Goal-directed Planning Research Articles

Meta-learning of human motor adaptation via the dorsal premotor cortex

Meta-learning enables us to learn how to learn the same or similar tasks more efficiently. Decision-making literature theorizes that a prefrontal network, including the orbitofrontal and anterior cingulate cortices, underlies meta-learning of decision making by reinforcement learning. Recently, computationally similar meta-learning has been theorized and empirically demonstrated in motor adaptation. However, it remains unclear whether meta-learning of motor adaptation also relies on a prefrontal network. Considering hierarchical information flow from the prefrontal to motor cortices, this study explores whether meta-learning is processed in the dorsolateral prefrontal cortex (DLPFC) or in the dorsal premotor cortex (PMd), which is situated upstream of the primary motor cortex, but downstream of the DLPFC. Transcranial magnetic stimulation (TMS) was delivered to either PMd or DLPFC during a motor meta-learning task, in which human participants were trained to regulate the rate and retention of motor adaptation to maximize rewards. While motor adaptation itself was intact, TMS to PMd, but not DLPFC, attenuated meta-learning, impairing the ability to regulate motor adaptation to maximize rewards. Further analyses revealed that TMS to PMd attenuated meta-learning of memory retention. These results suggest that meta-learning of motor adaptation relies more on the premotor area than on a prefrontal network. Thus, while PMd is traditionally viewed as crucial for planning motor actions, this study suggests that PMd is also crucial for meta-learning of motor adaptation, processing goal-directed planning of how long motor memory should be retained to fit the long-term goal of motor adaptation.

Comparative Roles of the Caudate and Putamen in the Serial Order of Behavior: Effects of Striatal Glutamate Receptor Blockade on Variable versus Fixed Spatial Self-Ordered Sequencing in Marmosets.

Self-ordered sequencing is an important executive function involving planning and executing a series of steps to achieve goal-directed outcomes. The lateral frontal cortex is implicated in this behavior, but downstream striatal outputs remain relatively unexplored. We trained marmosets on a three-stimulus self-ordered spatial sequencing task using a touch-sensitive screen to explore the role of the caudate nucleus and putamen in random and fixed response arrays. By transiently blocking glutamatergic inputs to these regions, using intrastriatal CNQX microinfusions, we demonstrate that the caudate and putamen are both required for, but contribute differently to, flexible and fixed sequencing. CNQX into either the caudate or putamen impaired variable array accuracy, and infusions into both simultaneously elicited greater impairment. We demonstrated that continuous perseverative errors in variable array were caused by putamen infusions, likely due to interference with the putamen's established role in monitoring motor feedback. Caudate infusions, however, did not affect continuous errors, but did cause an upward trend in recurrent perseveration, possibly reflecting interference with the caudate's established role in spatial working memory and goal-directed planning. In contrast to variable array performance, while both caudate and putamen infusions impaired fixed array responding, the combined effects were not additive, suggesting possible competing roles. Infusions into either region individually, but not simultaneously, led to continuous perseveration. Recurrent perseveration in fixed arrays was caused by putamen, but not caudate, infusions. These results are consistent overall with a role of caudate in planning and flexible responding and the putamen in more rigid habitual or automatic responding.

The construction and use of cognitive maps in model-based control.

When making decisions, we sometimes rely on habit and at other times plan toward goals. Planning requires the construction and use of an internal representation of the environment, a cognitive map. How are these maps constructed, and how do they guide goal-directed decisions? We coupled a sequential decision-making task with a behavioral representational similarity analysis approach to examine how relationships between choice options change when people build a cognitive map of the task structure. We found that participants who encoded stronger higher-order relationships among choice options showed increased planning and better performance. These higher-order relationships were more strongly encoded among objects encountered in high-reward contexts, indicating a role for motivation during cognitive map construction. In contrast, lower-order relationships such as simple visual co-occurrence of objects did not predict goal-directed planning. These results show that the construction of cognitive maps is an active process, with motivation dictating the degree to which higher-order relationships are encoded and used for planning. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Defining adaption measures for organisational multi-agent systems

Designing adaptive multi-agent Systems (MAS) is a challenging development effort. A key point of adaptive systems is that they provide alternative options for acting and designers have to weight the number and elaboration of these alternatives. Here, we concentrate on organisation-oriented MAS and show that organisational models provide suitable means for identifying key measures of this adaptivity. We derive measures both from the static description of the organisation as well as from the induced adaptation dynamics grounded on a Markovian analysis of the behaviour. These measures allow for a goal-directed planning of adaption.

