Deep temporal models and active inference

'Snakes and ladders' in paleoanthropology: From cognitive surprise to skillfulness a million years ago

A paradigmatic account may suffice to explain behavioral evolution in early Homo. We propose a parsimonious account that (1) could explain a particular, frequently-encountered, archeological outcome of behavior in early Homo - namely, the fashioning of a Paleolithic stone 'handaxe' - from a biological theoretic perspective informed by the free energy principle (FEP); and that (2) regards instances of the outcome as postdictive or retrodictive, circumstantial corroboration. Our proposal considers humankind evolving as a self-organizing biological ecosystem at a geological time-scale. We offer a narrative treatment of this self-organization in terms of the FEP. Specifically, we indicate how 'cognitive surprises' could underwrite an evolving propensity in early Homo to express sporadic unorthodox or anomalous behavior. This co-evolutionary propensity has left us a legacy of Paleolithic artifacts that is reminiscent of a 'snakes and ladders' board game of appearances, disappearances, and reappearances of particular archeological traces of Paleolithic behavior. When detected in the Early and Middle Pleistocene record, anthropologists and archeologists often imagine evidence of unusual or novel behavior in terms of early humankind ascending the rungs of a figurative phylogenetic 'ladder' - as if these corresponded to progressive evolution of cognitive abilities that enabled incremental achievements of increasingly innovative technical prowess, culminating in the cognitive ascendancy of Homo sapiens. The conjecture overlooks a plausible likelihood that behavior by an individual who was atypical among her conspecifics could have been disregarded in a community of Hominina (for definition see Appendix 1) that failed to recognize, imagine, or articulate potential advantages of adopting hitherto unorthodox behavior. Such failure, as well as diverse fortuitous demographic accidents, would cause exceptional personal behavior to be ignored and hence unremembered. It could disappear by a pitfall, down a 'snake', as it were, in the figurative evolutionary board game; thereby causing a discontinuity in the evolution of human behavior that presents like an evolutionary puzzle. The puzzle discomforts some paleoanthropologists trained in the natural and life sciences. They often dismiss it, explaining it away with such self-justifying conjectures as that, maybe, separate paleospecies of Homo differentially possessed different cognitive abilities, which, supposedly, could account for the presence or absence in the Pleistocene archeological record of traces of this or that behavioral outcome or skill. We argue that an alternative perspective - that inherits from the FEP and an individual's 'active inference' about its surroundings and of its own responses - affords a prosaic, deflationary, and parsimonious way to account for appearances, disappearances, and reappearances of particular behavioral outcomes and skills of early humankind.

Synergizing habits and goals with variational Bayes

Behaving efficiently and flexibly is crucial for biological and artificial embodied agents. Behavior is generally classified into two types: habitual (fast but inflexible), and goal-directed (flexible but slow). While these two types of behaviors are typically considered to be managed by two distinct systems in the brain, recent studies have revealed a more sophisticated interplay between them. We introduce a theoretical framework using variational Bayesian theory, incorporating a Bayesian intention variable. Habitual behavior depends on the prior distribution of intention, computed from sensory context without goal-specification. In contrast, goal-directed behavior relies on the goal-conditioned posterior distribution of intention, inferred through variational free energy minimization. Assuming that an agent behaves using a synergized intention, our simulations in vision-based sensorimotor tasks explain the key properties of their interaction as observed in experiments. Our work suggests a fresh perspective on the neural mechanisms of habits and goals, shedding light on future research in decision making.

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.

Neurophysiological signatures of prediction in language: A critical review of anticipatory negativities

Recent event-related potential (ERP) studies in language comprehension converge in finding anticipatory negativities preceding words or word segments that can be pre-activated based on either sentence contexts or phonological cues. We review these findings from different paradigms in the light of evidence from other cognitive domains in which slow negative potentials have long been associated with anticipatory processes and discuss their potential underlying mechanisms. We propose that this family of anticipatory negativities captures common mechanisms associated with the pre-activation of linguistic information both within words and within sentences. Future studies could utilize these anticipatory negativities in combination with other, well-established ERPs, to simultaneously track prediction-related processes emerging at different time intervals (before and after the perception of pre-activated input) and with distinct time courses (shorter-lived and longer-lived cognitive operations).

An Overview of the Free Energy Principle and Related Research

The free energy principle and its corollary, the active inference framework, serve as theoretical foundations in the domain of neuroscience, explaining the genesis of intelligent behavior. This principle states that the processes of perception, learning, and decision making-within an agent-are all driven by the objective of "minimizing free energy," evincing the following behaviors: learning and employing a generative model of the environment to interpret observations, thereby achieving perception, and selecting actions to maintain a stable preferred state and minimize the uncertainty about the environment, thereby achieving decision making. This fundamental principle can be used to explain how the brain processes perceptual information, learns about the environment, and selects actions. Two pivotal tenets are that the agent employs a generative model for perception and planning and that interaction with the world (and other agents) enhances the performance of the generative model and augments perception. With the evolution of control theory and deep learning tools, agents based on the FEP have been instantiated in various ways across different domains, guiding the design of a multitude of generative models and decision-making algorithms. This letter first introduces the basic concepts of the FEP, followed by its historical development and connections with other theories of intelligence, and then delves into the specific application of the FEP to perception and decision making, encompassing both low-dimensional simple situations and high-dimensional complex situations. It compares the FEP with model-based reinforcement learning to show that the FEP provides a better objective function. We illustrate this using numerical studies of Dreamer3 by adding expected information gain into the standard objective function. In a complementary fashion, existing reinforcement learning, and deep learning algorithms can also help implement the FEP-based agents. Finally, we discuss the various capabilities that agents need to possess in complex environments and state that the FEP can aid agents in acquiring these capabilities.

