The Bayesian brain: the role of uncertainty in neural coding and computation

Representing stimulus motion with waves in adaptive neural fields

Traveling waves of neural activity emerge in cortical networks both spontaneously and in response to stimuli. The spatiotemporal structure of waves can indicate the information they encode and the physiological processes that sustain them. Here, we investigate the stimulus-response relationships of traveling waves emerging in adaptive neural fields as a model of visual motion processing. Neural field equations model the activity of cortical tissue as a continuum excitable medium, and adaptive processes provide negative feedback, generating localized activity patterns. Synaptic connectivity in our model is described by an integral kernel that weakens dynamically due to activity-dependent synaptic depression, leading to marginally stable traveling fronts (with attenuated backs) or pulses of a fixed speed. Our analysis quantifies how weak stimuli shift the relative position of these waves over time, characterized by a wave response function we obtain perturbatively. Persistent and continuously visible stimuli model moving visual objects. Intermittent flashes that hop across visual space can produce the experience of smooth apparent visual motion. Entrainment of waves to both kinds of moving stimuli are well characterized by our theory and numerical simulations, providing a mechanistic description of the perception of visual motion.

Chaotic neural dynamics facilitate probabilistic computations through sampling

Cortical neurons exhibit highly variable responses over trials and time. Theoretical works posit that this variability arises potentially from chaotic network dynamics of recurrently connected neurons. Here, we demonstrate that chaotic neural dynamics, formed through synaptic learning, allow networks to perform sensory cue integration in a sampling-based implementation. We show that the emergent chaotic dynamics provide neural substrates for generating samples not only of a static variable but also of a dynamical trajectory, where generic recurrent networks acquire these abilities with a biologically plausible learning rule through trial and error. Furthermore, the networks generalize their experience in the stimulus-evoked samples to the inference without partial or all sensory information, which suggests a computational role of spontaneous activity as a representation of the priors as well as a tractable biological computation for marginal distributions. These findings suggest that chaotic neural dynamics may serve for the brain function as a Bayesian generative model.

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.

Prediction error processing and sharpening of expected information across the face-processing hierarchy

The perception and neural processing of sensory information are strongly influenced by prior expectations. The integration of prior and sensory information can manifest through distinct underlying mechanisms: focusing on unexpected input, denoted as prediction error (PE) processing, or amplifying anticipated information via sharpened representation. In this study, we employed computational modeling using deep neural networks combined with representational similarity analyses of fMRI data to investigate these two processes during face perception. Participants were cued to see face images, some generated by morphing two faces, leading to ambiguity in face identity. We show that expected faces were identified faster and perception of ambiguous faces was shifted towards priors. Multivariate analyses uncovered evidence for PE processing across and beyond the face-processing hierarchy from the occipital face area (OFA), via the fusiform face area, to the anterior temporal lobe, and suggest sharpened representations in the OFA. Our findings support the proposition that the brain represents faces grounded in prior expectations.

The effect of task load, information reliability and interdependency on anticipation performance

In sport, coaches often explicitly provide athletes with stable contextual information related to opponent action preferences to enhance anticipation performance. This information can be dependent on, or independent of, dynamic contextual information that only emerges during the sequence of play (e.g. opponent positioning). The interdependency between contextual information sources, and the associated cognitive demands of integrating information sources during anticipation, has not yet been systematically examined. We used a temporal occlusion paradigm to alter the reliability of contextual and kinematic information during the early, mid- and final phases of a two-versus-two soccer anticipation task. A dual-task paradigm was incorporated to investigate the impact of task load on skilled soccer players' ability to integrate information and update their judgements in each phase. Across conditions, participants received no contextual information (control) or stable contextual information (opponent preferences) that was dependent on, or independent of, dynamic contextual information (opponent positioning). As predicted, participants used reliable contextual and kinematic information to enhance anticipation. Further exploratory analysis suggested that increased task load detrimentally affected anticipation accuracy but only when both reliable contextual and kinematic information were available for integration in the final phase. This effect was observed irrespective of whether the stable contextual information was dependent on, or independent of, dynamic contextual information. Findings suggest that updating anticipatory judgements in the final phase of a sequence of play based on the integration of reliable contextual and kinematic information requires cognitive resources.

Predictive coding with spiking neurons and feedforward gist signaling

Predictive coding (PC) is an influential theory in neuroscience, which suggests the existence of a cortical architecture that is constantly generating and updating predictive representations of sensory inputs. Owing to its hierarchical and generative nature, PC has inspired many computational models of perception in the literature. However, the biological plausibility of existing models has not been sufficiently explored due to their use of artificial neurons that approximate neural activity with firing rates in the continuous time domain and propagate signals synchronously. Therefore, we developed a spiking neural network for predictive coding (SNN-PC), in which neurons communicate using event-driven and asynchronous spikes. Adopting the hierarchical structure and Hebbian learning algorithms from previous PC neural network models, SNN-PC introduces two novel features: (1) a fast feedforward sweep from the input to higher areas, which generates a spatially reduced and abstract representation of input (i.e., a neural code for the gist of a scene) and provides a neurobiological alternative to an arbitrary choice of priors; and (2) a separation of positive and negative error-computing neurons, which counters the biological implausibility of a bi-directional error neuron with a very high baseline firing rate. After training with the MNIST handwritten digit dataset, SNN-PC developed hierarchical internal representations and was able to reconstruct samples it had not seen during training. SNN-PC suggests biologically plausible mechanisms by which the brain may perform perceptual inference and learning in an unsupervised manner. In addition, it may be used in neuromorphic applications that can utilize its energy-efficient, event-driven, local learning, and parallel information processing nature.

Probabilistic causal reasoning under time pressure

While causal reasoning is a core facet of our cognitive abilities, its time-course has not received proper attention. As the duration of reasoning might prove crucial in understanding the underlying cognitive processes, we asked participants in two experiments to make probabilistic causal inferences while manipulating time pressure. We found that participants are less accurate under time pressure, a speed-accuracy-tradeoff, and that they respond more conservatively. Surprisingly, two other persistent reasoning errors-Markov violations and failures to explain away-appeared insensitive to time pressure. These observations seem related to confidence: Conservative inferences were associated with low confidence, whereas Markov violations and failures to explain were not. These findings challenge existing theories that predict an association between time pressure and all causal reasoning errors including conservatism. Our findings suggest that these errors should not be attributed to a single cognitive mechanism and emphasize that causal judgements are the result of multiple processes.

