Perceptions as hypotheses: saccades as experiments

The behavioural preview effect with faces is susceptible to statistical regularities: Evidence for predictive processing across the saccade

The world around us appears stable and continuous despite saccadic eye movements. This apparent visual stability is achieved by trans-saccadic perception leading at the behavioural level to preview effects: performance in processing a foveal stimulus is better if the stimulus remained unchanged (valid) compared to when it changed (invalid) during the saccade that brought it into focus. Trans-saccadic perception is known to predictively adapt to the statistics of the environment. Here, we asked whether the behavioural preview effect shows the same characteristics, employing a between-participants training design. Participants made saccades to faces which could change their orientation (upright/inverted) during the saccade. In addition, the post-saccadic face was slightly tilted and participants reported this tilt upon fixation. In a training phase, one group of participants conducted only invalid trials whereas another group conducted only valid trials. In a subsequent test phase with 50% valid and 50% invalid trials, we measured the preview effect. Invalid training reduced the preview effect. With a mixed-model analysis, we could show how this training effect gradually declines in the course of the test phase. These results show that the behavioural preview effect adapts to the statistics of the environment suggesting that it results from predictive processes.

Tea With Milk? A Hierarchical Generative Framework of Sequential Event Comprehension

To make sense of the world around us, we must be able to segment a continual stream of sensory inputs into discrete events. In this review, I propose that in order to comprehend events, we engage hierarchical generative models that "reverse engineer" the intentions of other agents as they produce sequential action in real time. By generating probabilistic predictions for upcoming events, generative models ensure that we are able to keep up with the rapid pace at which perceptual inputs unfold. By tracking our certainty about other agents' goals and the magnitude of prediction errors at multiple temporal scales, generative models enable us to detect event boundaries by inferring when a goal has changed. Moreover, by adapting flexibly to the broader dynamics of the environment and our own comprehension goals, generative models allow us to optimally allocate limited resources. Finally, I argue that we use generative models not only to comprehend events but also to produce events (carry out goal-relevant sequential action) and to continually learn about new events from our surroundings. Taken together, this hierarchical generative framework provides new insights into how the human brain processes events so effortlessly while highlighting the fundamental links between event comprehension, production, and learning.

A Bayesian brain model of adaptive behavior: an application to the Wisconsin Card Sorting Task

Adaptive behavior emerges through a dynamic interaction between cognitive agents and changing environmental demands. The investigation of information processing underlying adaptive behavior relies on controlled experimental settings in which individuals are asked to accomplish demanding tasks whereby a hidden regularity or an abstract rule has to be learned dynamically. Although performance in such tasks is considered as a proxy for measuring high-level cognitive processes, the standard approach consists in summarizing observed response patterns by simple heuristic scoring measures. With this work, we propose and validate a new computational Bayesian model accounting for individual performance in the Wisconsin Card Sorting Test (WCST), a renowned clinical tool to measure set-shifting and deficient inhibitory processes on the basis of environmental feedback. We formalize the interaction between the task's structure, the received feedback, and the agent's behavior by building a model of the information processing mechanisms used to infer the hidden rules of the task environment. Furthermore, we embed the new model within the mathematical framework of the Bayesian Brain Theory (BBT), according to which beliefs about hidden environmental states are dynamically updated following the logic of Bayesian inference. Our computational model maps distinct cognitive processes into separable, neurobiologically plausible, information-theoretic constructs underlying observed response patterns. We assess model identification and expressiveness in accounting for meaningful human performance through extensive simulation studies. We then validate the model on real behavioral data in order to highlight the utility of the proposed model in recovering cognitive dynamics at an individual level. We highlight the potentials of our model in decomposing adaptive behavior in the WCST into several information-theoretic metrics revealing the trial-by-trial unfolding of information processing by focusing on two exemplary individuals whose behavior is examined in depth. Finally, we focus on the theoretical implications of our computational model by discussing the mapping between BBT constructs and functional neuroanatomical correlates of task performance. We further discuss the empirical benefit of recovering the assumed dynamics of information processing for both clinical and research practices, such as neurological assessment and model-based neuroscience.

Deep Active Inference and Scene Construction

Adaptive agents must act in intrinsically uncertain environments with complex latent structure. Here, we elaborate a model of visual foraging-in a hierarchical context-wherein agents infer a higher-order visual pattern (a "scene") by sequentially sampling ambiguous cues. Inspired by previous models of scene construction-that cast perception and action as consequences of approximate Bayesian inference-we use active inference to simulate decisions of agents categorizing a scene in a hierarchically-structured setting. Under active inference, agents develop probabilistic beliefs about their environment, while actively sampling it to maximize the evidence for their internal generative model. This approximate evidence maximization (i.e., self-evidencing) comprises drives to both maximize rewards and resolve uncertainty about hidden states. This is realized via minimization of a free energy functional of posterior beliefs about both the world as well as the actions used to sample or perturb it, corresponding to perception and action, respectively. We show that active inference, in the context of hierarchical scene construction, gives rise to many empirical evidence accumulation phenomena, such as noise-sensitive reaction times and epistemic saccades. We explain these behaviors in terms of the principled drives that constitute the expected free energy, the key quantity for evaluating policies under active inference. In addition, we report novel behaviors exhibited by these active inference agents that furnish new predictions for research on evidence accumulation and perceptual decision-making. We discuss the implications of this hierarchical active inference scheme for tasks that require planned sequences of information-gathering actions to infer compositional latent structure (such as visual scene construction and sentence comprehension). This work sets the stage for future experiments to investigate active inference in relation to other formulations of evidence accumulation (e.g., drift-diffusion models) in tasks that require planning in uncertain environments with higher-order structure.

Closed loop motor-sensory dynamics in human vision

Vision is obtained with a continuous motion of the eyes. The kinematic analysis of eye motion, during any visual or ocular task, typically reveals two (kinematic) components: saccades, which quickly replace the visual content in the retinal fovea, and drifts, which slowly scan the image after each saccade. While the saccadic exchange of regions of interest (ROIs) is commonly considered to be included in motor-sensory closed-loops, it is commonly assumed that drifts function in an open-loop manner, that is, independent of the concurrent visual input. Accordingly, visual perception is assumed to be based on a sequence of open-loop processes, each initiated by a saccade-triggered retinal snapshot. Here we directly challenged this assumption by testing the dependency of drift kinematics on concurrent visual inputs using real-time gaze-contingent-display. Our results demonstrate a dependency of the trajectory on the concurrent visual input, convergence of speed to condition-specific values and maintenance of selected drift-related motor-sensory controlled variables, all strongly indicative of drifts being included in a closed-loop brain-world process, and thus suggesting that vision is inherently a closed-loop process.