Semiology and determinants of apathy across neurodegenerative motor disorders: A comparison between amyotrophic lateral sclerosis, Parkinson's and Huntington's disease.

The semiology and determinants of apathy are largely unknown across amyotrophic lateral sclerosis (ALS), Parkinson's disease (PD), and Huntington's disease (HD), due to both motor and non-motor confounders. This study thus aimed at (1) profiling apathy in ALS, PD, and HD and (2) exploring its clinical determinants. Consecutive ALS (N = 99), PD (N = 73), and HD (N = 25) patients underwent a motor-free assessment of apathy (Dimensional Apathy Scale, DAS), global cognition, anxiety and depression. Function was assessed through disease-specific scales. The DAS was also completed by N = 101 healthy controls (HCs). Between-group comparisons on DAS scores were implemented by covarying for all applicable confounders. Predictive models on DAS scores were built through multiple, stepwise regressions. Parkinson's disease and HD, but not ALS, patients were more apathetic than HCs-with HD patients also selectively showing lower initiation and poorer goal-directed planning than HCs. Higher apathetic features were detected in PD and HD as compared to ALS. Education was a protective factor against apathy in ALS. Anxiety was a risk factor for global apathy in ALS, HD, and to a lesser extent, in PD, whereas, protective only toward affective disintegration in PD and ALS. Cognitive inefficiency was a risk factor toward apathy in both PD and ALS. Depression was a risk factor for executive-related apathy in PD. This study provides unprecedented insights into the heterogeneous semiology and determinants of apathy across ALS, PD, and HD via the DAS, in turn informing clinical practice and research.

The Relationship between Exercise and Mental Health Outcomes during the COVID-19 Pandemic: From the Perspective of Hope.

The unexpected outbreak of COVID-19 triggered fear and anxiety in the general population. Exercise was one of the most widely promoted methods to improve body function when socially restricted. This study aims to examine the role of exercise in relieving stressful mental health outcomes (anxiety and depressive symptoms) during the COVID-19 pandemic and explore the underlying mechanism from the perspective of hope, using a combination of goal-directed planning (pathways) and motivation (agency). A cross-sectional online survey recruiting 2390 Chinese participants was conducted during the COVID-19 pandemic in China. A series of questions and scales, including the self-designed exercise questionnaire, the Adult Dispositional Hope Scale, the Generalized Anxiety Disorder Scale-7 and the Patient Health Questionnaire-9, were used to measure exercise, hope, anxiety symptoms and depressive symptoms, respectively. A structural equation model was constructed to test the hypothesis that exercise benefits mental health outcomes through the mediating role of hope. Our results showed that exercise relieved stressful mental health outcomes via three paths: one direct path (β = −0.077, 95% CI = (−0.138, −0.017), p < 0.01), one indirect path through hope of pathways thinking (β = −0.046, 95% CI = (−0.064, −0.027), p < 0.001) and another indirect path through hope of agency thinking (β = −0.060, 95% CI = (−0.081, −0.039), p < 0.001). Our results showed that exercise could alleviate stressful mental health outcomes by promoting both hope of pathway thinking and agency thinking. It provided practical insights into psychological prevention and intervention by means of exercise during the COVID-19 pandemic.

Goal-Directed Planning and Goal Understanding by Extended Active Inference: Evaluation through Simulated and Physical Robot Experiments.

We show that goal-directed action planning and generation in a teleological framework can be formulated by extending the active inference framework. The proposed model, which is built on a variational recurrent neural network model, is characterized by three essential features. These are that (1) goals can be specified for both static sensory states, e.g., for goal images to be reached and dynamic processes, e.g., for moving around an object, (2) the model cannot only generate goal-directed action plans, but can also understand goals through sensory observation, and (3) the model generates future action plans for given goals based on the best estimate of the current state, inferred from past sensory observations. The proposed model is evaluated by conducting experiments on a simulated mobile agent as well as on a real humanoid robot performing object manipulation.