A computationally informed distinction of interoception and exteroception

While interoception is of major neuroscientific interest, its precise definition and delineation from exteroception continue to be debated. Here, we propose a functional distinction between interoception and exteroception based on computational concepts of sensor-effector loops. Under this view, the classification of sensory inputs as serving interoception or exteroception depends on the sensor-effector loop they feed into, for the control of either bodily (physiological and biochemical) or environmental states. We explain the utility of this perspective by examining the perception of skin temperature, one of the most challenging cases for distinguishing between interoception and exteroception. Specifically, we propose conceptualising thermoception as inference about the thermal state of the body (including the skin), which is directly coupled to thermoregulatory processes. This functional view emphasises the coupling to regulation (control) as a defining property of perception (inference) and connects the definition of interoception to contemporary computational theories of brain-body interactions.

Modeling road user response timing in naturalistic traffic conflicts: A surprise-based framework

There is currently no established method for evaluating human response timing across a range of naturalistic traffic conflict types. Traditional notions derived from controlled experiments, such as perception-response time, fail to account for the situation-dependency of human responses and offer no clear way to define the stimulus in many common traffic conflict scenarios. As a result, they are not well suited for application in naturalistic settings. We present a novel framework for measuring and modeling response times in naturalistic traffic conflicts applicable to automated driving systems as well as other traffic safety domains. The framework suggests that response timing must be understood relative to the subject's current (prior) belief and is always embedded in, and dependent on, the dynamically evolving situation. The response process is modeled as a belief update process driven by perceived violations to this prior belief, that is, by surprising stimuli. The framework resolves two key limitations with traditional notions of response time when applied in naturalistic scenarios: (1) The strong situation dependence of response timing and (2) how to unambiguously define the stimulus. Resolving these issues is a challenge that must be addressed by any response timing model intended to be applied in naturalistic traffic conflicts. We show how the framework can be implemented by means of a relatively simple heuristic model fit to naturalistic human response data from real crashes and near crashes from the SHRP2 dataset and discuss how it is, in principle, generalizable to any traffic conflict scenario. We also discuss how the response timing framework can be implemented computationally based on evidence accumulation enhanced by machine learning-based generative models and the information-theoretic concept of surprise.

Predictive coding networks for temporal prediction

One of the key problems the brain faces is inferring the state of the world from a sequence of dynamically changing stimuli, and it is not yet clear how the sensory system achieves this task. A well-established computational framework for describing perceptual processes in the brain is provided by the theory of predictive coding. Although the original proposals of predictive coding have discussed temporal prediction, later work developing this theory mostly focused on static stimuli, and key questions on neural implementation and computational properties of temporal predictive coding networks remain open. Here, we address these questions and present a formulation of the temporal predictive coding model that can be naturally implemented in recurrent networks, in which activity dynamics rely only on local inputs to the neurons, and learning only utilises local Hebbian plasticity. Additionally, we show that temporal predictive coding networks can approximate the performance of the Kalman filter in predicting behaviour of linear systems, and behave as a variant of a Kalman filter which does not track its own subjective posterior variance. Importantly, temporal predictive coding networks can achieve similar accuracy as the Kalman filter without performing complex mathematical operations, but just employing simple computations that can be implemented by biological networks. Moreover, when trained with natural dynamic inputs, we found that temporal predictive coding can produce Gabor-like, motion-sensitive receptive fields resembling those observed in real neurons in visual areas. In addition, we demonstrate how the model can be effectively generalized to nonlinear systems. Overall, models presented in this paper show how biologically plausible circuits can predict future stimuli and may guide research on understanding specific neural circuits in brain areas involved in temporal prediction.

Response sub-additivity and variability quenching in visual cortex

Sub-additivity and variability are ubiquitous response motifs in the primary visual cortex (V1). Response sub-additivity enables the construction of useful interpretations of the visual environment, whereas response variability indicates the factors that limit the precision with which the brain can do this. There is increasing evidence that experimental manipulations that elicit response sub-additivity often also quench response variability. Here, we provide an overview of these phenomena and suggest that they may have common origins. We discuss empirical findings and recent model-based insights into the functional operations, computational objectives and circuit mechanisms underlying V1 activity. These different modelling approaches all predict that response sub-additivity and variability quenching often co-occur. The phenomenology of these two response motifs, as well as many of the insights obtained about them in V1, generalize to other cortical areas. Thus, the connection between response sub-additivity and variability quenching may be a canonical motif across the cortex.

Surprisal From Language Models Can Predict ERPs in Processing Predicate-Argument Structures Only if Enriched by an Agent Preference Principle

Language models based on artificial neural networks increasingly capture key aspects of how humans process sentences. Most notably, model-based surprisals predict event-related potentials such as N400 amplitudes during parsing. Assuming that these models represent realistic estimates of human linguistic experience, their success in modeling language processing raises the possibility that the human processing system relies on no other principles than the general architecture of language models and on sufficient linguistic input. Here, we test this hypothesis on N400 effects observed during the processing of verb-final sentences in German, Basque, and Hindi. By stacking Bayesian generalised additive models, we show that, in each language, N400 amplitudes and topographies in the region of the verb are best predicted when model-based surprisals are complemented by an Agent Preference principle that transiently interprets initial role-ambiguous noun phrases as agents, leading to reanalysis when this interpretation fails. Our findings demonstrate the need for this principle independently of usage frequencies and structural differences between languages. The principle has an unequal force, however. Compared to surprisal, its effect is weakest in German, stronger in Hindi, and still stronger in Basque. This gradient is correlated with the extent to which grammars allow unmarked NPs to be patients, a structural feature that boosts reanalysis effects. We conclude that language models gain more neurobiological plausibility by incorporating an Agent Preference. Conversely, theories of human processing profit from incorporating surprisal estimates in addition to principles like the Agent Preference, which arguably have distinct evolutionary roots.