To be or not to be: The active inference of suicide

Suicide presents an apparent paradox as a behavior whose motivation is not obvious since its outcome is non-existence and cannot be experienced. To address this paradox, we propose to frame suicide in the integrated theory of stress and active inference. We present an active inference-based cognitive model of suicide as a type of stress response hanging in cognitive balance between predicting self-preservation and self-destruction. In it, self-efficacy emerges as a meta-cognitive regulator that can bias the model toward either survival or suicide. The model suggests conditions under which cognitive homeostasis can override physiological homeostasis in motivating self-destruction. We also present a model proto-suicidal behavior, programmed cell death (apoptosis), in active inference terms to illustrate how an active inference model of self-destruction can be embodied in molecular mechanisms and to offer a hypothesis on another puzzle of suicide: why only humans among brain-endowed animals are known to practice it.

I like therefore I can, and I can therefore I like: the role of self-efficacy and affect in active inference of allostasis

Active inference (AIF) is a theory of the behavior of information-processing open dynamic systems. It describes them as generative models (GM) generating inferences on the causes of sensory input they receive from their environment. Based on these inferences, GMs generate predictions about sensory input. The discrepancy between a prediction and the actual input results in prediction error. GMs then execute action policies predicted to minimize the prediction error. The free-energy principle provides a rationale for AIF by stipulating that information-processing open systems must constantly minimize their free energy (through suppressing the cumulative prediction error) to avoid decay. The theory of homeostasis and allostasis has a similar logic. Homeostatic set points are expectations of living organisms. Discrepancies between set points and actual states generate stress. For optimal functioning, organisms avoid stress by preserving homeostasis. Theories of AIF and homeostasis have recently converged, with AIF providing a formal account for homeo- and allostasis. In this paper, we present bacterial chemotaxis as molecular AIF, where mutual constraints by extero- and interoception play an essential role in controlling bacterial behavior supporting homeostasis. Extending this insight to the brain, we propose a conceptual model of the brain homeostatic GM, in which we suggest partition of the brain GM into cognitive and physiological homeostatic GMs. We outline their mutual regulation as well as their integration based on the free-energy principle. From this analysis, affect and self-efficacy emerge as the main regulators of the cognitive homeostatic GM. We suggest fatigue and depression as target neurocognitive phenomena for studying the neural mechanisms of such regulation.

Hierarchical network structure as the source of hierarchical dynamics (power-law frequency spectra) in living and non-living systems: How state-trait continua (body plans, personalities) emerge from first principles in biophysics

Living systems are hierarchical control systems that display a small world network structure. In such structures, many smaller clusters are nested within fewer larger ones, producing a fractal-like structure with a 'power-law' cluster size distribution (a mereology). Just like their structure, the dynamics of living systems shows fractal-like qualities: the timeseries of inner message passing and overt behavior contain high frequencies or 'states' (treble) that are nested within lower frequencies or 'traits' (bass), producing a power-law frequency spectrum that is known as a 'state-trait continuum' in the behavioral sciences. Here, we argue that the power-law dynamics of living systems results from their power-law network structure: organisms 'vertically encode' the deep spatiotemporal structure of their (anticipated) environments, to the effect that many small clusters near the base of the hierarchy produce high frequency signal changes and fewer larger clusters at its top produce ultra-low frequencies. Such ultra-low frequencies exert a tonic regulatory pressure that produces morphological as well as behavioral traits (i.e., body plans and personalities). Nested-modular structure causes higher frequencies to be embedded within lower frequencies, producing a power-law state-trait continuum. At the heart of such dynamics lies the need for efficient energy dissipation through networks of coupled oscillators, which also governs the dynamics of non-living systems (e.q., earthquakes, stock market fluctuations). Since hierarchical structure produces hierarchical dynamics, the development and collapse of hierarchical structure (e.g., during maturation and disease) should leave specific traces in system dynamics (shifts in lower frequencies, i.e. morphological and behavioral traits) that may serve as early warning signs to system failure. The applications of this idea range from (bio)physics and phylogenesis to ontogenesis and clinical medicine.

The technology, opportunities, and challenges of Synthetic Biological Intelligence

Integrating neural cultures developed through synthetic biology methods with digital computing has enabled the early development of Synthetic Biological Intelligence (SBI). Recently, key studies have emphasized the advantages of biological neural systems in some information processing tasks. However, neither the technology behind this early development, nor the potential ethical opportunities or challenges, have been explored in detail yet. Here, we review the key aspects that facilitate the development of SBI and explore potential applications. Considering these foreseeable use cases, various ethical implications are proposed. Ultimately, this work aims to provide a robust framework to structure ethical considerations to ensure that SBI technology can be both researched and applied responsibly.

An Introduction to Predictive Processing Models of Perception and Decision-Making

The predictive processing framework includes a broad set of ideas, which might be articulated and developed in a variety of ways, concerning how the brain may leverage predictive models when implementing perception, cognition, decision-making, and motor control. This article provides an up-to-date introduction to the two most influential theories within this framework: predictive coding and active inference. The first half of the paper (Sections 2-5) reviews the evolution of predictive coding, from early ideas about efficient coding in the visual system to a more general model encompassing perception, cognition, and motor control. The theory is characterized in terms of the claims it makes at Marr's computational, algorithmic, and implementation levels of description, and the conceptual and mathematical connections between predictive coding, Bayesian inference, and variational free energy (a quantity jointly evaluating model accuracy and complexity) are explored. The second half of the paper (Sections 6-8) turns to recent theories of active inference. Like predictive coding, active inference models assume that perceptual and learning processes minimize variational free energy as a means of approximating Bayesian inference in a biologically plausible manner. However, these models focus primarily on planning and decision-making processes that predictive coding models were not developed to address. Under active inference, an agent evaluates potential plans (action sequences) based on their expected free energy (a quantity that combines anticipated reward and information gain). The agent is assumed to represent the world as a partially observable Markov decision process with discrete time and discrete states. Current research applications of active inference models are described, including a range of simulation work, as well as studies fitting models to empirical data. The paper concludes by considering future research directions that will be important for further development of both models.