NeuroLISP: High-level symbolic programming with attractor neural networks

Despite significant improvements in contemporary machine learning, symbolic methods currently outperform artificial neural networks on tasks that involve compositional reasoning, such as goal-directed planning and logical inference. This illustrates a computational explanatory gap between cognitive and neurocomputational algorithms that obscures the neurobiological mechanisms underlying cognition and impedes progress toward human-level artificial intelligence. Because of the strong relationship between cognition and working memory control, we suggest that the cognitive abilities of contemporary neural networks are limited by biologically-implausible working memory systems that rely on persistent activity maintenance and/or temporal nonlocality. Here we present NeuroLISP, an attractor neural network that can represent and execute programs written in the LISP programming language. Unlike previous approaches to high-level programming with neural networks, NeuroLISP features a temporally-local working memory based on itinerant attractor dynamics, top-down gating, and fast associative learning, and implements several high-level programming constructs such as compositional data structures, scoped variable binding, and the ability to manipulate and execute programmatic expressions in working memory (i.e., programs can be treated as data). Our computational experiments demonstrate the correctness of the NeuroLISP interpreter, and show that it can learn non-trivial programs that manipulate complex derived data structures (multiway trees), perform compositional string manipulation operations (PCFG SET task), and implement high-level symbolic AI algorithms (first-order unification). We conclude that NeuroLISP is an effective neurocognitive controller that can replace the symbolic components of hybrid models, and serves as a proof of concept for further development of high-level symbolic programming in neural networks.

Emergence of Content-Agnostic Information Processing by a Robot Using Active Inference, Visual Attention, Working Memory, and Planning.

Generalization by learning is an essential cognitive competency for humans. For example, we can manipulate even unfamiliar objects and can generate mental images before enacting a preplan. How is this possible? Our study investigated this problem by revisiting our previous study (Jung, Matsumoto, & Tani, 2019), which examined the problem of vision-based, goal-directed planning by robots performing a task of block stacking. By extending the previous study, our work introduces a large network comprising dynamically interacting submodules, including visual working memory (VWMs), a visual attention module, and an executive network. The executive network predicts motor signals, visual images, and various controls for attention, as well as masking of visual information. The most significant difference from the previous study is that our current model contains an additional VWM. The entire network is trained by using predictive coding and an optimal visuomotor plan to achieve a given goal state is inferred using active inference. Results indicate that our current model performs significantly better than that used in Jung et al. (2019), especially when manipulating blocks with unlearned colors and textures. Simulation results revealed that the observed generalization was achieved because content-agnostic information processing developed through synergistic interaction between the second VWM and other modules during the course of learning, in which memorizing image contents and transforming them are dissociated. This letter verifies this claim by conducting both qualitative and quantitative analysis of simulation results.

Orbitofrontal cortex and learning predictions of state transitions.

The orbitofrontal cortex (OFC) has been implicated in goal-directed planning and model-based decision-making. One key prerequisite for model-based decision-making is learning the transition structure of the environment-the probabilities of transitioning from one environmental state to another. In this work, we investigated how the OFC might be involved in learning this transition structure, by using fMRI to assess OFC activity while humans experienced probabilistic cue-outcome transitions. We found that OFC activity was indeed correlated with behavioral measures of learning about transition structure. On a trial-by-trial basis, OFC activity was associated with subsequently increased expectation of the more probable outcome; that is, with subsequently more optimal cue-outcome predictions. Interestingly, this relationship was observed no matter what outcome occurred at the time of the OFC activity, and thus is inconsistent with an interpretation of the OFC activity as representing a "state prediction error" that would facilitate learning transitions via error-correcting mechanisms. Finally, OFC activity was related to more optimal predictions only for subsequent trials involving the same cue that was observed at the time of OFC activity-this relationship was not observed for subsequent trials involving a different cue. All together, these results indicate that the OFC is involved in updating or reinforcing a learned transition model on a trial-by-trial basis, specifically for the currently observed cue-outcome associations. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

A - 01 Baseline Executive Functioning and Mood in Older Adults before a Remote Physical Exercise Training Program