A unifying theory explains seemingly contradictory biases in perceptual estimation

Perceptual biases are widely regarded as offering a window into the neural computations underlying perception. To understand these biases, previous work has proposed a number of conceptually different, and even seemingly contradictory, explanations, including attraction to a Bayesian prior, repulsion from the prior due to efficient coding and central tendency effects on a bounded range. We present a unifying Bayesian theory of biases in perceptual estimation derived from first principles. We demonstrate theoretically an additive decomposition of perceptual biases into attraction to a prior, repulsion away from regions with high encoding precision and regression away from the boundary. The results reveal a simple and universal rule for predicting the direction of perceptual biases. Our theory accounts for, and yields, new insights regarding biases in the perception of a variety of stimulus attributes, including orientation, color and magnitude. These results provide important constraints on the neural implementations of Bayesian computations.

Visualizing threat and trustworthiness prior beliefs in face perception in high versus low paranoia

Predictive processing accounts of psychosis conceptualize delusions as overly strong learned expectations (prior beliefs) that shape cognition and perception. Paranoia, the most prevalent form of delusions, involves threat prior beliefs that are inherently social. Here, we investigated whether paranoia is related to overly strong threat prior beliefs in face perception. Participants with subclinical levels of high (n = 109) versus low (n = 111) paranoia viewed face stimuli paired with written descriptions of threatening versus trustworthy behaviors, thereby activating their threat versus trustworthiness prior beliefs. Subsequently, they completed an established social-psychological reverse correlation image classification (RCIC) paradigm. This paradigm used participants' responses to randomly varying face stimuli to generate individual classification images (ICIs) that intend to visualize either facial prior belief (threat vs. trust). An independent sample (n = 76) rated these ICIs as more threatening in the threat compared to the trust condition, validating the causal effect of prior beliefs on face perception. Contrary to expectations derived from predictive processing accounts, there was no evidence for a main effect of paranoia. This finding suggests that paranoia was not related to stronger threat prior beliefs that directly affected face perception, challenging the assumption that paranoid beliefs operate on a perceptual level.

Active Inference and Social Actors: Towards a Neuro-Bio-Social Theory of Brains and Bodies in Their Worlds

Although research including biological concepts and variables has gained more prominence in sociology, progress assimilating the organ of experience, the brain, has been theoretically and technically challenging. Formal uptake and assimilation have thus been slow. Within psychology and neuroscience, the traditional brain, which has made brief appearances in sociological research, is a "bottom-up" processor in which sensory signals are passed up the neural hierarchy where they are eventually cognitively and emotionally processed, after which actions and responses are generated. In this paper, we introduce the Active Inference Framework (AIF), which casts the brain as a Bayesian "inference engine" that tests its "top-down" predictive models against "bottom-up" sensory error streams in its attempts to resolve uncertainty and make the world more predictable. After assembling and presenting key concepts in the AIF, we describe an integrated neuro-bio-social model that prioritizes the microsociological assertion that the scene of action is the situation, wherein brains enculturate. Through such social dynamics, enculturated brains share models of the world with one another, enabling collective realities that disclose the actions afforded in those times and places. We conclude by discussing this neuro-bio-social model within the context of exemplar sociological research areas, including the sociology of stress and health, the sociology of emotions, and cognitive cultural sociology, all areas where the brain has received some degree of recognition and incorporation. In each case, sociological insights that do not fit naturally with the traditional brain model emerge intuitively from the predictive AIF model, further underscoring the interconnections and interdependencies between these areas, while also providing a foundation for a probabilistic sociology.

The police hunch: the Bayesian brain, active inference, and the free energy principle in action

In the realm of law enforcement, the "police hunch" has long been a mysterious but crucial aspect of decision-making. Drawing on the developing framework of Active Inference from cognitive science, this theoretical article examines the genesis, mechanics, and implications of the police hunch. It argues that hunches - often vital in high-stakes situations - should not be described as mere intuitions, but as intricate products of our mind's generative models. These models, shaped by observations of the social world and assimilated and enacted through active inference, seek to reduce surprise and make hunches an indispensable tool for officers, in exactly the same way that hypotheses are indispensable for scientists. However, the predictive validity of hunches is influenced by a range of factors, including experience and bias, thus warranting critical examination of their reliability. This article not only explores the formation of police hunches but also provides practical insights for officers and researchers on how to harness the power of active inference to fully understand policing decisions and subsequently explore new avenues for future research.

Tracking subjects' strategies in behavioural choice experiments at trial resolution

Investigating how, when, and what subjects learn during decision-making tasks requires tracking their choice strategies on a trial-by-trial basis. Here, we present a simple but effective probabilistic approach to tracking choice strategies at trial resolution using Bayesian evidence accumulation. We show this approach identifies both successful learning and the exploratory strategies used in decision tasks performed by humans, non-human primates, rats, and synthetic agents. Both when subjects learn and when rules change the exploratory strategies of win-stay and lose-shift, often considered complementary, are consistently used independently. Indeed, we find the use of lose-shift is strong evidence that subjects have latently learnt the salient features of a new rewarded rule. Our approach can be extended to any discrete choice strategy, and its low computational cost is ideally suited for real-time analysis and closed-loop control.

People with a tobacco use disorder exhibit misaligned Bayesian belief updating by falsely attributing non-drug cues as worse predictors of positive outcomes compared to drug cues

Adaptive behaviours depend on dynamically updating internal representations of the world based on the ever-changing environmental contingencies. People with a substance use disorder (pSUD) show maladaptive behaviours with high persistence in drug-taking, despite severe negative consequences. We recently proposed a salience misattribution model for addiction (SMMA; Kalhan et al., 2021), arguing that pSUD have aberrations in their updating processes where drug cues are misattributed as strong predictors of positive outcomes, but weaker predictors of negative outcomes. We also argued that conversely, non-drug cues are misattributed as weak predictors of positive outcomes, but stronger predictors of negative outcomes. We tested these hypotheses using a multi-cue reversal learning task, with reversals in whether drug or non-drug cues are relevant in predicting the outcome (monetary win or loss). We show that people with a tobacco use disorder (pTUD), do form misaligned internal representations. We found that pTUD updated less towards learning the drug cue's relevance in predicting a loss. Further, when neither drug nor non-drug cue predicted a win, pTUD updated more towards the drug cue being relevant predictors of that win. Our Bayesian belief updating model revealed that pTUD had a low estimated likelihood of non-drug cues being predictors of wins, compared to drug cues, which drove the misaligned updating. Overall, several hypotheses of the SMMA were supported, but not all. Our results implicate that strengthening the non-drug cue association with positive outcomes may help restore the misaligned internal representation in pTUD, and offers a quantifiable, computational account of these updating processes.

