Active inference and robot control: a case study

Active Inference for Learning and Development in Embodied Neuromorphic Agents

Taking inspiration from humans can help catalyse embodied AI solutions for important real-world applications. Current human-inspired tools include neuromorphic systems and the developmental approach to learning. However, this developmental neurorobotics approach is currently lacking important frameworks for human-like computation and learning. We propose that human-like computation is inherently embodied, with its interface to the world being neuromorphic, and its learning processes operating across different timescales. These constraints necessitate a unified framework: active inference, underpinned by the free energy principle (FEP). Herein, we describe theoretical and empirical support for leveraging this framework in embodied neuromorphic agents with autonomous mental development. We additionally outline current implementation approaches (including toolboxes) and challenges, and we provide suggestions for next steps to catalyse this important field.

Active Inference With Empathy Mechanism for Socially Behaved Artificial Agents in Diverse Situations

This article proposes a method for an artificial agent to behave in a social manner. Although defining proper social behavior is difficult because it differs from situation to situation, the agent following the proposed method adaptively behaves appropriately in each situation by empathizing with the surrounding others. The proposed method is achieved by incorporating empathy into active inference. We evaluated the proposed method regarding control of autonomous mobile robots in diverse situations. From the evaluation results, an agent controlled by the proposed method could behave more adaptively socially than an agent controlled by the standard active inference in the diverse situations. In the case of two agents, the agent controlled with the proposed method behaved in a social way that reduced the other agent's travel distance by 13.7% and increased the margin between the agents by 25.8%, even though it increased the agent's travel distance by 8.2%. Also, the agent controlled with the proposed method behaved more socially when it was surrounded by altruistic others but less socially when it was surrounded by selfish others.

The empirical status of predictive coding and active inference

Research on predictive processing models has focused largely on two specific algorithmic theories: Predictive Coding for perception and Active Inference for decision-making. While these interconnected theories possess broad explanatory potential, they have only recently begun to receive direct empirical evaluation. Here, we review recent studies of Predictive Coding and Active Inference with a focus on evaluating the degree to which they are empirically supported. For Predictive Coding, we find that existing empirical evidence offers modest support. However, some positive results can also be explained by alternative feedforward (e.g., feature detection-based) models. For Active Inference, most empirical studies have focused on fitting these models to behavior as a means of identifying and explaining individual or group differences. While Active Inference models tend to explain behavioral data reasonably well, there has not been a focus on testing empirical validity of active inference theory per se, which would require formal comparison to other models (e.g., non-Bayesian or model-free reinforcement learning models). This review suggests that, while promising, a number of specific research directions are still necessary to evaluate the empirical adequacy and explanatory power of these algorithms.

Deep kinematic inference affords efficient and scalable control of bodily movements

Performing goal-directed movements requires mapping goals from extrinsic (workspace-relative) to intrinsic (body-relative) coordinates and then to motor signals. Mainstream approaches based on optimal control realize the mappings by minimizing cost functions, which is computationally demanding. Instead, active inference uses generative models to produce sensory predictions, which allows a cheaper inversion to the motor signals. However, devising generative models to control complex kinematic chains like the human body is challenging. We introduce an active inference architecture that affords a simple but effective mapping from extrinsic to intrinsic coordinates via inference and easily scales up to drive complex kinematic chains. Rich goals can be specified in both intrinsic and extrinsic coordinates using attractive or repulsive forces. The proposed model reproduces sophisticated bodily movements and paves the way for computationally efficient and biologically plausible control of actuated systems.

Active Vision in Binocular Depth Estimation: A Top-Down Perspective

Depth estimation is an ill-posed problem; objects of different shapes or dimensions, even if at different distances, may project to the same image on the retina. Our brain uses several cues for depth estimation, including monocular cues such as motion parallax and binocular cues such as diplopia. However, it remains unclear how the computations required for depth estimation are implemented in biologically plausible ways. State-of-the-art approaches to depth estimation based on deep neural networks implicitly describe the brain as a hierarchical feature detector. Instead, in this paper we propose an alternative approach that casts depth estimation as a problem of active inference. We show that depth can be inferred by inverting a hierarchical generative model that simultaneously predicts the eyes' projections from a 2D belief over an object. Model inversion consists of a series of biologically plausible homogeneous transformations based on Predictive Coding principles. Under the plausible assumption of a nonuniform fovea resolution, depth estimation favors an active vision strategy that fixates the object with the eyes, rendering the depth belief more accurate. This strategy is not realized by first fixating on a target and then estimating the depth; instead, it combines the two processes through action-perception cycles, with a similar mechanism of the saccades during object recognition. The proposed approach requires only local (top-down and bottom-up) message passing, which can be implemented in biologically plausible neural circuits.