Movie viewing elicits rich and reliable brain state dynamics

Adaptive brain function requires that sensory impressions of the social and natural milieu are dynamically incorporated into intrinsic brain activity. While dynamic switches between brain states have been well characterised in resting state acquisitions, the remodelling of these state transitions by engagement in naturalistic stimuli remains poorly understood. Here, we show that the temporal dynamics of brain states, as measured in fMRI, are reshaped from predominantly bistable transitions between two relatively indistinct states at rest, toward a sequence of well-defined functional states during movie viewing whose transitions are temporally aligned to specific features of the movie. The expression of these brain states covaries with different physiological states and reflects subjectively rated engagement in the movie. In sum, a data-driven decoding of brain states reveals the distinct reshaping of functional network expression and reliable state transitions that accompany the switch from resting state to perceptual immersion in an ecologically valid sensory experience.

A dual foveal-peripheral visual processing model implements efficient saccade selection

We develop a visuomotor model that implements visual search as a focal accuracy-seeking policy, with the target's position and category drawn independently from a common generative process. Consistently with the anatomical separation between the ventral versus dorsal pathways, the model is composed of two pathways that respectively infer what to see and where to look. The "What" network is a classical deep learning classifier that only processes a small region around the center of fixation, providing a "foveal" accuracy. In contrast, the "Where" network processes the full visual field in a biomimetic fashion, using a log-polar retinotopic encoding, which is preserved up to the action selection level. In our model, the foveal accuracy is used as a monitoring signal to train the "Where" network, much like in the "actor/critic" framework. After training, the "Where" network provides an "accuracy map" that serves to guide the eye toward peripheral objects. Finally, the comparison of both networks' accuracies amounts to either selecting a saccade or keeping the eye focused at the center to identify the target. We test this setup on a simple task of finding a digit in a large, cluttered image. Our simulation results demonstrate the effectiveness of this approach, increasing by one order of magnitude the radius of the visual field toward which the agent can detect and recognize a target, either through a single saccade or with multiple ones. Importantly, our log-polar treatment of the visual information exploits the strong compression rate performed at the sensory level, providing ways to implement visual search in a sublinear fashion, in contrast with mainstream computer vision.

Pupil Diameter Tracks Statistical Structure in the Environment to Increase Visual Sensitivity

Pupil diameter determines how much light hits the retina and, thus, how much information is available for visual processing. This is regulated by a brainstem reflex pathway. Here, we investigate whether this pathway is under the control of internal models about the environment. This would allow adjusting pupil dynamics to environmental statistics to augment information transmission. We present image sequences containing internal temporal structure to humans of either sex and male macaque monkeys. We then measure whether the pupil tracks this temporal structure not only at the rate of luminance variations, but also at the rate of statistics not available from luminance information alone. We find entrainment to environmental statistics in both species. This entrainment directly affects visual processing by increasing sensitivity at the environmentally relevant temporal frequency. Thus, pupil dynamics are matched to the temporal structure of the environment to optimize perception, in line with an active sensing account.SIGNIFICANCE STATEMENT When light hits the retina, the pupil reflexively constricts. This determines how much light and thus how much information is available for visual processing. We show that the rate at which the pupil constricts and dilates is matched to the temporal structure of our visual environment, although this information is not directly contained in the light variations that usually trigger reflexive pupil constrictions. Adjusting pupil diameter in accordance with environmental regularities optimizes information transmission at ecologically relevant temporal frequencies. We show that this is the case in humans and macaque monkeys, suggesting that the reflex pathways that regulate pupil diameter are under some degree of cognitive control across primate species.

The growth of cognition: Free energy minimization and the embryogenesis of cortical computation

The assumption that during cortical embryogenesis neurons and synaptic connections are selected to form an ensemble maximising synchronous oscillation explains mesoscopic cortical development, and a mechanism for cortical information processing is implied by consistency with the Free Energy Principle and Dynamic Logic. A heteroclinic network emerges, with stable and unstable fixed points of oscillation corresponding to activity in symmetrically connected, versus asymmetrically connected, sets of neurons. Simulations of growth explain a wide range of anatomical observations for columnar and non-columnar cortex, superficial patch connections, and the organization and dynamic interactions of neurone response properties. An antenatal scaffold is created, upon which postnatal learning can establish continuously ordered neuronal representations, permitting matching of co-synchronous fields in multiple cortical areas to solve optimization problems as in Dynamic Logic. Fast synaptic competition partitions equilibria, minimizing "the curse of dimensionality", while perturbations between imperfectly partitioned synchronous fields, under internal reinforcement, enable the cortex to become adaptively self-directed. As learning progresses variational free energy is minimized and entropy bounded.

Inferring What to Do (And What Not to)

In recent years, the "planning as inference" paradigm has become central to the study of behaviour. The advance offered by this is the formalisation of motivation as a prior belief about "how I am going to act". This paper provides an overview of the factors that contribute to this prior. These are rooted in optimal experimental design, information theory, and statistical decision making. We unpack how these factors imply a functional architecture for motivated behaviour. This raises an important question: how can we put this architecture to work in the service of understanding observed neurobiological structure? To answer this question, we draw from established techniques in experimental studies of behaviour. Typically, these examine the influence of perturbations of the nervous system-which include pathological insults or optogenetic manipulations-to see their influence on behaviour. Here, we argue that the message passing that emerges from inferring what to do can be similarly perturbed. If a given perturbation elicits the same behaviours as a focal brain lesion, this provides a functional interpretation of empirical findings and an anatomical grounding for theoretical results. We highlight examples of this approach that influence different sorts of goal-directed behaviour, active learning, and decision making. Finally, we summarise their implications for the neuroanatomy of inferring what to do (and what not to).