From psychophysiology to brain imaging: forty-five years MMN history of investigating acoustic change sensitivity

Forty-five years have passed since the first publication of the mismatch negativity (MMN) event-related brain potential (ERP) component. The first 10 years of research hardly gained any particular attention of the scientific community interested in acoustic perception. Debates on the nature of sensation versus perception were going on, and the technical possibilities to record ERPs, called in general evoked potentials, were very limited. Subtle changes in pure tone frequency or intensity giving rise to the MMN component were first investigated in humans. The background of the theoretical model developed by Risto Näätänen was the orientation reaction model of E.N. Sokolov published in 1963 so that the MMN was seen first as an electrophysiological correlate of auditory change detection. This fundamental ability of the auditory system seen as crucial for survival led to the development of the first animal model of the MMN (Csépe et al. in Clin Neurophysiol 66: 571-578, 1987). Indeed, it was confirmed that the MMN was the brain correlate of subtle changes detected that might alert to potential threats in the environment and direct the behavioral orientation. The investigations performed after 2000 introduced complex models and more sophisticated methods, both in animal and human studies, so that the MMN method was on the way to become a tool on the first place and not the main goal of research. This approach was further strengthened by the increasing number of studies on different clinical populations aiming at future applications. The aim of our review is to describe and redefine what the MMN may reflect in auditory perception and to show why and how this brain correlate of changes in the auditory scene can be used as a valuable tool in cognitive neuroscience research. We refer to publications selected to underly the argument the MMN cannot be classified anymore as a sign of simple change detection and not all the indicators used to confirm how genuine the MMN elicited by variations of tones are valid for those to  speech contrasts. We provide a fresh view on the broadly used MMN models, provided by some influential publications as well as on the unwritten history of MMN research aiming to give revised picture on what the MMN may truly reflect. We show how the focus and terminology of the MMN research have changed and what kind of misunderstandings and seemingly contradictive results prevent the MMN community to accept a generally usable cognitive model.

The Human Affectome

Over the last decades, theoretical perspectives in the interdisciplinary field of the affective sciences have proliferated rather than converged due to differing assumptions about what human affective phenomena are and how they work. These metaphysical and mechanistic assumptions, shaped by academic context and values, have dictated affective constructs and operationalizations. However, an assumption about the purpose of affective phenomena can guide us to a common set of metaphysical and mechanistic assumptions. In this capstone paper, we home in on a nested teleological principle for human affective phenomena in order to synthesize metaphysical and mechanistic assumptions. Under this framework, human affective phenomena can collectively be considered algorithms that either adjust based on the human comfort zone (affective concerns) or monitor those adaptive processes (affective features). This teleologically-grounded framework offers a principled agenda and launchpad for both organizing existing perspectives and generating new ones. Ultimately, we hope the Human Affectome brings us a step closer to not only an integrated understanding of human affective phenomena, but an integrated field for affective research.

Test-retest reliability of Bayesian estimations of the effects of stimulation, prior information and individual traits on pain perception

BACKGROUND

There is inter-individual variability in the influence of different components (e.g. nociception and expectations) on pain perception. Identifying the individual effect of these components could serve for patient stratification, but only if these influences are stable in time.

METHODS

In this study, 30 healthy participants underwent a cognitive pain paradigm in which they rated pain after viewing a probabilistic cue informing of forthcoming pain intensity and then receiving electrical stimulation. The trial information was then used in a Bayesian probability model to compute the relative weight each participant put on stimulation, cue, cue uncertainty and trait-like bias. The same procedure was repeated 2 weeks later. Relative and absolute test-retest reliability of all measures was assessed.

RESULTS

Intraclass correlation results showed good reliability for the effect of the stimulation (0.83), the effect of the cue (0.75) and the trait-like bias (0.75 and 0.75), and a moderate reliability for the effect of the cue uncertainty (0.55). Absolute reliability measures also supported the temporal stability of the results and indicated that a change in parameters corresponding to a difference in pain ratings ranging between 0.47 and 1.45 (depending on the parameters) would be needed to consider differences in outcomes significant. The comparison of these measures with the closest clinical data we possess supports the reliability of our results.

CONCLUSIONS

These findings support the hypothesis that inter-individual differences in the weight placed on different pain factors are stable in time and could therefore be a possible target for patient stratification.

SIGNIFICANCE

Our results demonstrate the temporal stability of the weight healthy individuals place on the different factors leading to the pain response. These findings give validity to the idea of using Bayesian estimations of the influence of different factors on pain as a way to stratify patients for treatment personalization.

Bayesian inference is facilitated by modular neural networks with different time scales

Various animals, including humans, have been suggested to perform Bayesian inferences to handle noisy, time-varying external information. In performing Bayesian inference by the brain, the prior distribution must be acquired and represented by sampling noisy external inputs. However, the mechanism by which neural activities represent such distributions has not yet been elucidated. Our findings reveal that networks with modular structures, composed of fast and slow modules, are adept at representing this prior distribution, enabling more accurate Bayesian inferences. Specifically, the modular network that consists of a main module connected with input and output layers and a sub-module with slower neural activity connected only with the main module outperformed networks with uniform time scales. Prior information was represented specifically by the slow sub-module, which could integrate observed signals over an appropriate period and represent input means and variances. Accordingly, the neural network could effectively predict the time-varying inputs. Furthermore, by training the time scales of neurons starting from networks with uniform time scales and without modular structure, the above slow-fast modular network structure and the division of roles in which prior knowledge is selectively represented in the slow sub-modules spontaneously emerged. These results explain how the prior distribution for Bayesian inference is represented in the brain, provide insight into the relevance of modular structure with time scale hierarchy to information processing, and elucidate the significance of brain areas with slower time scales.