An active inference model of hierarchical action understanding, learning and imitation

We advance a novel active inference model of the cognitive processing that underlies the acquisition of a hierarchical action repertoire and its use for observation, understanding and imitation. We illustrate the model in four simulations of a tennis learner who observes a teacher performing tennis shots, forms hierarchical representations of the observed actions, and imitates them. Our simulations show that the agent's oculomotor activity implements an active information sampling strategy that permits inferring the kinematic aspects of the observed movement, which lie at the lowest level of the action hierarchy. In turn, this low-level kinematic inference supports higher-level inferences about deeper aspects of the observed actions: proximal goals and intentions. Finally, the inferred action representations can steer imitative responses, but interfere with the execution of different actions. Our simulations show that hierarchical active inference provides a unified account of action observation, understanding, learning and imitation and helps explain the neurobiological underpinnings of visuomotor cognition, including the multiple routes for action understanding in the dorsal and ventral streams and mirror mechanisms.

Cessations of consciousness in meditation: Advancing a scientific understanding of nirodha samāpatti

Absence of consciousness can occur due to a concussion, anesthetization, intoxication, epileptic seizure, or other fainting/syncope episode caused by lack of blood flow to the brain. However, some meditation practitioners also report that it is possible to undergo a total absence of consciousness during meditation, lasting up to 7 days, and that these "cessations" can be consistently induced. One form of extended cessation (i.e., nirodha samāpatti) is thought to be different from sleep because practitioners are said to be completely impervious to external stimulation. That is, they cannot be 'woken up' from the cessation state as one might be from a dream. Cessations are also associated with the absence of any time experience or tiredness, and are said to involve a stiff rather than a relaxed body. Emergence from meditation-induced cessations is said to have profound effects on subsequent cognition and experience (e.g., resulting in a sudden sense of clarity, openness, and possibly insights). In this paper, we briefly outline the historical context for cessation events, present preliminary data from two labs, set a research agenda for their study, and provide an initial framework for understanding what meditation induced cessation may reveal about the mind and brain. We conclude by integrating these so-called nirodha and nirodha samāpatti experiences-as they are known in classical Buddhism-into current cognitive-neurocomputational and active inference frameworks of meditation.

A neural active inference model of perceptual-motor learning

The active inference framework (AIF) is a promising new computational framework grounded in contemporary neuroscience that can produce human-like behavior through reward-based learning. In this study, we test the ability for the AIF to capture the role of anticipation in the visual guidance of action in humans through the systematic investigation of a visual-motor task that has been well-explored-that of intercepting a target moving over a ground plane. Previous research demonstrated that humans performing this task resorted to anticipatory changes in speed intended to compensate for semi-predictable changes in target speed later in the approach. To capture this behavior, our proposed "neural" AIF agent uses artificial neural networks to select actions on the basis of a very short term prediction of the information about the task environment that these actions would reveal along with a long-term estimate of the resulting cumulative expected free energy. Systematic variation revealed that anticipatory behavior emerged only when required by limitations on the agent's movement capabilities, and only when the agent was able to estimate accumulated free energy over sufficiently long durations into the future. In addition, we present a novel formulation of the prior mapping function that maps a multi-dimensional world-state to a uni-dimensional distribution of free-energy/reward. Together, these results demonstrate the use of AIF as a plausible model of anticipatory visually guided behavior in humans.

Time-consciousness in computational phenomenology: a temporal analysis of active inference

Time plays a significant role in science and everyday life. Despite being experienced as a continuous flow, computational models of consciousness are typically restricted to a sequential temporal structure. This difference poses a serious challenge for computational phenomenology-a novel field combining phenomenology and computational modelling. By analysing the temporal structure of the active inference framework, we show that an integrated continuity of time can be achieved by merging Husserlian temporality with a sequential order of time. We also show that a Markov blanket of the present moment integrates past and future moments of both subjective temporality and objective time in an asynchronous manner. By applying the integrated continuity, it is clear that active inference makes use of both subjective temporality and objective time in an integrated fashion. We conclude that active inference, on a temporal note, qualifies as a computational model for phenomenological investigations.

In the Body’s Eye: The computational anatomy of interoceptive inference

A growing body of evidence highlights the intricate linkage of exteroceptive perception to the rhythmic activity of the visceral body. In parallel, interoceptive inference theories of affective perception and self-consciousness are on the rise in cognitive science. However, thus far no formal theory has emerged to integrate these twin domains; instead, most extant work is conceptual in nature. Here, we introduce a formal model of cardiac active inference, which explains how ascending cardiac signals entrain exteroceptive sensory perception and uncertainty. Through simulated psychophysics, we reproduce the defensive startle reflex and commonly reported effects linking the cardiac cycle to affective behaviour. We further show that simulated ‘interoceptive lesions’ blunt affective expectations, induce psychosomatic hallucinations, and exacerbate biases in perceptual uncertainty. Through synthetic heart-rate variability analyses, we illustrate how the balance of arousal-priors and visceral prediction errors produces idiosyncratic patterns of physiological reactivity. Our model thus offers a roadmap for computationally phenotyping disordered brain-body interaction.