Sentience and the Origins of Consciousness: From Cartesian Duality to Markovian Monism

This essay addresses Cartesian duality and how its implicit dialectic might be repaired using physics and information theory. Our agenda is to describe a key distinction in the physical sciences that may provide a foundation for the distinction between mind and matter, and between sentient and intentional systems. From this perspective, it becomes tenable to talk about the physics of sentience and 'forces' that underwrite our beliefs (in the sense of probability distributions represented by our internal states), which may ground our mental states and consciousness. We will refer to this view as Markovian monism, which entails two claims: (1) fundamentally, there is only one type of thing and only one type of irreducible property (hence monism). (2) All systems possessing a Markov blanket have properties that are relevant for understanding the mind and consciousness: if such systems have mental properties, then they have them partly by virtue of possessing a Markov blanket (hence Markovian). Markovian monism rests upon the information geometry of random dynamic systems. In brief, the information geometry induced in any system-whose internal states can be distinguished from external states-must acquire a dual aspect. This dual aspect concerns the (intrinsic) information geometry of the probabilistic evolution of internal states and a separate (extrinsic) information geometry of probabilistic beliefs about external states that are parameterised by internal states. We call these intrinsic (i.e., mechanical, or state-based) and extrinsic (i.e., Markovian, or belief-based) information geometries, respectively. Although these mathematical notions may sound complicated, they are fairly straightforward to handle, and may offer a means through which to frame the origins of consciousness.

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.

Face Selective Neural Activity: Comparisons Between Fixed and Free Viewing

Event Related Potentials (ERPs) are widely used to study category-selective EEG responses to visual stimuli, such as the face-selective N170 component. Typically, this is done by flashing stimuli at the point of static gaze fixation. While allowing for good experimental control, these paradigms ignore the dynamic role of eye-movements in natural vision. Fixation-related potentials (FRPs), obtained using simultaneous EEG and eye-tracking, overcome this limitation. Various studies have used FRPs to study processes such as lexical processing, target detection and attention allocation. The goal of this study was to carefully compare face-sensitive activity time-locked to an abrupt stimulus onset at fixation, with that time-locked to a self-generated fixation on a stimulus. Twelve participants participated in three experimental conditions: Free-viewing (FRPs), Cued-viewing (FRPs) and Control (ERPs). We used a multiple regression approach to disentangle overlapping activity components. Our results show that the N170 face-effect is evident for the first fixation on a stimulus, whether it follows a self-generated saccade or stimulus appearance at fixation point. The N170 face-effect has similar topography across viewing conditions, but there were major differences within each stimulus category. We ascribe these differences to an overlap of the fixation-related lambda response and the N170. We tested the plausibility of this account using dipole simulations. Finally, the N170 exhibits category-specific adaptation in free viewing. This study establishes the comparability of the free-viewing N170 face-effect with the classic event-related effect, while highlighting the importance of accounting for eye-movement related effects.

A World Unto Itself: Human Communication as Active Inference

Recent theoretical work in developmental psychology suggests that humans are predisposed to align their mental states with those of other individuals. One way this manifests is in cooperative communication; that is, intentional communication aimed at aligning individuals' mental states with respect to events in their shared environment. This idea has received strong empirical support. The purpose of this paper is to extend this account by proposing an integrative model of the biobehavioral dynamics of cooperative communication. Our formulation is based on active inference. Active inference suggests that action-perception cycles operate to minimize uncertainty and optimize an individual's internal model of the world. We propose that humans are characterized by an evolved adaptive prior belief that their mental states are aligned with, or similar to, those of conspecifics (i.e., that 'we are the same sort of creature, inhabiting the same sort of niche'). The use of cooperative communication emerges as the principal means to gather evidence for this belief, allowing for the development of a shared narrative that is used to disambiguate interactants' (hidden and inferred) mental states. Thus, by using cooperative communication, individuals effectively attune to a hermeneutic niche composed, in part, of others' mental states; and, reciprocally, attune the niche to their own ends via epistemic niche construction. This means that niche construction enables features of the niche to encode precise, reliable cues about the deontic or shared value of certain action policies (e.g., the utility of using communicative constructions to disambiguate mental states, given expectations about shared prior beliefs). In turn, the alignment of mental states (prior beliefs) enables the emergence of a novel, contextualizing scale of cultural dynamics that encompasses the actions and mental states of the ensemble of interactants and their shared environment. The dynamics of this contextualizing layer of cultural organization feedback, across scales, to constrain the variability of the prior expectations of the individuals who constitute it. Our theory additionally builds upon the active inference literature by introducing a new set of neurobiologically plausible computational hypotheses for cooperative communication. We conclude with directions for future research.

Generalised free energy and active inference

Active inference is an approach to understanding behaviour that rests upon the idea that the brain uses an internal generative model to predict incoming sensory data. The fit between this model and data may be improved in two ways. The brain could optimise probabilistic beliefs about the variables in the generative model (i.e. perceptual inference). Alternatively, by acting on the world, it could change the sensory data, such that they are more consistent with the model. This implies a common objective function (variational free energy) for action and perception that scores the fit between an internal model and the world. We compare two free energy functionals for active inference in the framework of Markov decision processes. One of these is a functional of beliefs (i.e. probability distributions) about states and policies, but a function of observations, while the second is a functional of beliefs about all three. In the former (expected free energy), prior beliefs about outcomes are not part of the generative model (because they are absorbed into the prior over policies). Conversely, in the second (generalised free energy), priors over outcomes become an explicit component of the generative model. When using the free energy function, which is blind to future observations, we equip the generative model with a prior over policies that ensure preferred (i.e. priors over) outcomes are realised. In other words, if we expect to encounter a particular kind of outcome, this lends plausibility to those policies for which this outcome is a consequence. In addition, this formulation ensures that selected policies minimise uncertainty about future outcomes by minimising the free energy expected in the future. When using the free energy functional-that effectively treats future observations as hidden states-we show that policies are inferred or selected that realise prior preferences by minimising the free energy of future expectations. Interestingly, the form of posterior beliefs about policies (and associated belief updating) turns out to be identical under both formulations, but the quantities used to compute them are not.

Information generation as a functional basis of consciousness

What is the biological advantage of having consciousness? Functions of consciousness have been elusive due to the subjective nature of consciousness and ample empirical evidence showing the presence of many nonconscious cognitive performances in the human brain. Drawing upon empirical literature, here, we propose that a core function of consciousness be the ability to internally generate representations of events possibly detached from the current sensory input. Such representations are constructed by generative models learned through sensory-motor interactions with the environment. We argue that the ability to generate information underlies a variety of cognitive functions associated with consciousness such as intention, imagination, planning, short-term memory, attention, curiosity, and creativity, all of which contribute to non-reflexive behavior. According to this view, consciousness emerged in evolution when organisms gained the ability to perform internal simulations using internal models, which endowed them with flexible intelligent behavior. To illustrate the notion of information generation, we take variational autoencoders (VAEs) as an analogy and show that information generation corresponds the decoding (or decompression) part of VAEs. In biological brains, we propose that information generation corresponds to top-down predictions in the predictive coding framework. This is compatible with empirical observations that recurrent feedback activations are linked with consciousness whereas feedforward processing alone seems to occur without evoking conscious experience. Taken together, the information generation hypothesis captures many aspects of existing ideas about potential functions of consciousness and provides new perspectives on the functional roles of consciousness.