An ERP study on the certainty of epistemic modality in predictive inference processing

Previous psychological experiments have shown that predictive inference processing under different textual constraints is modulated by the directionality function of epistemic modality (EM) certainty within the context. Nevertheless, recent neuroscientific studies have not presented positive evidence for such a function during text reading. Consequently, the current study deposited Chinese EMs "" (possibly) and "" (surely) into a predictive inference context to examine whether a directionality of EM certainty influences the processing of predictive inference via the ERP technique. Two independent variables, namely textual constraint and EM certainty, were manipulated, and 36 participants were recruited. The results revealed that, in the anticipatory stage of predictive inference processing while under a weak textual constraint, low certainty evoked a larger N400 (300-500 ms) in the fronto-central and centro-parietal regions, indicating the augmentation of cognitive loads in calculating the possibility of representations of the forthcoming information. Meanwhile, high certainty elicited a right fronto-central late positive component (LPC) (500-700 ms) associated with semantically congruent but lexically unpredicted words. In the integration stage, low certainty resulted in larger right fronto-central and centro-frontal N400 (300-500 ms) effects in the weak textual constraint condition, associated with the facilitation of lexical-semantic retrieval or pre-activation, and high certainty successively elicited right fronto-central and centro-parietal LPC (500-700 ms) effects, associated respectively with lexical unpredictability and reanalysis of the sentence meaning. The results support the directionality function of EM certainty and reveal the complete neural processing of predictive inferences with high and low certainties under different textual constraint conditions.

Bayesianism and wishful thinking are compatible

Bayesian principles show up across many domains of human cognition, but wishful thinking-where beliefs are updated in the direction of desired outcomes rather than what the evidence implies-seems to threaten the universality of Bayesian approaches to the mind. In this Article, we show that Bayesian optimality and wishful thinking are, despite first appearances, compatible. The setting of opposing goals can cause two groups of people with identical prior beliefs to reach opposite conclusions about the same evidence through fully Bayesian calculations. We show that this is possible because, when people set goals, they receive privileged information in the form of affective experiences, and this information systematically supports goal-consistent conclusions. We ground this idea in a formal, Bayesian model in which affective prediction errors drive wishful thinking. We obtain empirical support for our model across five studies.

Are depressive symptoms linked to a reduced pupillary response to novel positive information?-An eye tracking proof-of-concept study

Introduction

Depressive symptoms have been linked to difficulties in revising established negative beliefs in response to novel positive information. Recent predictive processing accounts have suggested that this bias in belief updating may be related to a blunted processing of positive prediction errors at the neural level. In this proof-of-concept study, pupil dilation in response to unexpected positive emotional information was examined as a psychophysiological marker of an attenuated processing of positive prediction errors associated with depressive symptoms.

Methods

Participants (N = 34) completed a modified version of the emotional Bias Against Disconfirmatory Evidence (BADE) task in which scenarios initially suggest negative interpretations that are later either confirmed or disconfirmed by additional information. Pupil dilation in response to the confirmatory and disconfirmatory information was recorded.

Results

Behavioral results showed that depressive symptoms were related to difficulties in revising negative interpretations despite disconfirmatory positive information. The eye tracking results pointed to a reduced pupil response to unexpected positive information among people with elevated depressive symptoms.

Discussion

Altogether, the present study demonstrates that the adapted emotional BADE task can be appropriate for examining psychophysiological aspects such as changes in pupil size along with behavioral responses. Furthermore, the results suggest that depression may be characterized by deviations in both behavioral (i.e., reduced updating of negative beliefs) and psychophysiological (i.e., decreased pupil dilation) responses to unexpected positive information. Future work should focus on a larger sample including clinically depressed patients to further explore these findings.

Amyloid induced hyperexcitability in default mode network drives medial temporal hyperactivity and early tau accumulation

In early Alzheimer's disease (AD) β-amyloid (Aβ) deposits throughout association cortex and tau appears in the entorhinal cortex (EC). Why these initially appear in disparate locations is not understood. Using task-based fMRI and multimodal PET imaging, we assess the impact of local AD pathology on network-to-network interactions. We show that AD pathologies flip interactions between the default mode network (DMN) and the medial temporal lobe (MTL) from inhibitory to excitatory. The DMN is hyperexcited with increasing levels of Aβ, which drives hyperexcitability within the MTL and this directed hyperexcitation of the MTL by the DMN predicts the rate of tau accumulation within the EC. Our results support a model whereby Aβ induces disruptions to local excitatory-inhibitory balance in the DMN, driving hyperexcitability in the MTL, leading to tau accumulation. We propose that Aβ-induced disruptions to excitatory-inhibitory balance is a candidate causal route between Aβ and remote EC-tau accumulation.

Enhanced electrophysiological responses to explicitly predicted and pre-imagined inputs, with confirmation from online decoding with neuro-feedback

Neural responses to sensory inputs can scale with the likelihood of encountering the input. This is consistent with the predictive coding framework, in that the human brain is expected to be less responsive to predicted inputs. Typically, however, prediction is not explicitly measured. It is inferred from the probability of encountering an event. When an input is explicitly predicted, responses to predicted inputs can be enhanced. Here, we ask if this effect can be ascribed to a generic priming effect, from pre-cogitating about one of two possible inputs. Consistent with this, we find that P300s (a relatively late event-related potential measured with electroencephalography) are greater for explicitly predicted audio and visual inputs, and that this effect cannot be distinguished from a priming effect from pre-imagining audio or visual presentations. Evidence indicates that participants engaged in pre-imagining presentations, as we were able to decode online what type of presentation (audio or visual) they were imagining with a high success rate (approx. 73%), and we encouraged compliance with neuro-feedback regarding this success rate. Our data confirm that human cortex can be more responsive to inputs that have been subject to pre-cogitation-including explicit predictions. This highlights that while anticipatory processes can reduce responding to likely inputs, they can also enhance responding to explicitly predicted inputs.

Emergent human-like covert attention in feedforward convolutional neural networks

Covert attention allows the selection of locations or features of the visual scene without moving the eyes. Cues and contexts predictive of a target's location orient covert attention and improve perceptual performance. The performance benefits are widely attributed to theories of covert attention as a limited resource, zoom, spotlight, or weighting of visual information. However, such concepts are difficult to map to neuronal populations. We show that a feedforward convolutional neural network (CNN) trained on images to optimize target detection accuracy and with no explicit incorporation of an attention mechanism, a limited resource, or feedback connections learns to utilize cues and contexts in the three most prominent covert attention tasks (Posner cueing, set size effects in search, and contextual cueing) and predicts the cue/context influences on human accuracy. The CNN's cueing/context effects generalize across network training schemes, to peripheral and central pre-cues, discrimination tasks, and reaction time measures, and critically do not vary with reductions in network resources (size). The CNN shows comparable cueing/context effects to a model that optimally uses image information to make decisions (Bayesian ideal observer) but generalizes these effects to cue instances unseen during training. Together, the findings suggest that human-like behavioral signatures of covert attention in the three landmark paradigms might be an emergent property of task accuracy optimization in neuronal populations without positing limited attentional resources. The findings might explain recent behavioral results showing cueing and context effects across a variety of simple organisms with no neocortex, from archerfish to fruit flies.