Understanding brain functional architecture through robotics

Robotics is increasingly seen as a useful test bed for computational models of the brain functional architecture underlying animal behavior. We provide an overview of past and current work, focusing on probabilistic and dynamical models, including approaches premised on the free energy principle, situating this endeavor in relation to evidence that the brain constitutes a layered control system. We argue that future neurorobotic models should integrate multiple neurobiological constraints and be hybrid in nature.

Interpersonal attunement in social interactions: from collective psychophysiology to inter-personalized psychiatry and beyond

In this article, we analyse social interactions, drawing on diverse points of views, ranging from dialectics, second-person neuroscience and enactivism to dynamical systems, active inference and machine learning. To this end, we define interpersonal attunement as a set of multi-scale processes of building up and materializing social expectations-put simply, anticipating and interacting with others and ourselves. While cultivating and negotiating common ground, via communication and culture-building activities, are indispensable for the survival of the individual, the relevant multi-scale mechanisms have been largely considered in isolation. Here, collective psychophysiology, we argue, can lend itself to the fine-tuned analysis of social interactions, without neglecting the individual. On the other hand, an interpersonal mismatch of expectations can lead to a breakdown of communication and social isolation known to negatively affect mental health. In this regard, we review psychopathology in terms of interpersonal misattunement, conceptualizing psychiatric disorders as disorders of social interaction, to describe how individual mental health is inextricably linked to social interaction. By doing so, we foresee avenues for an inter-personalized psychiatry, which moves from a static spectrum of disorders to a dynamic relational space, focusing on how the multi-faceted processes of social interaction can help to promote mental health. This article is part of the theme issue 'Concepts in interaction: social engagement and inner experiences'.

Turn-Taking Mechanisms in Imitative Interaction: Robotic Social Interaction Based on the Free Energy Principle

This study explains how the leader-follower relationship and turn-taking could develop in a dyadic imitative interaction by conducting robotic simulation experiments based on the free energy principle. Our prior study showed that introducing a parameter during the model training phase can determine leader and follower roles for subsequent imitative interactions. The parameter is defined as w, the so-called meta-prior, and is a weighting factor used to regulate the complexity term versus the accuracy term when minimizing the free energy. This can be read as sensory attenuation, in which the robot's prior beliefs about action are less sensitive to sensory evidence. The current extended study examines the possibility that the leader-follower relationship shifts depending on changes in w during the interaction phase. We identified a phase space structure with three distinct types of behavioral coordination using comprehensive simulation experiments with sweeps of w of both robots during the interaction. Ignoring behavior in which the robots follow their own intention was observed in the region in which both ws were set to large values. One robot leading, followed by the other robot was observed when one w was set larger and the other was set smaller. Spontaneous, random turn-taking between the leader and the follower was observed when both ws were set at smaller or intermediate values. Finally, we examined a case of slowly oscillating w in anti-phase between the two agents during the interaction. The simulation experiment resulted in turn-taking in which the leader-follower relationship switched during determined sequences, accompanied by periodic shifts of ws. An analysis using transfer entropy found that the direction of information flow between the two agents also shifted along with turn-taking. Herein, we discuss qualitative differences between random/spontaneous turn-taking and agreed-upon sequential turn-taking by reviewing both synthetic and empirical studies.

Flexible intentions: An Active Inference theory

We present a normative computational theory of how the brain may support visually-guided goal-directed actions in dynamically changing environments. It extends the Active Inference theory of cortical processing according to which the brain maintains beliefs over the environmental state, and motor control signals try to fulfill the corresponding sensory predictions. We propose that the neural circuitry in the Posterior Parietal Cortex (PPC) compute flexible intentions-or motor plans from a belief over targets-to dynamically generate goal-directed actions, and we develop a computational formalization of this process. A proof-of-concept agent embodying visual and proprioceptive sensors and an actuated upper limb was tested on target-reaching tasks. The agent behaved correctly under various conditions, including static and dynamic targets, different sensory feedbacks, sensory precisions, intention gains, and movement policies; limit conditions were individuated, too. Active Inference driven by dynamic and flexible intentions can thus support goal-directed behavior in constantly changing environments, and the PPC might putatively host its core intention mechanism. More broadly, the study provides a normative computational basis for research on goal-directed behavior in end-to-end settings and further advances mechanistic theories of active biological systems.