Understanding interactions between the brain and the body has become a topic of increased interest in computational neuroscience and psychiatry. A particular question here concerns how visceral, homeostatic rhythms such as the heart beat influence sensory, affective, and cognitive processing. To better understand these and other oscillatory brain-body interactions, we here introduce a novel computational model of interoceptive inference in which a synthetic agent’s perceptual beliefs are coupled to the rhythm of the heart. Our model both helps to explain emerging empirical data indicating that perceptual inference depends upon beat-to-beat heart rhythms, and can be used to better quantify intra- and inter-individual differences in heart-brain coupling. Using proof-of-principle simulations, we demonstrate how future empirical works could utilize our model to better understand and stratify disorders of interoception and brain-body interaction.

Learned uncertainty: The free energy principle in anxiety

Generalized anxiety disorder is among the world’s most prevalent psychiatric disorders and often manifests as persistent and difficult to control apprehension. Despite its prevalence, there is no integrative, formal model of how anxiety and anxiety disorders arise. Here, we offer a perspective derived from the free energy principle; one that shares similarities with established constructs such as learned helplessness. Our account is simple: anxiety can be formalized as learned uncertainty. A biological system, having had persistent uncertainty in its past, will expect uncertainty in its future, irrespective of whether uncertainty truly persists. Despite our account’s intuitive simplicity—which can be illustrated with the mere flip of a coin—it is grounded within the free energy principle and hence situates the formation of anxiety within a broader explanatory framework of biological self-organization and self-evidencing. We conclude that, through conceptualizing anxiety within a framework of working generative models, our perspective might afford novel approaches in the clinical treatment of anxiety and its key symptoms.

Rational arbitration between statistics and rules in human sequence processing

Detecting and learning temporal regularities is essential to accurately predict the future. A long-standing debate in cognitive science concerns the existence in humans of a dissociation between two systems, one for handling statistical regularities governing the probabilities of individual items and their transitions, and another for handling deterministic rules. Here, to address this issue, we used finger tracking to continuously monitor the online build-up of evidence, confidence, false alarms and changes-of-mind during sequence processing. All these aspects of behaviour conformed tightly to a hierarchical Bayesian inference model with distinct hypothesis spaces for statistics and rules, yet linked by a single probabilistic currency. Alternative models based either on a single statistical mechanism or on two non-commensurable systems were rejected. Our results indicate that a hierarchical Bayesian inference mechanism, capable of operating over distinct hypothesis spaces for statistics and rules, underlies the human capability for sequence processing.

Variability and task-responsiveness of electrophysiological dynamics: scale-free stability and oscillatory flexibility

Cortical oscillations and scale-free neural activity are thought to influence a variety of cognitive functions, but their differential relationships to neural stability and flexibility has never been investigated. Based on the existing literature, we hypothesize that scale-free and oscillatory processes in the brain exhibit different trade-offs between stability and flexibility; specifically, cortical oscillations may reflect variable, task-responsive aspects of brain activity, while scale-free activity is proposed to reflect a more stable and task-unresponsive aspect. We test this hypothesis using data from two large-scale MEG studies (HCP: n = 89; CamCAN: n = 195), operationalizing stability and flexibility by task-responsiveness and spontaneous intra-subject variability in resting state. We demonstrate that the power-law exponent of scale-free activity is a highly stable parameter, which responds little to external cognitive demands and shows minimal spontaneous fluctuations over time. In contrast, oscillatory power, particularly in the alpha range (8-13 Hz), responds strongly to tasks and exhibits comparatively large spontaneous fluctuations over time. In sum, our data support differential roles for oscillatory and scale-free activity in the brain with respect to neural stability and flexibility. This result carries implications for criticality-based theories of scale-free activity, state-trait models of variability, and homeostatic views of the brain with regulated variables vs. effectors.

A step-by-step tutorial on active inference and its application to empirical data

The active inference framework, and in particular its recent formulation as a partially observable Markov decision process (POMDP), has gained increasing popularity in recent years as a useful approach for modeling neurocognitive processes. This framework is highly general and flexible in its ability to be customized to model any cognitive process, as well as simulate predicted neuronal responses based on its accompanying neural process theory. It also affords both simulation experiments for proof of principle and behavioral modeling for empirical studies. However, there are limited resources that explain how to build and run these models in practice, which limits their widespread use. Most introductions assume a technical background in programming, mathematics, and machine learning. In this paper we offer a step-by-step tutorial on how to build POMDPs, run simulations using standard MATLAB routines, and fit these models to empirical data. We assume a minimal background in programming and mathematics, thoroughly explain all equations, and provide exemplar scripts that can be customized for both theoretical and empirical studies. Our goal is to provide the reader with the requisite background knowledge and practical tools to apply active inference to their own research. We also provide optional technical sections and multiple appendices, which offer the interested reader additional technical details. This tutorial should provide the reader with all the tools necessary to use these models and to follow emerging advances in active inference research.

Model Reduction Through Progressive Latent Space Pruning in Deep Active Inference

Although still not fully understood, sleep is known to play an important role in learning and in pruning synaptic connections. From the active inference perspective, this can be cast as learning parameters of a generative model and Bayesian model reduction, respectively. In this article, we show how to reduce dimensionality of the latent space of such a generative model, and hence model complexity, in deep active inference during training through a similar process. While deep active inference uses deep neural networks for state space construction, an issue remains in that the dimensionality of the latent space must be specified beforehand. We investigate two methods that are able to prune the latent space of deep active inference models. The first approach functions similar to sleep and performs model reduction post hoc. The second approach is a novel method which is more similar to reflection, operates during training and displays “aha” moments when the model is able to reduce latent space dimensionality. We show for two well-known simulated environments that model performance is retained in the first approach and only diminishes slightly in the second approach. We also show that reconstructions from a real world example are indistinguishable before and after reduction. We conclude that the most important difference constitutes a trade-off between training time and model performance in terms of accuracy and the ability to generalize, via minimization of model complexity.