ObjectivePhysical exercise shows benefits to executive functioning (EF), a cognitive process that is relevant to goal-directed planning, application of complex rules, and dynamic control of action. However, many older adults have not engaged in exercise due to social distancing measures implemented to combat the Covid-19 pandemic. The present study aims to administer a remote physical exercise training program to older adults in Canada and examine its effects on EF and mood, beginning with the collection of the following baseline data.MethodsSo far, nine older adults (78% females, 66–78 years-old) have completed a remote assessment that examines current physical exercise engagement, Covid-related distress, general mood, and executive behavior. In addition, the assessment included computerized tasks measuring various aspects of EF. Pattern analyses were used to characterize trends in baseline data.Results67% of participants reported undergoing no moderate-vigorous physical exercise in a typical week. Compared to other participants, more individuals in this subgroup performed at a lower rate on a higher-order EF task (i.e., the Balloon Analogue Risk Task). Similarly, more participants who endorsed higher Covid-related psychological distress performed worse in the same task compared to others (40% in the lower third of performance vs. 0%), and also endorsed more difficulties with organization.ConclusionsThese results suggest that those who do not engage in physical exercise are potential candidates to experience the mental health and cognitive benefits of a physical exercise training program.

Set-shifting-related basal ganglia deformation as a novel familial marker of obsessive-compulsive disorder.

The symptoms of obsessive-compulsive disorder (OCD) are suggestive of cognitive rigidity, and previous work identified impaired flexible responding on set-shifting tasks in such patients. The basal ganglia are central to habit learning and are thought to be abnormal in OCD, contributing to inflexible, rigid habitual patterns of behaviour. Here, we demonstrate that increased cognitive inflexibility, indexed by poor performance on the set-shifting task, correlated with putamen morphology, and that patients and their asymptomatic relatives had common curvature abnormalities within this same structure. The association between the structure of the putamen and the extradimensional errors was found to be significantly familial in OCD proband-relative pairs. The data implicate changes in basal ganglia structure linked to cognitive inflexibility as a familial marker of OCD. This may reflect a predisposing heightened propensity toward habitual response patterns and deficits in goal-directed planning.

A global framework for a systemic view of brain modeling

The brain is a complex system, due to the heterogeneity of its structure, the diversity of the functions in which it participates and to its reciprocal relationships with the body and the environment. A systemic description of the brain is presented here, as a contribution to developing a brain theory and as a general framework where specific models in computational neuroscience can be integrated and associated with global information flows and cognitive functions. In an enactive view, this framework integrates the fundamental organization of the brain in sensorimotor loops with the internal and the external worlds, answering four fundamental questions (what, why, where and how). Our survival-oriented definition of behavior gives a prominent role to pavlovian and instrumental conditioning, augmented during phylogeny by the specific contribution of other kinds of learning, related to semantic memory in the posterior cortex, episodic memory in the hippocampus and working memory in the frontal cortex. This framework highlights that responses can be prepared in different ways, from pavlovian reflexes and habitual behavior to deliberations for goal-directed planning and reasoning, and explains that these different kinds of responses coexist, collaborate and compete for the control of behavior. It also lays emphasis on the fact that cognition can be described as a dynamical system of interacting memories, some acting to provide information to others, to replace them when they are not efficient enough, or to help for their improvement. Describing the brain as an architecture of learning systems has also strong implications in Machine Learning. Our biologically informed view of pavlovian and instrumental conditioning can be very precious to revisit classical Reinforcement Learning and provide a basis to ensure really autonomous learning.

A generative spiking neural-network model of goal-directed behaviour and one-step planning.

In mammals, goal-directed and planning processes support flexible behaviour used to face new situations that cannot be tackled through more efficient but rigid habitual behaviours. Within the Bayesian modelling approach of brain and behaviour, models have been proposed to perform planning as probabilistic inference but this approach encounters a crucial problem: explaining how such inference might be implemented in brain spiking networks. Recently, the literature has proposed some models that face this problem through recurrent spiking neural networks able to internally simulate state trajectories, the core function at the basis of planning. However, the proposed models have relevant limitations that make them biologically implausible, namely their world model is trained ‘off-line’ before solving the target tasks, and they are trained with supervised learning procedures that are biologically and ecologically not plausible. Here we propose two novel hypotheses on how brain might overcome these problems, and operationalise them in a novel architecture pivoting on a spiking recurrent neural network. The first hypothesis allows the architecture to learn the world model in parallel with its use for planning: to this purpose, a new arbitration mechanism decides when to explore, for learning the world model, or when to exploit it, for planning, based on the entropy of the world model itself. The second hypothesis allows the architecture to use an unsupervised learning process to learn the world model by observing the effects of actions. The architecture is validated by reproducing and accounting for the learning profiles and reaction times of human participants learning to solve a visuomotor learning task that is new for them. Overall, the architecture represents the first instance of a model bridging probabilistic planning and spiking-processes that has a degree of autonomy analogous to the one of real organisms.

