.

Enhanced efficiency in visually guided online motor control for actions redirected towards the body midline

“…With a similar reasoning we could expect that the systematic bias observed in the proprioceptively sensed location of static and unseen hands towards the body midline [25,58,60,61] may play a role in the reported effect. The effect of embodiment on the motor control of a virtual arm under visuo-proprioceptive conflict has been recently documented [55,62] and could be explained as an implicit motor component that acts to minimize multisensory prediction errors [56]. According to the model, this motor component is modulated by the perceived mismatch between the visually and proprioceptively sensed locations of the arm and acts akin to a small force in the direction of the visually perceived location.…”

“…R. Soc. B 378: 20210453 on the motor control of a virtual arm under visuo-proprioceptive conflict has been recently documented [55,62] and could be explained as an implicit motor component that acts to minimize multisensory prediction errors [56]. According to the model, this motor component is modulated by the perceived mismatch between the visually and proprioceptively sensed locations of the arm and acts akin to a small force in the direction of the visually perceived location.…”

“…This is a major difference with respect to most previous studies on online corrections, in which the visual feedback from the moving hand is rendered as a small circle. Because embodiment is known to affect the perception and control of reaching actions under multisensory (visuo-proprioceptive) conflict [55,56], we could further test for possible confounds associated with the illusory embodiment of the virtual hands.…”

Unintentional synchronization with self-avatar for upper- and lower-body movements

“…From a cognitive point of view, the follower effect has been proposed to arise from the need to minimize multisensory conflicts induced by the embodiment of an avatar altering the self-body perception (Gonzalez-Franco et al, 2020;Maselli et al, 2022), which would emphasize the self-specificity of the follower effect, compared to mimicry or motor contagion. This hypothesis is grounded in the active inference theory, in which actions are understood as participating in the global goal of prediction error minimization (Friston et al, 2010;Sajid et al, 2021).…”

Skilled motor control of an inverted pendulum implies low entropy of states but high entropy of actions

“…For example, Johnson et al realized a 4-level perceptual controller for an inverted pendulum that is especially effective in withstanding external perturbations [19]. Other related approaches, such as planning-as-inference, active inference, surpriseminimising RL and KL control are also based on the idea that goal-directed behavior amounts to reducing the entropy or variance of the final (goal) state(s) of the controlled dynamical system [20][21][22][23][24][25][26][27]. For example, when balancing an inverted pendulum, the task ends in one single state, which is the one with the pendulum upwards and standing still.…”