Bayesian Filtering with Multiple Internal Models: Toward a Theory of Social Intelligence

To exhibit social intelligence, animals have to recognize whom they are communicating with. One way to make this inference is to select among internal generative models of each conspecific who may be encountered. However, these models also have to be learned via some form of Bayesian belief updating. This induces an interesting problem: When receiving sensory input generated by a particular conspecific, how does an animal know which internal model to update? We consider a theoretical and neurobiologically plausible solution that enables inference and learning of the processes that generate sensory inputs (e.g., listening and understanding) and reproduction of those inputs (e.g., talking or singing), under multiple generative models. This is based on recent advances in theoretical neurobiology-namely, active inference and post hoc (online) Bayesian model selection. In brief, this scheme fits sensory inputs under each generative model. Model parameters are then updated in proportion to the probability that each model could have generated the input (i.e., model evidence). The proposed scheme is demonstrated using a series of (real zebra finch) birdsongs, where each song is generated by several different birds. The scheme is implemented using physiologically plausible models of birdsong production. We show that generalized Bayesian filtering, combined with model selection, leads to successful learning across generative models, each possessing different parameters. These results highlight the utility of having multiple internal models when making inferences in social environments with multiple sources of sensory information.

The Predictive Processing Model of EMDR

Eye Movement Desensitization and Reprocessing Therapy (EMDR) is an effective treatment for Post-traumatic Stress Disorder (PTSD). The Adaptive Information Processing Model (AIP) guides the development and practice of EMDR. The AIP postulates inadequately processed memory as the foundation of PTSD pathology. Predictive Processing postulates that the primary function of the brain is prediction that serves to anticipate the next moment of experience in order to resist the dissipative force of entropy thus facilitating continued survival. Memory is the primary substrate of prediction, and is optimized by an ongoing process of precision weighted prediction error minimization that refines prediction by updating the memories on which it is based. The Predictive Processing model of EMDR postulates that EMDR facilitates the predictive processing of traumatic memory by overcoming the bias against exploration and evidence accumulation. The EMDR protocol brings the traumatic memory into an active state of re-experiencing. Defensive responding and/or low sensory precision preclude evidence accumulation to test the predictions of the traumatic memory in the present. Sets of therapist guided eye movements repeatedly challenge the bias against evidence accumulation and compel sensory sampling of the benign present. Eye movements reset the theta rhythm organizing the flow of information through the brain, facilitating the deployment of both overt and covert attention, and the mnemonic search for associations. Sampling of sensation does not support the predictions of the traumatic memory resulting in prediction error that the brain then attempts to minimize. The net result is a restoration of the integrity of the rhythmic deployment of attention, a recalibration of sensory precision, and the updating (reconsolidation) of the traumatic memory. Thus one prediction of the model is a decrease in Attention Bias Variability, a core dysfunction in PTSD, following successful treatment with EMDR.

Distinguishing volumetric content from perceptual presence within a predictive processing framework

I argue for an overlooked distinction between perceptual presence and volumetric content, and flesh it out in terms of predictive processing. Within the predictive processing framework we can distinguish between agent-active and object-active expectations. The former expectations account for perceptual presence, while the latter account for volumetric content. I then support this position with reference to how experiences of presence are created by virtual reality technologies, and end by reflecting on what this means for the relationship between sensorimotor enactivism and predictive processing.

The peripheral preview effect with faces: Combined EEG and eye-tracking suggests multiple stages of trans-saccadic predictive and non-predictive processing

The world appears stable despite saccadic eye-movements. One possible explanation for this phenomenon is that the visual system predicts upcoming input across saccadic eye-movements based on peripheral preview of the saccadic target. We tested this idea using concurrent electroencephalography (EEG) and eye-tracking. Participants made cued saccades to peripheral upright or inverted face stimuli that changed orientation (invalid preview) or maintained orientation (valid preview) while the saccade was completed. Experiment 1 demonstrated better discrimination performance and a reduced fixation-locked N170 component (fN170) with valid than with invalid preview, demonstrating integration of pre- and post-saccadic information. Moreover, the early fixation-related potentials (FRP) showed a preview face inversion effect suggesting that some pre-saccadic input was represented in the brain until around 170 ms post fixation-onset. Experiment 2 replicated Experiment 1 and manipulated the proportion of valid and invalid trials to test whether the preview effect reflects context-based prediction across trials. A whole-scalp Bayes factor analysis showed that this manipulation did not alter the fN170 preview effect but did influence the face inversion effect before the saccade. The pre-saccadic inversion effect declined earlier in the mostly invalid block than in the mostly valid block, which is consistent with the notion of pre-saccadic expectations. In addition, in both studies, we found strong evidence for an interaction between the pre-saccadic preview stimulus and the post-saccadic target as early as 50 ms (Experiment 2) or 90 ms (Experiment 1) into the new fixation. These findings suggest that visual stability may involve three temporal stages: prediction about the saccadic target, integration of pre-saccadic and post-saccadic information at around 50-90 ms post fixation onset, and post-saccadic facilitation of rapid categorization.

Morphogenesis as Bayesian inference: A variational approach to pattern formation and control in complex biological systems