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.

Aesthetics and predictive processing: grounds and prospects of a fruitful encounter

In the last few years, a remarkable convergence of interests and results has emerged between scholars interested in the arts and aesthetics from a variety of perspectives and cognitive scientists studying the mind and brain within the predictive processing (PP) framework. This convergence has so far proven fruitful for both sides: while PP is increasingly adopted as a framework for understanding aesthetic phenomena, the arts and aesthetics, examined under the lens of PP, are starting to be seen as important windows into our mental functioning. The result is a vast and fast-growing research programme that promises to deliver important insights into our aesthetic encounters as well as a wide range of psychological phenomena of general interest. Here, we present this developing research programme, describing its grounds and highlighting its prospects. We start by clarifying how the study of the arts and aesthetics encounters the PP picture of mental functioning (§1). We then go on to outline the prospects of this encounter for the fields involved: philosophy and history of art (§2), psychology of aesthetics and neuroaesthetics (§3) and psychology and neuroscience more generally (§4). The upshot is an ambitious but well-defined framework within which aesthetics and cognitive science can partner up to illuminate crucial aspects of the human mind. This article is part of the theme issue 'Art, aesthetics and predictive processing: theoretical and empirical perspectives'.

Development, Insults and Predisposing Factors of the Brain's Predictive Coding System to Chronic Perceptual Disorders-A Life-Course Examination

The predictive coding theory is currently widely accepted as the theoretical basis of perception and chronic perceptual disorders are explained as the maladaptive compensation of the brain to a prediction error. Although this gives us a general framework to work with, it is still not clear who may be more susceptible and/or vulnerable to aberrations in this system. In this paper, we study changes in predictive coding through the lens of tinnitus and pain. We take a step back to understand how the predictive coding system develops from infancy, what are the different neural and bio markers that characterise this system in the acute, transition and chronic phases and what may be the factors that pose a risk to the aberration of this system. Through this paper, we aim to identify people who may be at a higher risk of developing chronic perceptual disorders as a reflection of aberrant predictive coding, thereby giving future studies more facets to incorporate in their investigation of early markers of tinnitus, pain and other disorders of predictive coding. We therefore view this paper to encourage the thinking behind the development of preclinical biomarkers to maladaptive predictive coding.

Dynamic causal model application on hierarchical human motor control estimation in visuomotor tasks

Introduction

In brain function research, each brain region has been investigated independently, and how different parts of the brain work together has been examined using the correlations among them. However, the dynamics of how different brain regions interact with each other during time-varying tasks, such as voluntary motion tasks, are still not well-understood.

Methods

To address this knowledge gap, we conducted functional magnetic resonance imaging (fMRI) using target tracking tasks with and without feedback. We identified the motor cortex, cerebellum, and visual cortex by using a general linear model during the tracking tasks. We then employed a dynamic causal model (DCM) and parametric empirical Bayes to quantitatively elucidate the interactions among the left motor cortex (ML), right cerebellum (CBR) and left visual cortex (VL), and their roles as higher and lower controllers in the hierarchical model.

Results

We found that the tracking task with visual feedback strongly affected the modulation of connection strength in ML → CBR and ML↔VL. Moreover, we found that the modulation of VL → ML, ML → ML, and ML → CBR by the tracking task with visual feedback could explain individual differences in tracking performance and muscle activity, and we validated these findings by leave-one-out cross-validation.

Discussion

We demonstrated the effectiveness of our approach for understanding the mechanisms underlying human motor control. Our proposed method may have important implications for the development of new technologies in personalized interventions and technologies, as it sheds light on how different brain regions interact and work together during a motor task.

Modelling phenomenological differences in aetiologically distinct visual hallucinations using deep neural networks

Visual hallucinations (VHs) are perceptions of objects or events in the absence of the sensory stimulation that would normally support such perceptions. Although all VHs share this core characteristic, there are substantial phenomenological differences between VHs that have different aetiologies, such as those arising from Neurodegenerative conditions, visual loss, or psychedelic compounds. Here, we examine the potential mechanistic basis of these differences by leveraging recent advances in visualising the learned representations of a coupled classifier and generative deep neural network-an approach we call 'computational (neuro)phenomenology'. Examining three aetiologically distinct populations in which VHs occur-Neurodegenerative conditions (Parkinson's Disease and Lewy Body Dementia), visual loss (Charles Bonnet Syndrome, CBS), and psychedelics-we identified three dimensions relevant to distinguishing these classes of VHs: realism (veridicality), dependence on sensory input (spontaneity), and complexity. By selectively tuning the parameters of the visualisation algorithm to reflect influence along each of these phenomenological dimensions we were able to generate 'synthetic VHs' that were characteristic of the VHs experienced by each aetiology. We verified the validity of this approach experimentally in two studies that examined the phenomenology of VHs in Neurodegenerative and CBS patients, and in people with recent psychedelic experience. These studies confirmed the existence of phenomenological differences across these three dimensions between groups, and crucially, found that the appropriate synthetic VHs were rated as being representative of each group's hallucinatory phenomenology. Together, our findings highlight the phenomenological diversity of VHs associated with distinct causal factors and demonstrate how a neural network model of visual phenomenology can successfully capture the distinctive visual characteristics of hallucinatory experience.