Temporo-spatial Theory of Consciousness (TTC) - Bridging the gap of neuronal activity and phenomenal states

Consciousness and its neural mechanisms remain a mystery. Current neuroscientific theories focus predominantly on the external input/stimulus and the associated stimulus-related activity during conscious contents. Despite all progress, we encounter two gaps: (i) a gap between spontaneous and stimulus-related activity; (ii) a gap between neuronal and phenomenal features. A novel, different, and unique approach, Temporo-spatial theory of consciousness (TTC) aims to bridge both gaps. The TTC focuses on the brain's spontaneous activity and how its spatial topography and temporal dynamic shape stimulus-related activity and resurface in the corresponding spatial and temporal features of consciousness, i.e., 'common currency'. The TTC introduces four temporo-spatial mechanisms: expansion, globalization, alignment, and nestedness. These are associated with distinct dimensions of consciousness including phenomenal content, access, form/structure, and level/state, respectively. Following up on the first introduction of the TTC in 2017, we review updates, further develop these temporo-spatial mechanisms, and postulate specific neurophenomenal hypotheses. We conclude that the TTC offers a viable approach for (i) linking spontaneous and stimulus-related activity in conscious states; (ii) determining specific neuronal and neurophenomenal mechanisms for the distinct dimensions of consciousness; (iii) an integrative and unifying framework of different neuroscientific theories of consciousness; and (iv) offers novel empirically grounded conceptual assumptions about the biological and ontological nature of consciousness and its relation to the brain.

World model learning and inference

Understanding information processing in the brain-and creating general-purpose artificial intelligence-are long-standing aspirations of scientists and engineers worldwide. The distinctive features of human intelligence are high-level cognition and control in various interactions with the world including the self, which are not defined in advance and are vary over time. The challenge of building human-like intelligent machines, as well as progress in brain science and behavioural analyses, robotics, and their associated theoretical formalisations, speaks to the importance of the world-model learning and inference. In this article, after briefly surveying the history and challenges of internal model learning and probabilistic learning, we introduce the free energy principle, which provides a useful framework within which to consider neuronal computation and probabilistic world models. Next, we showcase examples of human behaviour and cognition explained under that principle. We then describe symbol emergence in the context of probabilistic modelling, as a topic at the frontiers of cognitive robotics. Lastly, we review recent progress in creating human-like intelligence by using novel probabilistic programming languages. The striking consensus that emerges from these studies is that probabilistic descriptions of learning and inference are powerful and effective ways to create human-like artificial intelligent machines and to understand intelligence in the context of how humans interact with their world.

Self, Me and I in the repertoire of spontaneously occurring altered states of Selfhood: eight neurophenomenological case study reports

This study investigates eight case reports of spontaneously emerging, brief episodes of vivid altered states of Selfhood (ASoSs) that occurred during mental exercise in six long-term meditators by using a neurophenomenological electroencephalography (EEG) approach. In agreement with the neurophenomenological methodology, first-person reports were used to identify such spontaneous ASoSs and to guide the neural analysis, which involved the estimation of three operational modules of the brain self-referential network (measured by EEG operational synchrony). The result of such analysis demonstrated that the documented ASoSs had unique neurophenomenological profiles, where several aspects or components of Selfhood (measured neurophysiologically and phenomenologically) are affected and expressed differently, but still in agreement with the neurophysiological three-dimensional construct model of the complex experiential Selfhood proposed in our earlier work (Fingelkurts et al. in Conscious Cogn 86:103031. 10.1016/j.concog.2020.103031, 2020).

Realizing Active Inference in Variational Message Passing: The Outcome-Blind Certainty Seeker

Active inference is a state-of-the-art framework in neuroscience that offers a unified theory of brain function. It is also proposed as a framework for planning in AI. Unfortunately, the complex mathematics required to create new models can impede application of active inference in neuroscience and AI research. This letter addresses this problem by providing a complete mathematical treatment of the active inference framework in discrete time and state spaces and the derivation of the update equations for any new model. We leverage the theoretical connection between active inference and variational message passing as described by John Winn and Christopher M. Bishop in 2005. Since variational message passing is a well-defined methodology for deriving Bayesian belief update equations, this letter opens the door to advanced generative models for active inference. We show that using a fully factorized variational distribution simplifies the expected free energy, which furnishes priors over policies so that agents seek unambiguous states. Finally, we consider future extensions that support deep tree searches for sequential policy optimization based on structure learning and belief propagation.

Meta-control of the exploration-exploitation dilemma emerges from probabilistic inference over a hierarchy of time scales

Cognitive control is typically understood as a set of mechanisms that enable humans to reach goals that require integrating the consequences of actions over longer time scales. Importantly, using routine behaviour or making choices beneficial only at short time scales would prevent one from attaining these goals. During the past two decades, researchers have proposed various computational cognitive models that successfully account for behaviour related to cognitive control in a wide range of laboratory tasks. As humans operate in a dynamic and uncertain environment, making elaborate plans and integrating experience over multiple time scales is computationally expensive. Importantly, it remains poorly understood how uncertain consequences at different time scales are integrated into adaptive decisions. Here, we pursue the idea that cognitive control can be cast as active inference over a hierarchy of time scales, where inference, i.e., planning, at higher levels of the hierarchy controls inference at lower levels. We introduce the novel concept of meta-control states, which link higher-level beliefs with lower-level policy inference. Specifically, we conceptualize cognitive control as inference over these meta-control states, where solutions to cognitive control dilemmas emerge through surprisal minimisation at different hierarchy levels. We illustrate this concept using the exploration-exploitation dilemma based on a variant of a restless multi-armed bandit task. We demonstrate that beliefs about contexts and meta-control states at a higher level dynamically modulate the balance of exploration and exploitation at the lower level of a single action. Finally, we discuss the generalisation of this meta-control concept to other control dilemmas.