Recent advances in molecular biology such as gene editing [1], bioelectric recording and manipulation [2] and live cell microscopy using fluorescent reporters [3], [4] - especially with the advent of light-controlled protein activation through optogenetics [5] - have provided the tools to measure and manipulate molecular signaling pathways with unprecedented spatiotemporal precision. This has produced ever increasing detail about the molecular mechanisms underlying development and regeneration in biological organisms. However, an overarching concept - that can predict the emergence of form and the robust maintenance of complex anatomy - is largely missing in the field. Classic (i.e., dynamic systems and analytical mechanics) approaches such as least action principles are difficult to use when characterizing open, far-from equilibrium systems that predominate in Biology. Similar issues arise in neuroscience when trying to understand neuronal dynamics from first principles. In this (neurobiology) setting, a variational free energy principle has emerged based upon a formulation of self-organization in terms of (active) Bayesian inference. The free energy principle has recently been applied to biological self-organization beyond the neurosciences [6], [7]. For biological processes that underwrite development or regeneration, the Bayesian inference framework treats cells as information processing agents, where the driving force behind morphogenesis is the maximization of a cell's model evidence. This is realized by the appropriate expression of receptors and other signals that correspond to the cell's internal (i.e., generative) model of what type of receptors and other signals it should express. The emerging field of the free energy principle in pattern formation provides an essential quantitative formalism for understanding cellular decision-making in the context of embryogenesis, regeneration, and cancer suppression. In this paper, we derive the mathematics behind Bayesian inference - as understood in this framework - and use simulations to show that the formalism can reproduce experimental, top-down manipulations of complex morphogenesis. First, we illustrate this 'first principle' approach to morphogenesis through simulated alterations of anterior-posterior axial polarity (i.e., the induction of two heads or two tails) as in planarian regeneration. Then, we consider aberrant signaling and functional behavior of a single cell within a cellular ensemble - as a first step in carcinogenesis as false 'beliefs' about what a cell should 'sense' and 'do'. We further show that simple modifications of the inference process can cause - and rescue - mis-patterning of developmental and regenerative events without changing the implicit generative model of a cell as specified, for example, by its DNA. This formalism offers a new road map for understanding developmental change in evolution and for designing new interventions in regenerative medicine settings.

Making the Environment an Informative Place: A Conceptual Analysis of Epistemic Policies and Sensorimotor Coordination

How do living organisms decide and act with limited and uncertain information? Here, we discuss two computational approaches to solving these challenging problems: a "cognitive" and a "sensorimotor" enrichment of stimuli, respectively. In both approaches, the key notion is that agents can strategically modulate their behavior in informative ways, e.g., to disambiguate amongst alternative hypotheses or to favor the perception of stimuli providing the information necessary to later act appropriately. We discuss how, despite their differences, both approaches appeal to the notion that actions must obey both epistemic (i.e., information-gathering or uncertainty-reducing) and pragmatic (i.e., goal- or reward-maximizing) imperatives and balance them. Our computationally-guided analysis reveals that epistemic behavior is fundamental to understanding several facets of cognitive processing, including perception, decision making, and social interaction.

A Computational Model of Visual Recognition Memory via Grid Cells

Models of face, object, and scene recognition traditionally focus on massively parallel processing of low-level features, with higher-order representations emerging at later processing stages. However, visual perception is tightly coupled to eye movements, which are necessarily sequential. Recently, neurons in entorhinal cortex have been reported with grid cell-like firing in response to eye movements, i.e., in visual space. Following the presumed role of grid cells in vector navigation, we propose a model of recognition memory for familiar faces, objects, and scenes, in which grid cells encode translation vectors between salient stimulus features. A sequence of saccadic eye-movement vectors, moving from one salient feature to the expected location of the next, potentially confirms an initial hypothesis (accumulating evidence toward a threshold) about stimulus identity, based on the relative feature layout (i.e., going beyond recognition of individual features). The model provides an explicit neural mechanism for the long-held view that directed saccades support hypothesis-driven, constructive perception and recognition; is compatible with holistic face processing; and constitutes the first quantitative proposal for a role of grid cells in visual recognition. The variance of grid cell activity along saccade trajectories exhibits 6-fold symmetry across 360 degrees akin to recently reported fMRI data. The model suggests that disconnecting grid cells from occipitotemporal inputs may yield prosopagnosia-like symptoms. The mechanism is robust with regard to partial visual occlusion, can accommodate size and position invariance, and suggests a functional explanation for medial temporal lobe involvement in visual memory for relational information and memory-guided attention.

Developing attentional control in naturalistic dynamic road crossing situations

In the last 20 years, there has been increasing interest in studying visual attentional processes under more natural conditions. In the present study, we propose to determine the critical age at which children show similar to adult performance and attentional control in a visually guided task; in a naturalistic dynamic and socially relevant context: road crossing. We monitored visual exploration and crossing decisions in adults and children aged between 5 and 15 while they watched road traffic videos containing a range of traffic densities with or without pedestrians. 5–10 year old (y/o) children showed less systematic gaze patterns. More specifically, adults and 11–15 y/o children look mainly at the vehicles’ appearing point, which is an optimal location to sample diagnostic information for the task. In contrast, 5–10 y/os look more at socially relevant stimuli and attend to moving vehicles further down the trajectory when the traffic density is high. Critically, 5-10 y/o children also make an increased number of crossing decisions compared to 11–15 y/os and adults. Our findings reveal a critical shift around 10 y/o in attentional control and crossing decisions in a road crossing task.

Time course of spatiotopic updating across saccades

Humans move their eyes several times per second, yet we perceive the outside world as continuous despite the sudden disruptions created by each eye movement. To date, the mechanism that the brain employs to achieve visual continuity across eye movements remains unclear. While it has been proposed that the oculomotor system quickly updates and informs the visual system about the upcoming eye movement, behavioral studies investigating the time course of this updating suggest the involvement of a slow mechanism, estimated to take more than 500 ms to operate effectively. This is a surprisingly slow estimate, because both the visual system and the oculomotor system process information faster. If spatiotopic updating is indeed this slow, it cannot contribute to perceptual continuity, because it is outside the temporal regime of typical oculomotor behavior. Here, we argue that the behavioral paradigms that have been used previously are suboptimal to measure the speed of spatiotopic updating. In this study, we used a fast gaze-contingent paradigm, using high phi as a continuous stimulus across eye movements. We observed fast spatiotopic updating within 150 ms after stimulus onset. The results suggest the involvement of a fast updating mechanism that predictively influences visual perception after an eye movement. The temporal characteristics of this mechanism are compatible with the rate at which saccadic eye movements are typically observed in natural viewing.

Reconsidering the Mind-Wandering Reader: Predictive Processing, Probability Designs, and Enculturation

Studies on mind-wandering frequently use reading as an experimental task. In these studies, reading is conceived as a cognitive process that potentially offers a contrast to mind-wandering, because it seems to be task-related, goal-directed and stimulus-dependent. More recent work attempts to avoid the dichotomy of successful cognitive processes and processes of mind-wandering found in earlier studies. We approach the issue from the perspective that texts provoke modes of cognitive involvement different from the information processing and recall account that underlies many early mind-wandering studies. After all, reading itself is an umbrella term for activities that are characterized by a variety of phenomenological and functional signatures. We conceptualize reading and mind-wandering in relation to each other through the framework of enculturated predictive processing, which is informed by research in literary studies. Earlier accounts think of reading predominantly in terms of the construction of situation models that organize textual information. By contrast, enculturated predictive processing foregrounds shifting stances readers can take in relation to the text. Characters featuring in literary texts might be mind-wandering themselves, or passages in literary style might make the construction of a clearly defined situational model impossible. Furthermore, we take into account that texts often elicit mind-wandering through the construction of task-relevant and attention-driven virtual scenarios in readers' minds. This more complex account of reading can enrich recent attempts to understand mind-wandering as a complex, multi-dimensional phenomenon. The study of mind-wandering can benefit, we argue, from a closer attention to the process of reading and to the texts it deploys as stimuli. The emerging perspective from enculturated predictive processing and literary studies makes distinctions in reading that in turn enable research on mind-wandering to ask more precise questions about (1) different kinds of mind-wandering, (2) different modes of reading, and (3) how and where they interconnect across time.