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

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.

Accessing Active Inference Theory through Its Implicit and Deliberative Practice in Human Organizations

Active inference theory (AIT) is a corollary of the free-energy principle, which formalizes cognition of living system’s autopoietic organization. AIT comprises specialist terminology and mathematics used in theoretical neurobiology. Yet, active inference is common practice in human organizations, such as private companies, public institutions, and not-for-profits. Active inference encompasses three interrelated types of actions, which are carried out to minimize uncertainty about how organizations will survive. The three types of action are updating work beliefs, shifting work attention, and/or changing how work is performed. Accordingly, an alternative starting point for grasping active inference, rather than trying to understand AIT specialist terminology and mathematics, is to reflect upon lived experience. In other words, grasping active inference through autoethnographic research. In this short communication paper, accessing AIT through autoethnography is explained in terms of active inference in existing organizational practice (implicit active inference), new organizational methodologies that are informed by AIT (deliberative active inference), and combining implicit and deliberative active inference. In addition, these autoethnographic options for grasping AIT are related to generative learning.

Trust as Extended Control: Human-Machine Interactions as Active Inference

In order to interact seamlessly with robots, users must infer the causes of a robot's behavior-and be confident about that inference (and its predictions). Hence, trust is a necessary condition for human-robot collaboration (HRC). However, and despite its crucial role, it is still largely unknown how trust emerges, develops, and supports human relationship to technological systems. In the following paper we review the literature on trust, human-robot interaction, HRC, and human interaction at large. Early models of trust suggest that it is a trade-off between benevolence and competence; while studies of human to human interaction emphasize the role of shared behavior and mutual knowledge in the gradual building of trust. We go on to introduce a model of trust as an agent' best explanation for reliable sensory exchange with an extended motor plant or partner. This model is based on the cognitive neuroscience of active inference and suggests that, in the context of HRC, trust can be casted in terms of virtual control over an artificial agent. Interactive feedback is a necessary condition to the extension of the trustor's perception-action cycle. This model has important implications for understanding human-robot interaction and collaboration-as it allows the traditional determinants of human trust, such as the benevolence and competence attributed to the trustee, to be defined in terms of hierarchical active inference, while vulnerability can be described in terms of information exchange and empowerment. Furthermore, this model emphasizes the role of user feedback during HRC and suggests that boredom and surprise may be used in personalized interactions as markers for under and over-reliance on the system. The description of trust as a sense of virtual control offers a crucial step toward grounding human factors in cognitive neuroscience and improving the design of human-centered technology. Furthermore, we examine the role of shared behavior in the genesis of trust, especially in the context of dyadic collaboration, suggesting important consequences for the acceptability and design of human-robot collaborative systems.

The Acquisition of Culturally Patterned Attention Styles Under Active Inference

This paper presents an active inference based simulation study of visual foraging. The goal of the simulation is to show the effect of the acquisition of culturally patterned attention styles on cognitive task performance, under active inference. We show how cultural artefacts like antique vase decorations drive cognitive functions such as perception, action and learning, as well as task performance in a simple visual discrimination task. We thus describe a new active inference based research pipeline that future work may employ to inquire on deep guiding principles determining the manner in which material culture drives human thought, by building and rebuilding our patterns of attention.

Bidirectional interaction between visual and motor generative models using Predictive Coding and Active Inference

In this work, we build upon the Active Inference (AIF) and Predictive Coding (PC) frameworks to propose a neural architecture comprising a generative model for sensory prediction, and a distinct generative model for motor trajectories. We highlight how sequences of sensory predictions can act as rails guiding learning, control and online adaptation of motor trajectories. We furthermore inquire the effects of bidirectional interactions between the motor and the visual modules. The architecture is tested on the control of a simulated robotic arm learning to reproduce handwritten letters.