The Sync-Fire/deSync model: Modelling the reactivation of dynamic memories from cortical alpha oscillations

We propose a neural network model to explore how humans can learn and accurately retrieve temporal sequences, such as melodies, movies, or other dynamic content. We identify target memories by their neural oscillatory signatures, as shown in recent human episodic memory paradigms. Our model comprises three plausible components for the binding of temporal content, where each component imposes unique limitations on the encoding and representation of that content. A cortical component actively represents sequences through the disruption of an intrinsically generated alpha rhythm, where a desynchronisation marks information-rich operations as the literature predicts. A binding component converts each event into a discrete index, enabling repetitions through a sparse encoding of events. A timing component - consisting of an oscillatory "ticking clock" made up of hierarchical synfire chains - discretely indexes a moment in time. By encoding the absolute timing between discretised events, we show how one can use cortical desynchronisations to dynamically detect unique temporal signatures as they are reactivated in the brain. We validate this model by simulating a series of events where sequences are uniquely identifiable by analysing phasic information, as several recent EEG/MEG studies have shown. As such, we show how one can encode and retrieve complete episodic memories where the quality of such memories is modulated by the following: alpha gate keepers to content representation; binding limitations that induce a blink in temporal perception; and nested oscillations that provide preferential learning phases in order to temporally sequence events.

Whence the Expected Free Energy?

The expected free energy (EFE) is a central quantity in the theory of active inference. It is the quantity that all active inference agents are mandated to minimize through action, and its decomposition into extrinsic and intrinsic value terms is key to the balance of exploration and exploitation that active inference agents evince. Despite its importance, the mathematical origins of this quantity and its relation to the variational free energy (VFE) remain unclear. In this letter, we investigate the origins of the EFE in detail and show that it is not simply "the free energy in the future." We present a functional that we argue is the natural extension of the VFE but actively discourages exploratory behavior, thus demonstrating that exploration does not directly follow from free energy minimization into the future. We then develop a novel objective, the free energy of the expected future (FEEF), which possesses both the epistemic component of the EFE and an intuitive mathematical grounding as the divergence between predicted and desired futures.

Active inference on discrete state-spaces: A synthesis

Active inference is a normative principle underwriting perception, action, planning, decision-making and learning in biological or artificial agents. From its inception, its associated process theory has grown to incorporate complex generative models, enabling simulation of a wide range of complex behaviours. Due to successive developments in active inference, it is often difficult to see how its underlying principle relates to process theories and practical implementation. In this paper, we try to bridge this gap by providing a complete mathematical synthesis of active inference on discrete state-space models. This technical summary provides an overview of the theory, derives neuronal dynamics from first principles and relates this dynamics to biological processes. Furthermore, this paper provides a fundamental building block needed to understand active inference for mixed generative models; allowing continuous sensations to inform discrete representations. This paper may be used as follows: to guide research towards outstanding challenges, a practical guide on how to implement active inference to simulate experimental behaviour, or a pointer towards various in-silico neurophysiological responses that may be used to make empirical predictions.

The predictive global neuronal workspace: A formal active inference model of visual consciousness

The global neuronal workspace (GNW) model has inspired over two decades of hypothesis-driven research on the neural basis of consciousness. However, recent studies have reported findings that are at odds with empirical predictions of the model. Further, the macro-anatomical focus of current GNW research has limited the specificity of predictions afforded by the model. In this paper we present a neurocomputational model - based on Active Inference - that captures central architectural elements of the GNW and is able to address these limitations. The resulting 'predictive global workspace' casts neuronal dynamics as approximating Bayesian inference, allowing precise, testable predictions at both the behavioural and neural levels of description. We report simulations demonstrating the model's ability to reproduce: 1) the electrophysiological and behavioural results observed in previous studies of inattentional blindness; and 2) the previously introduced four-way taxonomy predicted by the GNW, which describes the relationship between consciousness, attention, and sensory signal strength. We then illustrate how our model can reconcile/explain (apparently) conflicting findings, extend the GNW taxonomy to include the influence of prior expectations, and inspire novel paradigms to test associated behavioural and neural predictions.

Better Safe Than Sorry: A Common Signature of General Vulnerability for Psychopathology

Several labels, such as neuroticism, negative emotionality, and dispositional negativity, indicate a broad dimension of psychopathology. However, largely separate, often disorder-specific research lines have developed that focus on different cognitive and affective characteristics that are associated with this dimension, such as perseverative cognition (worry, rumination), reduced autobiographical memory specificity, compromised fear learning, and enhanced somatic-symptom reporting. In this article, we present a theoretical perspective within a predictive-processing framework in which we trace these phenotypically different characteristics back to a common underlying “better-safe-than-sorry” processing strategy. This implies information processing that tends to be low in sensory-perceptual detail, which allows threat-related categorical priors to dominate conscious experience and for chronic uncertainty/surprise because of a stagnated error-reduction process. This common information-processing strategy has beneficial effects in the short term but important costs in the long term. From this perspective, we suggest that the phenomenally distinct cognitive and affective psychopathological characteristics mentioned above represent the same basic processing heuristic of the brain and are only different in relation to the particular type of information involved (e.g., in working memory, in autobiographical memory, in the external and internal world). Clinical implications of this view are discussed.