Individual prediction tendencies do not generalize across modalities

Predictive processing theories, which model the brain as a "prediction machine", explain a wide range of cognitive functions, including learning, perception and action. Furthermore, it is increasingly accepted that aberrant prediction tendencies play a crucial role in psychiatric disorders. Given this explanatory value for clinical psychiatry, prediction tendencies are often implicitly conceptualized as individual traits or as tendencies that generalize across situations. As this has not yet explicitly been shown, in the current study, we quantify to what extent the individual tendency to anticipate sensory features of high probability generalizes across modalities. Using magnetoencephalography (MEG), we recorded brain activity while participants were presented with a sequence of four different (either visual or auditory) stimuli, which changed according to predefined transitional probabilities of two entropy levels: ordered vs. random. Our results show that, on a group-level, under conditions of low entropy, stimulus features of high probability are preactivated in the auditory but not in the visual modality. Crucially, the magnitude of the individual tendency to predict sensory events seems not to correlate between the two modalities. Furthermore, reliability statistics indicate poor internal consistency, suggesting that the measures from the different modalities are unlikely to reflect a single, common cognitive process. In sum, our findings suggest that quantification and interpretation of individual prediction tendencies cannot be generalized across modalities.

Multisensory Calibration: A Variety of Slow and Fast Brain Processes Throughout the Lifespan

From before we are born, throughout development, adulthood, and aging, we are immersed in a multisensory world. At each of these stages, our sensory cues are constantly changing, due to body, brain, and environmental changes. While integration of information from our different sensory cues improves precision, this only improves accuracy if the underlying cues are unbiased. Thus, multisensory calibration is a vital and ongoing process. To meet this grand challenge, our brains have evolved a variety of mechanisms. First, in response to a systematic discrepancy between sensory cues (without external feedback) the cues calibrate one another (unsupervised calibration). Second, multisensory function is calibrated to external feedback (supervised calibration). These two mechanisms superimpose. While the former likely reflects a lower level mechanism, the latter likely reflects a higher level cognitive mechanism. Indeed, neural correlates of supervised multisensory calibration in monkeys were found in higher level multisensory cortical area VIP, but not in the relatively lower level multisensory area MSTd. In addition, even without a cue discrepancy (e.g., when experiencing stimuli from different sensory cues in series) the brain monitors supra-modal statistics of events in the environment and adapts perception cross-modally. This too comprises a variety of mechanisms, including confirmation bias to prior choices, and lower level cross-sensory adaptation. Further research into the neuronal underpinnings of the broad and diverse functions of multisensory calibration, with improved synthesis of theories is needed to attain a more comprehensive understanding of multisensory brain function.

Decentralized Neural Circuits of Multisensory Information Integration in the Brain

The brain combines multisensory inputs together to obtain a complete and reliable description of the world. Recent experiments suggest that several interconnected multisensory brain areas are simultaneously involved to integrate multisensory information. It was unknown how these mutually connected multisensory areas achieve multisensory integration. To answer this question, using biologically plausible neural circuit models we developed a decentralized system for information integration that comprises multiple interconnected multisensory brain areas. Through studying an example of integrating visual and vestibular cues to infer heading direction, we show that such a decentralized system is well consistent with experimental observations. In particular, we demonstrate that this decentralized system can optimally integrate information by implementing sampling-based Bayesian inference. The Poisson variability of spike generation provides appropriate variability to drive sampling, and the interconnections between multisensory areas store the correlation prior between multisensory stimuli. The decentralized system predicts that optimally integrated information emerges locally from the dynamics of the communication between brain areas and sheds new light on the interpretation of the connectivity between multisensory brain areas.

"What we talk about is creating a probability": Exploring the interaction between the anticipation and decision-making processes of professional bowlers and batters in Twenty20 cricket

Expert performers in time constrained sports use a range of information sources to facilitate anticipatory and decision-making processes. However, research has often focused on responders such as batters, goalkeepers, defenders, and returners of serve, and failed to capture the complex interaction between opponents, where responders can also manipulate probabilities in their favour. This investigation aimed to explore the interaction between top order batters and fast or medium paced bowlers in cricket and the information they use to inform their anticipatory and decision-making skills in Twenty20 competition. Eleven professional cricketers were interviewed (8 batters and 3 bowlers) using semi-structured questions and scenarios from Twenty20 matches. An inductive and deductive thematic analysis was conducted using the overarching themes of Situation Awareness (SA) and Option Awareness (OA). Within SA, the sub-themes identified related to information sources used by bowlers and batters (i.e., stable contextual information, dynamic contextual information, kinematic information). Within OA, the sub-themes identified highlighted how cricketers use these information sources to understand the options available and the likelihood of success associated with each option (e.g., risk and reward, personal strengths). A sub-theme of 'responder manipulation' was also identified within OA to provide insight into how batters and bowlers interact in a cat-and-mouse like manner to generate options that manipulate one another throughout the competition. A schematic has been developed based on the study findings to illustrate the complex interaction between the anticipation and decision-making processes of professional top order batters and fast or medium paced bowlers in Twenty20 cricket.

Psychedelia: The interplay of music and psychedelics

Music and psychedelics have been intertwined throughout the existence of Homo sapiens, from the early shamanic rituals of the Americas and Africa to the modern use of psychedelic-assisted therapy for a variety of mental health conditions. Across such settings, music has been highly prized for its ability to guide the psychedelic experience. Here, we examine the interplay between music and psychedelics, starting by describing their association with the brain's functional hierarchy that is relied upon for music perception and its psychedelic-induced manipulation, as well as an exploration of the limited research on their mechanistic neural overlap. We explore music's role in Western psychedelic therapy and the use of music in indigenous psychedelic rituals, with a specific focus on ayahuasca and the Santo Daime Church. Furthermore, we explore work relating to the evolution and onset of music and psychedelic use. Finally, we consider music's potential to lead to altered states of consciousness in the absence of psychedelics as well as the development of psychedelic music. Here, we provide an overview of several perspectives on the interaction between psychedelic use and music-a topic with growing interest given increasing excitement relating to the therapeutic efficacy of psychedelic interventions.

From Multisensory Integration to Multisensory Decision-Making

Organisms live in a dynamic environment in which sensory information from multiple sources is ever changing. A conceptually complex task for the organisms is to accumulate evidence across sensory modalities and over time, a process known as multisensory decision-making. This is a new concept, in terms of that previous researches have been largely conducted in parallel disciplines. That is, much efforts have been put either in sensory integration across modalities using activity summed over a duration of time, or in decision-making with only one sensory modality that evolves over time. Recently, a few studies with neurophysiological measurements emerge to study how different sensory modality information is processed, accumulated, and integrated over time in decision-related areas such as the parietal or frontal lobes in mammals. In this review, we summarize and comment on these studies that combine the long-existed two parallel fields of multisensory integration and decision-making. We show how the new findings provide insight into our understanding about neural mechanisms mediating multisensory information processing in a more complete way.