Perceptual Content, Not Physiological Signals, Determines Perceived Duration When Viewing Dynamic, Natural Scenes

The neural basis of time perception remains unknown. A prominent account is the pacemaker-accumulator model, wherein regular ticks of some physiological or neural pacemaker are read out as time. Putative candidates for the pacemaker have been suggested in physiological processes (heartbeat), or dopaminergic mid-brain neurons, whose activity has been associated with spontaneous blinking. However, such proposals have difficulty accounting for observations that time perception varies systematically with perceptual content. We examined physiological influences on human duration estimates for naturalistic videos between 1–64 seconds using cardiac and eye recordings. Duration estimates were biased by the amount of change in scene content. Contrary to previous claims, heart rate, and blinking were not related to duration estimates. Our results support a recent proposal that tracking change in perceptual classification networks provides a basis for human time perception, and suggest that previous assertions of the importance of physiological factors should be tempered.

Connectomics of bipolar disorder: a critical review, and evidence for dynamic instabilities within interoceptive networks

The notion that specific cognitive and emotional processes arise from functionally distinct brain regions has lately shifted toward a connectivity-based approach that emphasizes the role of network-mediated integration across regions. The clinical neurosciences have likewise shifted from a predominantly lesion-based approach to a connectomic paradigm-framing disorders as diverse as stroke, schizophrenia (SCZ), and dementia as "dysconnection syndromes". Here we position bipolar disorder (BD) within this paradigm. We first summarise the disruptions in structural, functional and effective connectivity that have been documented in BD. Not surprisingly, these disturbances show a preferential impact on circuits that support emotional processes, cognitive control and executive functions. Those at high risk (HR) for BD also show patterns of connectivity that differ from both matched control populations and those with BD, and which may thus speak to neurobiological markers of both risk and resilience. We highlight research fields that aim to link brain network disturbances to the phenotype of BD, including the study of large-scale brain dynamics, the principles of network stability and control, and the study of interoception (the perception of physiological states). Together, these findings suggest that the affective dysregulation of BD arises from dynamic instabilities in interoceptive circuits which subsequently impact on fear circuitry and cognitive control systems. We describe the resulting disturbance as a "psychosis of interoception".

Embodied Precision: Intranasal Oxytocin Modulates Multisensory Integration

Multisensory integration processes are fundamental to our sense of self as embodied beings. Bodily illusions, such as the rubber hand illusion (RHI) and the size-weight illusion (SWI), allow us to investigate how the brain resolves conflicting multisensory evidence during perceptual inference in relation to different facets of body representation. In the RHI, synchronous tactile stimulation of a participant's hidden hand and a visible rubber hand creates illusory body ownership; in the SWI, the perceived size of the body can modulate the estimated weight of external objects. According to Bayesian models, such illusions arise as an attempt to explain the causes of multisensory perception and may reflect the attenuation of somatosensory precision, which is required to resolve perceptual hypotheses about conflicting multisensory input. Recent hypotheses propose that the precision of sensorimotor representations is determined by modulators of synaptic gain, like dopamine, acetylcholine, and oxytocin. However, these neuromodulatory hypotheses have not been tested in the context of embodied multisensory integration. The present, double-blind, placebo-controlled, crossover study ( n = 41 healthy volunteers) aimed to investigate the effect of intranasal oxytocin (IN-OT) on multisensory integration processes, tested by means of the RHI and the SWI. Results showed that IN-OT enhanced the subjective feeling of ownership in the RHI, only when synchronous tactile stimulation was involved. Furthermore, IN-OT increased an embodied version of the SWI (quantified as estimation error during a weight estimation task). These findings suggest that oxytocin might modulate processes of visuotactile multisensory integration by increasing the precision of top-down signals against bottom-up sensory input.

Active Fovea-Based Vision Through Computationally-Effective Model-Based Prediction

What motivates an action in the absence of a definite reward? Taking the case of visuomotor control, we consider a minimal control problem that is how select the next saccade, in a sequence of discrete eye movements, when the final objective is to better interpret the current visual scene. The visual scene is modeled here as a partially-observed environment, with a generative model explaining how the visual data is shaped by action. This allows to interpret different action selection metrics proposed in the literature, including the Salience, the Infomax and the Variational Free Energy, under a single information theoretic construct, namely the view-based Information Gain. Pursuing this analytic track, two original action selection metrics named the Information Gain Lower Bound (IGLB) and the Information Gain Upper Bound (IGUB) are then proposed. Showing either a conservative or an optimistic bias regarding the Information Gain, they strongly simplify its calculation. An original fovea-based visual scene decoding setup is then proposed, with numerical experiments highlighting different facets of artificial fovea-based vision. A first and principal result is that state-of-the-art recognition rates are obtained with fovea-based saccadic exploration, using less than 10% of the original image's data. Those satisfactory results illustrate the advantage of mixing predictive control with accurate state-of-the-art predictors, namely a deep neural network. A second result is the sub-optimality of some classical action-selection metrics widely used in the literature, that is not manifest with finely-tuned inference models, but becomes patent when coarse or faulty models are used. Last, a computationally-effective predictive model is developed using the IGLB objective, with pre-processed visual scan-path read-out from memory, bypassing computationally-demanding predictive calculations. This last simplified setting is shown effective in our case, showing both a competing accuracy and a good robustness to model flaws.

Dopamine and eye movement control in Parkinson's disease: deficits in corollary discharge signals?