The science of consciousness does not need another theory, it needs a minimal unifying model

This article discusses a hypothesis recently put forward by Kanai et al., according to which information generation constitutes a functional basis of, and a sufficient condition for, consciousness. Information generation involves the ability to compress and subsequently decompress information, potentially after a temporal delay and adapted to current purposes. I will argue that information generation should not be regarded as a sufficient condition for consciousness, but could serve as what I will call a “minimal unifying model of consciousness.” A minimal unifying model (MUM) specifies at least one necessary feature of consciousness, characterizes it in a determinable way, and shows that it is entailed by (many) existing theories of consciousness. Information generation fulfills these requirements. A MUM of consciousness is useful, because it unifies existing theories of consciousness by highlighting their common assumptions, while enabling further developments from which empirical predictions can be derived. Unlike existing theories (which probably contain at least some false assumptions), a MUM is thus likely to be an adequate model of consciousness, albeit at a relatively general level. Assumptions embodied in such a model are less informative than assumptions made by more specific theories and hence function more in the way of guiding principles. Still, they enable further refinements, in line with new empirical results and broader theoretical and evolutionary considerations. This also allows developing the model in ways that facilitate more specific claims and predictions.

A challenge for predictive coding: Representational or experiential diversity?

To become a unifying theory of brain function, predictive processing (PP) must accommodate its rich representational diversity. Gilead et al. claim such diversity requires a multi-process theory, and thus is out of reach for PP, which postulates a universal canonical computation. We contend this argument and instead propose that PP fails to account for the experiential level of representations.

Learning action-oriented models through active inference

Converging theories suggest that organisms learn and exploit probabilistic models of their environment. However, it remains unclear how such models can be learned in practice. The open-ended complexity of natural environments means that it is generally infeasible for organisms to model their environment comprehensively. Alternatively, action-oriented models attempt to encode a parsimonious representation of adaptive agent-environment interactions. One approach to learning action-oriented models is to learn online in the presence of goal-directed behaviours. This constrains an agent to behaviourally relevant trajectories, reducing the diversity of the data a model need account for. Unfortunately, this approach can cause models to prematurely converge to sub-optimal solutions, through a process we refer to as a bad-bootstrap. Here, we exploit the normative framework of active inference to show that efficient action-oriented models can be learned by balancing goal-oriented and epistemic (information-seeking) behaviours in a principled manner. We illustrate our approach using a simple agent-based model of bacterial chemotaxis. We first demonstrate that learning via goal-directed behaviour indeed constrains models to behaviorally relevant aspects of the environment, but that this approach is prone to sub-optimal convergence. We then demonstrate that epistemic behaviours facilitate the construction of accurate and comprehensive models, but that these models are not tailored to any specific behavioural niche and are therefore less efficient in their use of data. Finally, we show that active inference agents learn models that are parsimonious, tailored to action, and which avoid bad bootstraps and sub-optimal convergence. Critically, our results indicate that models learned through active inference can support adaptive behaviour in spite of, and indeed because of, their departure from veridical representations of the environment. Our approach provides a principled method for learning adaptive models from limited interactions with an environment, highlighting a route to sample efficient learning algorithms.

Simulating Emotions: An Active Inference Model of Emotional State Inference and Emotion Concept Learning

The ability to conceptualize and understand one's own affective states and responses - or "Emotional awareness" (EA) - is reduced in multiple psychiatric populations; it is also positively correlated with a range of adaptive cognitive and emotional traits. While a growing body of work has investigated the neurocognitive basis of EA, the neurocomputational processes underlying this ability have received limited attention. Here, we present a formal Active Inference (AI) model of emotion conceptualization that can simulate the neurocomputational (Bayesian) processes associated with learning about emotion concepts and inferring the emotions one is feeling in a given moment. We validate the model and inherent constructs by showing (i) it can successfully acquire a repertoire of emotion concepts in its "childhood", as well as (ii) acquire new emotion concepts in synthetic "adulthood," and (iii) that these learning processes depend on early experiences, environmental stability, and habitual patterns of selective attention. These results offer a proof of principle that cognitive-emotional processes can be modeled formally, and highlight the potential for both theoretical and empirical extensions of this line of research on emotion and emotional disorders.

An Active Inference Approach to Dissecting Reasons for Nonadherence to Antidepressants

BACKGROUND

Antidepressant medication adherence is among the most important problems in health care worldwide. Interventions designed to increase adherence have largely failed, pointing toward a critical need to better understand the underlying decision-making processes that contribute to adherence. A computational decision-making model that integrates empirical data with a fundamental action selection principle could be pragmatically useful in 1) making individual-level predictions about adherence and 2) providing an explanatory framework that improves our understanding of nonadherence.

METHODS

Here we formulated a partially observable Markov decision process model based on the active inference framework that can simulate several processes that plausibly influence adherence decisions.

RESULTS

Using model simulations of the day-to-day decisions to take a prescribed selective serotonin reuptake inhibitor, we show that several distinct parameters in the model can influence adherence decisions in predictable ways. These parameters include differences in policy depth (i.e., how far into the future one considers when deciding), decision uncertainty, beliefs about the predictability (stochasticity) of symptoms, beliefs about the magnitude and time course of symptom reductions and side effects, and strength of medication-taking habits that one has acquired.

CONCLUSIONS

Clarifying these influential factors will be an important first step toward empirically determining which factors are contributing to nonadherence to antidepressants in individual patients. The model can also be seamlessly extended to simulate adherence to other medications (by incorporating the known symptom reduction and side effect trajectories of those medications), with the potential promise of identifying which treatments may be best suited for different patients.