The effects of selective serotonin reuptake inhibitors on brain functional networks during goal-directed planning in obsessive\u2013compulsive disorder

Whether brain network connectivity during goal-directed planning in patients with obsessive–compulsive disorder (OCD) is abnormal and restored by treatment with selective serotonin reuptake inhibitors (SSRIs) remains unknown. This study investigated whether the disrupted network connectivity during the Tower of London (ToL) planning task in medication-free OCD patients could be restored by SSRI treatment. Seventeen medication-free OCD patients and 21 matched healthy controls (HCs) underwent functional magnetic resonance imaging (fMRI) while performing the ToL task at baseline and again after 16 weeks of SSRI treatment. Internetwork connectivity was compared across the groups and treatment statuses (pretreatment versus posttreatment). At baseline, compared with the HCs, the OCD patients showed lower internetwork connectivity between the dorsal attention network and the default-mode network during the ToL planning task. After 16 weeks of SSRI treatment, the OCD patients showed improved clinical symptoms accompanied by normalized network connectivity, although their improved behavioral performance in the ToL task did not reach that of the HCs. Our findings support the conceptualization of OCD as a network disease characterized by an imbalance between brain networks during goal-directed planning and suggest that internetwork connectivity may serve as an early biomarker of the effects of SSRIs on goal-directed planning.

Dopamine, Updated: Reward Prediction Error and Beyond.

Dopamine neurons have been intensely studied for their roles in reinforcement learning. A dominant theory of how these neurons contribute to learning is through the encoding of a reward prediction error (RPE) signal. Recent advances in dopamine research have added nuance to RPE theory by incorporating the ideas of sensory prediction error, distributional encoding, and belief states. Further nuance is likely to be added shortly by convergent lines of research on dopamine neuron diversity. Finally, a major challenge is to reconcile RPE theory with other current theories of dopamine function to account for dopamine's role in movement, motivation, and goal-directed planning.

Computational Neural Mechanisms of Goal-Directed Planning and Problem Solving

The question of how animals and humans can solve arbitrary goal-driven problems remains open. Reinforcement learning (RL) methods have approached goal-directed control problems through model-based algorithms. However, RL focus on maximizing long-term reward is inconsistent with the psychological notion of planning to satisfy homeostatic drives, which involves setting goals first, then planning actions to achieve them. Optimal control theory suggests a solution: animals can learn a model of the world, learn where goals can be fulfilled, set a goal, and then act to minimize the difference between actual and desired world states. Here, we present a purely localist neural network model that can autonomously learn the structure of an environment and then achieve any arbitrary goal state in a changing environment without relearning reward values. The model, GOLSA, achieves this through a backwards spreading activation that propagates goal-values to an agent. The model elucidates how neural inhibitory mechanisms can support competition between goal representations, serving to push needs-based planning versus exploration. The model performs similar to humans in canonical revaluation tasks used to classify human and rodent behavior as goal-directed. The model revaluates optimal actions when goals, goal-values, world structure, and need to fulfill drive changes. The model also clarifies a number of issues inherent in other RL-based representations such as policy dependence in successor representations, while elucidating biological constraints such as the role of oscillations in gating information flow for learning versus action. Together, our proposed model suggests a biologically grounded framework for multi-step planning behaviors through consideration of how goal representations compete for behavioral expression in planning.

Goal-Directed Planning for Habituated Agents by Active Inference Using a Variational Recurrent Neural Network.