Movement in Parkinson's disease (PD) is fragmented, and the patients depend on visual information in their behavior. This suggests that the patients may have deficits in internally monitoring their own movements. Internal monitoring of movements is assumed to rely on corollary discharge signals that enable the brain to predict the sensory consequences of actions. We studied early-stage PD patients (N = 14), and age-matched healthy control participants (N = 14) to examine whether PD patients reveal deficits in updating their sensory representations after eye movements. The participants performed a double-saccade task where, in order to accurately fixate a second target, the participant must correct for the displacement caused by the first saccade. In line with previous reports, the patients had difficulties in fixating the second target when the eye movement was performed without visual guidance. Furthermore, the patients had difficulties in taking into account the error in the first saccade when making a saccade toward the second target, especially when eye movements were made toward the side with dominant motor symptoms. Across PD patients, the impairments in saccadic eye movements correlated with the integrity of the dopaminergic system as measured with [123I]FP-CIT SPECT: Patients with lower striatal (caudate, anterior putamen, and posterior putamen) dopamine transporter binding made larger errors in saccades. This effect was strongest when patients made memory-guided saccades toward the second target. Our results provide tentative evidence that the motor deficits in PD may be partly due to deficits in internal monitoring of movements.

The Anatomy of Inference: Generative Models and Brain Structure

To infer the causes of its sensations, the brain must call on a generative (predictive) model. This necessitates passing local messages between populations of neurons to update beliefs about hidden variables in the world beyond its sensory samples. It also entails inferences about how we will act. Active inference is a principled framework that frames perception and action as approximate Bayesian inference. This has been successful in accounting for a wide range of physiological and behavioral phenomena. Recently, a process theory has emerged that attempts to relate inferences to their neurobiological substrates. In this paper, we review and develop the anatomical aspects of this process theory. We argue that the form of the generative models required for inference constrains the way in which brain regions connect to one another. Specifically, neuronal populations representing beliefs about a variable must receive input from populations representing the Markov blanket of that variable. We illustrate this idea in four different domains: perception, planning, attention, and movement. In doing so, we attempt to show how appealing to generative models enables us to account for anatomical brain architectures. Ultimately, committing to an anatomical theory of inference ensures we can form empirical hypotheses that can be tested using neuroimaging, neuropsychological, and electrophysiological experiments.

A Comparison of Predictive Spatial Augmented Reality Cues for Procedural Tasks

Previous research has demonstrated that Augmented Reality can reduce a user's task response time and mental effort when completing a procedural task. This paper investigates techniques to improve user performance and reduce mental effort by providing projector-based Spatial Augmented Reality predictive cues for future responses. The objective of the two experiments conducted in this study was to isolate the performance and mental effort differences from several different annotation cueing techniques for simple (Experiment 1) and complex (Experiment 2) button-pressing tasks. Comporting with existing cognitive neuroscience literature on prediction, attentional orienting, and interference, we hypothesized that for both simple procedural tasks and complex search-based tasks, having a visual cue guiding to the next task's location would positively impact performance relative to a baseline, no-cue condition. Additionally, we predicted that direction-based cues would provide a more significant positive impact than target-based cues. The results indicated that providing a line to the next task was the most effective technique for improving the users' task time and mental effort in both the simple and complex tasks.

Rapid Eye Movements in Sleep Furnish a Unique Probe Into Consciousness

The neural correlates of rapid eye movements (REMs) in sleep are extraordinarily robust; including REM-locked multisensory-motor integration and accompanying activation in the retrosplenial cortex, the supplementary eye field and areas encompassing cholinergic basal nucleus (Hong et al., 2009). The phenomenology of REMs speaks to the notion that perceptual experience in both sleep and wakefulness is a constructive process - in which we generate predictions of sensory inputs and then test those predictions through actively sampling the sensorium with eye movements. On this view, REMs during sleep may index an internalized active sampling or 'scanning' of self-generated visual constructs that are released from the constraints of visual input. If this view is correct, it renders REMs an ideal probe to study consciousness as "an exclusively internal affair" (Metzinger, 2009). In other words, REMs offer a probe of active inference - in the sense of predictive coding - when the brain is isolated from the sensorium in virtue of the natural blockade of sensory afferents during REM sleep. Crucially, REMs are temporally precise events that enable powerful inferences based on time series analyses. As a natural, task-free probe, (REMs) could be used in non-compliant subjects, including infants and animals. In short, REMs constitute a promising probe to study the ontogenetic and phylogenetic development of consciousness and perhaps the psychopathology of schizophrenia and autism, which have been considered in terms of aberrant predictive coding.

Attention or salience?

While attention is widely recognised as central to perception, the term is often used to mean very different things. Prominent theories of attention - notably the premotor theory - relate it to planned or executed eye movements. This contrasts with the notion of attention as a gain control process that weights the information carried by different sensory channels. We draw upon recent advances in theoretical neurobiology to argue for a distinction between attentional gain mechanisms and salience attribution. The former depends upon estimating the precision of sensory data, while the latter is a consequence of the need to actively engage with the sensorium. Having established this distinction, we consider the intimate relationship between attention and salience.

Free-energy minimization in joint agent-environment systems: A niche construction perspective

The free-energy principle is an attempt to explain the structure of the agent and its brain, starting from the fact that an agent exists (Friston and Stephan, 2007; Friston et al., 2010). More specifically, it can be regarded as a systematic attempt to understand the 'fit' between an embodied agent and its niche, where the quantity of free-energy is a measure for the 'misfit' or disattunement (Bruineberg and Rietveld, 2014) between agent and environment. This paper offers a proof-of-principle simulation of niche construction under the free-energy principle. Agent-centered treatments have so far failed to address situations where environments change alongside agents, often due to the action of agents themselves. The key point of this paper is that the minimum of free-energy is not at a point in which the agent is maximally adapted to the statistics of a static environment, but can better be conceptualized an attracting manifold within the joint agent-environment state-space as a whole, which the system tends toward through mutual interaction. We will provide a general introduction to active inference and the free-energy principle. Using Markov Decision Processes (MDPs), we then describe a canonical generative model and the ensuing update equations that minimize free-energy. We then apply these equations to simulations of foraging in an environment; in which an agent learns the most efficient path to a pre-specified location. In some of those simulations, unbeknownst to the agent, the 'desire paths' emerge as a function of the activity of the agent (i.e. niche construction occurs). We will show how, depending on the relative inertia of the environment and agent, the joint agent-environment system moves to different attracting sets of jointly minimized free-energy.

Being a Beast Machine: The Somatic Basis of Selfhood

Modern psychology has long focused on the body as the basis of the self. Recently, predictive processing accounts of interoception (perception of the body 'from within') have become influential in accounting for experiences of body ownership and emotion. Here, we describe embodied selfhood in terms of 'instrumental interoceptive inference' that emphasises allostatic regulation and physiological integrity. We apply this approach to the distinctive phenomenology of embodied selfhood, accounting for its non-object-like character and subjective stability over time. Our perspective has implications for the development of selfhood and illuminates longstanding debates about relations between life and mind, implying, contrary to Descartes, that experiences of embodied selfhood arise because of, and not in spite of, our nature as 'beast machines'.