The hierarchically mechanistic mind: an evolutionary systems theory of the human brain, cognition, and behavior

The purpose of this review was to integrate leading paradigms in psychology and neuroscience with a theory of the embodied, situated human brain, called the Hierarchically Mechanistic Mind (HMM). The HMM describes the brain as a complex adaptive system that functions to minimize the entropy of our sensory and physical states via action-perception cycles generated by hierarchical neural dynamics. First, we review the extant literature on the hierarchical structure of the brain. Next, we derive the HMM from a broader evolutionary systems theory that explains neural structure and function in terms of dynamic interactions across four nested levels of biological causation (i.e., adaptation, phylogeny, ontogeny, and mechanism). We then describe how the HMM aligns with a global brain theory in neuroscience called the free-energy principle, leveraging this theory to mathematically formulate neural dynamics across hierarchical spatiotemporal scales. We conclude by exploring the implications of the HMM for psychological inquiry.

The hierarchically mechanistic mind: A free-energy formulation of the human psyche

This article presents a unifying theory of the embodied, situated human brain called the Hierarchically Mechanistic Mind (HMM). The HMM describes the brain as a complex adaptive system that actively minimises the decay of our sensory and physical states by producing self-fulfilling action-perception cycles via dynamical interactions between hierarchically organised neurocognitive mechanisms. This theory synthesises the free-energy principle (FEP) in neuroscience with an evolutionary systems theory of psychology that explains our brains, minds, and behaviour by appealing to Tinbergen's four questions: adaptation, phylogeny, ontogeny, and mechanism. After leveraging the FEP to formally define the HMM across different spatiotemporal scales, we conclude by exploring its implications for theorising and research in the sciences of the mind and behaviour.

Hierarchical Predictive Information Is Channeled by Asymmetric Oscillatory Activity

Predictive coding and neural oscillations are two descriptive levels of brain functioning whose overlap is not yet understood. Chao et al. (2018) now show that hierarchical predictive coding is instantiated by asymmetric information channeling in the γ and α/β oscillatory ranges.

REBUS and the Anarchic Brain: Toward a Unified Model of the Brain Action of Psychedelics

This paper formulates the action of psychedelics by integrating the free-energy principle and entropic brain hypothesis. We call this formulation relaxed beliefs under psychedelics (REBUS) and the anarchic brain, founded on the principle that-via their entropic effect on spontaneous cortical activity-psychedelics work to relax the precision of high-level priors or beliefs, thereby liberating bottom-up information flow, particularly via intrinsic sources such as the limbic system. We assemble evidence for this model and show how it can explain a broad range of phenomena associated with the psychedelic experience. With regard to their potential therapeutic use, we propose that psychedelics work to relax the precision weighting of pathologically overweighted priors underpinning various expressions of mental illness. We propose that this process entails an increased sensitization of high-level priors to bottom-up signaling (stemming from intrinsic sources), and that this heightened sensitivity enables the potential revision and deweighting of overweighted priors. We end by discussing further implications of the model, such as that psychedelics can bring about the revision of other heavily weighted high-level priors, not directly related to mental health, such as those underlying partisan and/or overly-confident political, religious, and/or philosophical perspectives. SIGNIFICANCE STATEMENT: Psychedelics are capturing interest, with efforts underway to bring psilocybin therapy to marketing authorisation and legal access within a decade, spearheaded by the findings of a series of phase 2 trials. In this climate, a compelling unified model of how psychedelics alter brain function to alter consciousness would have appeal. Towards this end, we have sought to integrate a leading model of global brain function, hierarchical predictive coding, with an often-cited model of the acute action of psychedelics, the entropic brain hypothesis. The resulting synthesis states that psychedelics work to relax high-level priors, sensitising them to liberated bottom-up information flow, which, with the right intention, care provision and context, can help guide and cultivate the revision of entrenched pathological priors.

Parameters as Trait Indicators: Exploring a Complementary Neurocomputational Approach to Conceptualizing and Measuring Trait Differences in Emotional Intelligence

Current assessments of trait emotional intelligence (EI) rely on self-report inventories. While this approach has seen considerable success, a complementary approach allowing objective assessment of EI-relevant traits would provide some potential advantages. Among others, one potential advantage is that it would aid in emerging efforts to assess the brain basis of trait EI, where self-reported competency levels do not always match real-world behavior. In this paper, we review recent experimental paradigms in computational cognitive neuroscience (CCN), which allow behavioral estimates of individual differences in range of parameter values within computational models of neurocognitive processes. Based on this review, we illustrate how several of these parameters appear to correspond well to EI-relevant traits (i.e., differences in mood stability, stress vulnerability, self-control, and flexibility, among others). In contrast, although estimated objectively, these parameters do not correspond well to the optimal performance abilities assessed within competing “ability models” of EI. We suggest that adapting this approach from CCN—by treating parameter value estimates as objective trait EI measures—could (1) provide novel research directions, (2) aid in characterizing the neural basis of trait EI, and (3) offer a promising complementary assessment method.

Multiscale integration: beyond internalism and externalism

We present a multiscale integrationist interpretation of the boundaries of cognitive systems, using the Markov blanket formalism of the variational free energy principle. This interpretation is intended as a corrective for the philosophical debate over internalist and externalist interpretations of cognitive boundaries; we stake out a compromise position. We first survey key principles of new radical (extended, enactive, embodied) views of cognition. We then describe an internalist interpretation premised on the Markov blanket formalism. Having reviewed these accounts, we develop our positive multiscale account. We argue that the statistical seclusion of internal from external states of the system-entailed by the existence of a Markov boundary-can coexist happily with the multiscale integration of the system through its dynamics. Our approach does not privilege any given boundary (whether it be that of the brain, body, or world), nor does it argue that all boundaries are equally prescient. We argue that the relevant boundaries of cognition depend on the level being characterised and the explanatory interests that guide investigation. We approach the issue of how and where to draw the boundaries of cognitive systems through a multiscale ontology of cognitive systems, which offers a multidisciplinary research heuristic for cognitive science.