It is crucial to ask how agents can achieve goals by generating action plans using only partial models of the world acquired through habituated sensory-motor experiences. Although many existing robotics studies use a forward model framework, there are generalization issues with high degrees of freedom. The current study shows that the predictive coding (PC) and active inference (AIF) frameworks, which employ a generative model, can develop better generalization by learning a prior distribution in a low dimensional latent state space representing probabilistic structures extracted from well habituated sensory-motor trajectories. In our proposed model, learning is carried out by inferring optimal latent variables as well as synaptic weights for maximizing the evidence lower bound, while goal-directed planning is accomplished by inferring latent variables for maximizing the estimated lower bound. Our proposed model was evaluated with both simple and complex robotic tasks in simulation, which demonstrated sufficient generalization in learning with limited training data by setting an intermediate value for a regularization coefficient. Furthermore, comparative simulation results show that the proposed model outperforms a conventional forward model in goal-directed planning, due to the learned prior confining the search of motor plans within the range of habituated trajectories.

Comparison of the Association Between Goal-Directed Planning and Self-reported Compulsivity vs Obsessive-Compulsive Disorder Diagnosis

Dimensional definitions of transdiagnostic mental health problems have been suggested as an alternative to categorical diagnoses, having the advantage of capturing heterogeneity within diagnostic categories and similarity across them and bridging more naturally psychological and neural substrates. To examine whether a self-reported compulsivity dimension has a stronger association with goal-directed and related higher-order cognitive deficits compared with a diagnosis of obsessive-compulsive disorder (OCD). In this cross-sectional study, patients with OCD and/or generalized anxiety disorder (GAD) from across the United States completed a telephone-based diagnostic interview by a trained rater, internet-based cognitive testing, and self-reported clinical assessments from October 8, 2015, to October 1, 2017. Follow-up data were collected to test for replicability. Performance was measured on a test of goal-directed planning and cognitive flexibility (Wisconsin Card Sorting Test [WCST]) and a test of abstract reasoning. Clinical variables included DSM-5 diagnosis of OCD and GAD and 3 psychiatric symptom dimensions (general distress, compulsivity, and obsessionality) derived from a factor analysis. Of 285 individuals in the analysis (mean [SD] age, 32 [12] years; age range, 18-77 years; 219 [76.8%] female), 111 had OCD; 82, GAD; and 92, OCD and GAD. A diagnosis of OCD was not associated with goal-directed performance compared with GAD at baseline (β [SE], -0.02 [0.02]; P = .18). In contrast, a compulsivity dimension was negatively associated with goal-directed performance (β [SE], -0.05 [0.02]; P = .003). Results for abstract reasoning task and WCST mirrored this pattern; the compulsivity dimension was associated with abstract reasoning (β [SE], 2.99 [0.63]; P < .001) and several indicators of WCST performance (eg, categories completed: β [SE], -0.57 [0.09]; P < .001), whereas OCD diagnosis was not (abstract reasoning: β [SE], 0.39 [0.66]; P = .56; categories completed: β [SE], -0.09 [0.10]; P = .38). Other symptom dimensions relevant to OCD, obsessionality, and general distress had no reliable association with goal-directed performance, WCST, or abstract reasoning. Obsessionality had a positive association with requiring more trials to reach the first category on the WCST at baseline (β [SE], 2.92 [1.39]; P = .04), and general distress was associated with impaired goal-directed performance at baseline (β [SE],-0.04 [0.02]; P = .01). However, unlike the key results of this study, neither survived correction for multiple comparisons or was replicated at follow-up testing. Deficits in goal-directed planning in OCD may be more strongly associated with a compulsivity dimension than with OCD diagnosis. This result may have implications for research assessing the association between brain mechanisms and clinical manifestations and for understanding the structure of mental illness.

Hope and Physical Exercise: The Contributions of Hope, Self-Efficacy, and Optimism in Accounting for Variance in Exercise Frequency.

We examine the relationship of hope with exercise using Snyder's hope model, which defines hope as a combination of goal-directed planning (pathways) and motivation (agency). We surveyed 102 participants online via Amazon MTurk. The Adult Hope Scale and Goal-Specific Hope Scale assessed general hope and hope regarding exercise. The Leisure-Time Exercise Questionnaire assessed frequency of exercise and of any activity "long enough to work up a sweat." We also measured optimism, exercise self-efficacy, depression, anxiety, and stress. Because hope emphasizes planning and motivation (both likely needed in fitness regimens), we expected it to relate to exercise frequency after controlling for these other variables. Indeed, exercise goal-specific hope was correlated with frequency of exercise and activity long enough to sweat, though general hope was not. This relationship held after controlling for all other variables. Future research may be useful in investigating causality, including whether hope-based interventions result in increased exercise engagement.