The Discrete and Continuous Brain: From Decisions to Movement-And Back Again

To act upon the world, creatures must change continuous variables such as muscle length or chemical concentration. In contrast, decision making is an inherently discrete process, involving the selection among alternative courses of action. In this article, we consider the interface between the discrete and continuous processes that translate our decisions into movement in a Newtonian world-and how movement informs our decisions. We do so by appealing to active inference, with a special focus on the oculomotor system. Within this exemplar system, we argue that the superior colliculus is well placed to act as a discrete-continuous interface. Interestingly, when the neuronal computations within the superior colliculus are formulated in terms of active inference, we find that many aspects of its neuroanatomy emerge from the computations it must perform in this role.

Uncertainty, epistemics and active inference

Biological systems—like ourselves—are constantly faced with uncertainty. Despite noisy sensory data, and volatile environments, creatures appear to actively maintain their integrity. To account for this remarkable ability to make optimal decisions in the face of a capricious world, we propose a generative model that represents the beliefs an agent might possess about their own uncertainty. By simulating a noisy and volatile environment, we demonstrate how uncertainty influences optimal epistemic (visual) foraging. In our simulations, saccades were deployed less frequently to regions with a lower sensory precision, while a greater volatility led to a shorter inhibition of return. These simulations illustrate a principled explanation for some cardinal aspects of visual foraging—and allow us to propose a correspondence between the representation of uncertainty and ascending neuromodulatory systems, complementing that suggested by Yu & Dayan (Yu & Dayan 2005 Neuron46, 681–692. (doi:10.1016/j.neuron.2005.04.026)).

Recognition Dynamics in the Brain under the Free Energy Principle

We formulate the computational processes of perception in the framework of the principle of least action by postulating the theoretical action as a time integral of the variational free energy in the neurosciences. The free energy principle is accordingly rephrased, on autopoetic grounds, as follows: all viable organisms attempt to minimize their sensory uncertainty about an unpredictable environment over a temporal horizon. By taking the variation of informational action, we derive neural recognition dynamics (RD), which by construction reduces to the Bayesian filtering of external states from noisy sensory inputs. Consequently, we effectively cast the gradient-descent scheme of minimizing the free energy into Hamiltonian mechanics by addressing only the positions and momenta of the organisms' representations of the causal environment. To demonstrate the utility of our theory, we show how the RD may be implemented in a neuronally based biophysical model at a single-cell level and subsequently in a coarse-grained, hierarchical architecture of the brain. We also present numerical solutions to the RD for a model brain and analyze the perceptual trajectories around attractors in neural state space.

The Discrete and Continuous Brain: From Decisions to Movement-And Back Again

To act upon the world, creatures must change continuous variables such as muscle length or chemical concentration. In contrast, decision making is an inherently discrete process, involving the selection among alternative courses of action. In this article, we consider the interface between the discrete and continuous processes that translate our decisions into movement in a Newtonian world—and how movement informs our decisions. We do so by appealing to active inference, with a special focus on the oculomotor system. Within this exemplar system, we argue that the superior colliculus is well placed to act as a discrete-continuous interface. Interestingly, when the neuronal computations within the superior colliculus are formulated in terms of active inference, we find that many aspects of its neuroanatomy emerge from the computations it must perform in this role.

Dynamic Changes in Cortical Effective Connectivity Underlie Transsaccadic Integration in Humans

Due to saccadic eye movements the retinal image is abruptly displaced 2-4 times a second, yet we experience a stable and continuous stream of vision. It is known that saccades modulate neural processing in various local brain areas, but the question of how saccades influence neural communication between different areas in the thalamo-cortical system has remained unanswered. By combining transcranial magnetic stimulation with electroencephalography, we found that saccades were accompanied by dynamic changes in causal communication between different brain areas in humans. These changes were anticipatory; they began before the actual eye movement. Compared with fixation, communication between posterior cortical areas was first briefly enhanced during saccades, but subsequently peri-saccadic information did not ignite sustained activity in fronto-parietal cortices. This suggests that the brain constructs a spatially stable and temporally continuous stream of conscious vision from discrete fixations by restricting the access of peri-saccadic visual information to sustained processing in fronto-parietal cortices.

Active inference and the anatomy of oculomotion

Given that eye movement control can be framed as an inferential process, how are the requisite forces generated to produce anticipated or desired fixation? Starting from a generative model based on simple Newtonian equations of motion, we derive a variational solution to this problem and illustrate the plausibility of its implementation in the oculomotor brainstem. We show, through simulation, that the Bayesian filtering equations that implement 'planning as inference' can generate both saccadic and smooth pursuit eye movements. Crucially, the associated message passing maps well onto the known connectivity and neuroanatomy of the brainstem - and the changes in these messages over time are strikingly similar to single unit recordings of neurons in the corresponding nuclei. Furthermore, we show that simulated lesions to axonal pathways reproduce eye movement patterns of neurological patients with damage to these tracts.

Computational Neuropsychology and Bayesian Inference

Computational theories of brain function have become very influential in neuroscience. They have facilitated the growth of formal approaches to disease, particularly in psychiatric research. In this paper, we provide a narrative review of the body of computational research addressing neuropsychological syndromes, and focus on those that employ Bayesian frameworks. Bayesian approaches to understanding brain function formulate perception and action as inferential processes. These inferences combine ‘prior’ beliefs with a generative (predictive) model to explain the causes of sensations. Under this view, neuropsychological deficits can be thought of as false inferences that arise due to aberrant prior beliefs (that are poor fits to the real world). This draws upon the notion of a Bayes optimal pathology – optimal inference with suboptimal priors – and provides a means for computational phenotyping. In principle, any given neuropsychological disorder could be characterized by the set of prior beliefs that would make a patient’s behavior appear Bayes optimal. We start with an overview of some key theoretical constructs and use these to motivate a form of computational neuropsychology that relates anatomical structures in the brain to the computations they perform. Throughout, we draw upon computational accounts of neuropsychological syndromes. These are selected to emphasize the key features of a Bayesian approach, and the possible types of pathological prior that may be present. They range from visual neglect through hallucinations to autism. Through these illustrative examples, we review the use of Bayesian approaches to understand the link between biology and computation that is at the heart of neuropsychology.