Robot navigation as hierarchical active inference

Localization and mapping has been a long standing area of research, both in neuroscience, to understand how mammals navigate their environment, as well as in robotics, to enable autonomous mobile robots. In this paper, we treat navigation as inferring actions that minimize (expected) variational free energy under a hierarchical generative model. We find that familiar concepts like perception, path integration, localization and mapping naturally emerge from this active inference formulation. Moreover, we show that this model is consistent with models of hippocampal functions, and can be implemented in silico on a real-world robot. Our experiments illustrate that a robot equipped with our hierarchical model is able to generate topologically consistent maps, and correct navigation behaviour is inferred when a goal location is provided to the system.

Predictive Processing in Cognitive Robotics: A Review

Predictive processing has become an influential framework in cognitive sciences. This framework turns the traditional view of perception upside down, claiming that the main flow of information processing is realized in a top-down, hierarchical manner. Furthermore, it aims at unifying perception, cognition, and action as a single inferential process. However, in the related literature, the predictive processing framework and its associated schemes, such as predictive coding, active inference, perceptual inference, and free-energy principle, tend to be used interchangeably. In the field of cognitive robotics, there is no clear-cut distinction on which schemes have been implemented and under which assumptions. In this letter, working definitions are set with the main aim of analyzing the state of the art in cognitive robotics research working under the predictive processing framework as well as some related nonrobotic models. The analysis suggests that, first, research in both cognitive robotics implementations and nonrobotic models needs to be extended to the study of how multiple exteroceptive modalities can be integrated into prediction error minimization schemes. Second, a relevant distinction found here is that cognitive robotics implementations tend to emphasize the learning of a generative model, while in nonrobotics models, it is almost absent. Third, despite the relevance for active inference, few cognitive robotics implementations examine the issues around control and whether it should result from the substitution of inverse models with proprioceptive predictions. Finally, limited attention has been placed on precision weighting and the tracking of prediction error dynamics. These mechanisms should help to explore more complex behaviors and tasks in cognitive robotics research under the predictive processing framework.

Neurorobotic Models of Neurological Disorders: A Mini Review

Modeling is widely used in biomedical research to gain insights into pathophysiology and treatment of neurological disorders but existing models, such as animal models and computational models, are limited in generalizability to humans and are restricted in the scope of possible experiments. Robotics offers a potential complementary modeling platform, with advantages such as embodiment and physical environmental interaction yet with easily monitored and adjustable parameters. In this review, we discuss the different types of models used in biomedical research and summarize the existing neurorobotics models of neurological disorders. We detail the pertinent findings of these robot models which would not have been possible through other modeling platforms. We also highlight the existing limitations in a wider uptake of robot models for neurological disorders and suggest future directions for the field.

Active Inference: Applicability to Different Types of Social Organization Explained through Reference to Industrial Engineering and Quality Management

Active inference is a physics of life process theory of perception, action and learning that is applicable to natural and artificial agents. In this paper, active inference theory is related to different types of practice in social organization. Here, the term social organization is used to clarify that this paper does not encompass organization in biological systems. Rather, the paper addresses active inference in social organization that utilizes industrial engineering, quality management, and artificial intelligence alongside human intelligence. Social organization referred to in this paper can be in private companies, public institutions, other for-profit or not-for-profit organizations, and any combination of them. The relevance of active inference theory is explained in terms of variational free energy, prediction errors, generative models, and Markov blankets. Active inference theory is most relevant to the social organization of work that is highly repetitive. By contrast, there are more challenges involved in applying active inference theory for social organization of less repetitive endeavors such as one-of-a-kind projects. These challenges need to be addressed in order for active inference to provide a unifying framework for different types of social organization employing human and artificial intelligence.

Learning Generative State Space Models for Active Inference

In this paper we investigate the active inference framework as a means to enable autonomous behavior in artificial agents. Active inference is a theoretical framework underpinning the way organisms act and observe in the real world. In active inference, agents act in order to minimize their so called free energy, or prediction error. Besides being biologically plausible, active inference has been shown to solve hard exploration problems in various simulated environments. However, these simulations typically require handcrafting a generative model for the agent. Therefore we propose to use recent advances in deep artificial neural networks to learn generative state space models from scratch, using only observation-action sequences. This way we are able to scale active inference to new and challenging problem domains, whilst still building on the theoretical backing of the free energy principle. We validate our approach on the mountain car problem to illustrate that our learnt models can indeed trade-off instrumental value and ambiguity. Furthermore, we show that generative models can also be learnt using high-dimensional pixel observations, both in the OpenAI Gym car racing environment and a real-world robotic navigation task. Finally we show that active inference based policies are an order of magnitude more sample efficient than Deep Q Networks on RL tasks.

Mental Effort and Information-Processing Costs Are Inversely Related to Global Brain Free Energy During Visual Categorization

Mental effort is a neurocognitive process that reflects the controlled expenditure of psychological information-processing resources during perception, cognition, and action. There is a practical need to operationalize and measure mental effort in order to minimize detrimental effects of mental fatigue on real-world human performance. Previous research has identified several neurocognitive indices of mental effort, but these indices are indirect measures that are also sensitive to experimental demands or general factors such as sympathetic arousal. The present study investigated a potential direct neurocognitive index of mental effort based in theories where bounded rational decision makers (realized as embodied brains) are modeled as generalized thermodynamic systems. This index is called free energy, an information-theoretic system property of the brain that reflects the difference between the brain's current and predicted states. Theory predicts that task-related differences in a decision makers' free energy are inversely related to information-processing costs related to task decisions. The present study tested this prediction by quantifying global brain free energy from electroencephalographic (EEG) measures of human brain function. EEG signals were recorded while participants engaged in two visual categorization tasks in which categorization decisions resulted from the allocation of different levels of mental information processing resources. A novel method was developed to quantify brain free energy from machine learning classification of EEG trials. Participant information-processing resource costs were estimated via computational analysis of behavior, whereas the subjective expression of mental effort was estimated via participant ratings of mental workload. Following theoretical predictions, task-related differences in brain free energy negatively correlated with increased allocation of information-processing resource costs. These brain free energy differences were smaller for the visual categorization task that required a greater versus lesser allocation of information-processing resources. Ratings of mental workload were positively correlated with information-processing resource costs, and negatively correlated with global brain free energy differences, only for the categorization task requiring the larger amount of information-processing resource costs. These findings support theoretical thermodynamic approaches to decision making and provide the first empirical evidence of a relationship between mental effort, brain free energy, and neurocognitive information-processing.

Simulating Active Inference Processes by Message Passing

The free energy principle (FEP) offers a variational calculus-based description for how biological agents persevere through interactions with their environment. Active inference (AI) is a corollary of the FEP, which states that biological agents act to fulfill prior beliefs about preferred future observations (target priors). Purposeful behavior then results from variational free energy minimization with respect to a generative model of the environment with included target priors. However, manual derivations for free energy minimizing algorithms on custom dynamic models can become tedious and error-prone. While probabilistic programming (PP) techniques enable automatic derivation of inference algorithms on free-form models, full automation of AI requires specialized tools for inference on dynamic models, together with the description of an experimental protocol that governs the interaction between the agent and its simulated environment. The contributions of the present paper are two-fold. Firstly, we illustrate how AI can be automated with the use of ForneyLab, a recent PP toolbox that specializes in variational inference on flexibly definable dynamic models. More specifically, we describe AI agents in a dynamic environment as probabilistic state space models (SSM) and perform inference for perception and control in these agents by message passing on a factor graph representation of the SSM. Secondly, we propose a formal experimental protocol for simulated AI. We exemplify how this protocol leads to goal-directed behavior for flexibly definable AI agents in two classical RL examples, namely the Bayesian thermostat and the mountain car parking problems.

The Active Inference Approach to Ecological Perception: General Information Dynamics for Natural and Artificial Embodied Cognition

The emerging neurocomputational vision of humans as embodied, ecologically embedded, social agents—who shape and are shaped by their environment—offers a golden opportunity to revisit and revise ideas about the physical and information-theoretic underpinnings of life, mind, and consciousness itself. In particular, the active inference framework (AIF) makes it possible to bridge connections from computational neuroscience and robotics/AI to ecological psychology and phenomenology, revealing common underpinnings and overcoming key limitations. AIF opposes the mechanistic to the reductive, while staying fully grounded in a naturalistic and information-theoretic foundation, using the principle of free energy minimization. The latter provides a theoretical basis for a unified treatment of particles, organisms, and interactive machines, spanning from the inorganic to organic, non-life to life, and natural to artificial agents. We provide a brief introduction to AIF, then explore its implications for evolutionary theory, ecological psychology, embodied phenomenology, and robotics/AI research. We conclude the paper by considering implications for machine consciousness.