Nested Selves: Self-Organization and Shared Markov Blankets in Prenatal Development in Humans

The immune system is a central component of organismic function in humans. This paper addresses self-organization of biological systems in relation to-and nested within-other biological systems in pregnancy. Pregnancy constitutes a fundamental state for human embodiment and a key step in the evolution and conservation of our species. While not all humans can be pregnant, our initial state of emerging and growing within another person's body is universal. Hence, the pregnant state does not concern some individuals but all individuals. Indeed, the hierarchical relationship in pregnancy reflects an even earlier autopoietic process in the embryo by which the number of individuals in a single blastoderm is dynamically determined by cell- interactions. The relationship and the interactions between the two self-organizing systems during pregnancy may play a pivotal role in understanding the nature of biological self-organization per se in humans. Specifically, we consider the role of the immune system in biological self-organization in addition to neural/brain systems that furnish us with a sense of self. We examine the complex case of pregnancy, whereby two immune systems need to negotiate the exchange of resources and information in order to maintain viable self-regulation of nested systems. We conclude with a proposal for the mechanisms-that scaffold the complex relationship between two self-organising systems in pregnancy-through the lens of the Active Inference, with a focus on shared Markov blankets.

Alpha oscillations encode Bayesian belief updating underlying attentional allocation in dynamic environments

In a dynamic environment, expectations of the future constantly change based on updated evidence and affect the dynamic allocation of attention. To further investigate the neural mechanisms underlying attentional expectancies, we employed a modified Central Cue Posner Paradigm in which the probability of cues being valid (that is, accurately indicated the upcoming target location) was manipulated. Attentional deployment to the cued location (α), which was governed by precision of predictions on previous trials, was estimated using a hierarchical Bayesian model and was included as a regressor in the analyses of electrophysiological (EEG) data. Our results revealed that before the target appeared, alpha oscillations (8∼13 Hz) for high-predictability cues (88 % valid) were significantly predicted by precision-dependent attention (α). This relationship was not observed under low-predictability conditions (69 % and 50 % valid cues). After the target appeared, precision-dependent attention (α) correlated with alpha band oscillations only in the valid cue condition and not in the invalid condition. Further analysis under conditions of significant attentional modulation by precision suggested a separate effect of cue orientation. These results provide new insights on how trial-by-trial Bayesian belief updating relates to alpha band encoding of environmentally-sensitive allocation of visual spatial attention.

Cognitive biases as an adaptive strategy in autism and schizophrenia spectrum: the compensation perspective on neurodiversity

This article presents a novel theoretical perspective on the role of cognitive biases within the autism and schizophrenia spectrum by integrating the evolutionary and computational approaches. Against the background of neurodiversity, cognitive biases are presented as primary adaptive strategies, while the compensation of their shortcomings is a potential cognitive advantage. The article delineates how certain subtypes of autism represent a unique cognitive strategy to manage cognitive biases at the expense of rapid and frugal heuristics. In contrast, certain subtypes of schizophrenia emerge as distinctive cognitive strategies devised to navigate social interactions, albeit with a propensity for overdetecting intentional behaviors. In conclusion, the paper emphasizes that while extreme manifestations might appear non-functional, they are merely endpoints of a broader, primarily functional spectrum of cognitive strategies. The central argument hinges on the premise that cognitive biases in both autism and schizophrenia spectrums serve as compensatory mechanisms tailored for specific ecological niches.

Predictive coding and stochastic resonance as fundamental principles of auditory phantom perception

Mechanistic insight is achieved only when experiments are employed to test formal or computational models. Furthermore, in analogy to lesion studies, phantom perception may serve as a vehicle to understand the fundamental processing principles underlying healthy auditory perception. With a special focus on tinnitus-as the prime example of auditory phantom perception-we review recent work at the intersection of artificial intelligence, psychology and neuroscience. In particular, we discuss why everyone with tinnitus suffers from (at least hidden) hearing loss, but not everyone with hearing loss suffers from tinnitus. We argue that intrinsic neural noise is generated and amplified along the auditory pathway as a compensatory mechanism to restore normal hearing based on adaptive stochastic resonance. The neural noise increase can then be misinterpreted as auditory input and perceived as tinnitus. This mechanism can be formalized in the Bayesian brain framework, where the percept (posterior) assimilates a prior prediction (brain's expectations) and likelihood (bottom-up neural signal). A higher mean and lower variance (i.e. enhanced precision) of the likelihood shifts the posterior, evincing a misinterpretation of sensory evidence, which may be further confounded by plastic changes in the brain that underwrite prior predictions. Hence, two fundamental processing principles provide the most explanatory power for the emergence of auditory phantom perceptions: predictive coding as a top-down and adaptive stochastic resonance as a complementary bottom-up mechanism. We conclude that both principles also play a crucial role in healthy auditory perception. Finally, in the context of neuroscience-inspired artificial intelligence, both processing principles may serve to improve contemporary machine learning techniques.

Predictions about prosody facilitate lexical access: Evidence from P50/N100 and MMN components

Research into the neural foundation of perception asserts a model where top-down predictions modulate the bottom-up processing of sensory input. Despite becoming increasingly influential in cognitive neuroscience, the precise account of this predictive coding framework remains debated. In this study, we aim to contribute to this debate by investigating how predictions about prosody facilitate speech perception, and to shed light especially on lexical access influenced by simultaneous predictions in different domains, inter alia, prosodic and semantic. Using a passive auditory oddball paradigm, we examined neural responses to prosodic changes, leading to a semantic change as in Dutch nouns canon ['kaːnɔn] 'canon' vs kanon [kaː'nɔn] 'cannon', and used acoustically identical pseudowords as controls. Results from twenty-eight native speakers of Dutch (age range 18-32 years) indicated an enhanced P50/N100 complex to prosodic change in pseudowords as well as an MMN response to both words and pseudowords. The enhanced P50/N100 response to pseudowords is claimed to indicate that all relevant auditory information is still processed by the brain, whereas the reduced response to words might reflect the suppression of information that has already been encoded. The MMN response to pseudowords and words, on the other hand, is best justified by the unification of previously established prosodic representations with sensory and semantic input respectively. This pattern of results is in line with the predictive coding framework acting on multiple levels and is of crucial importance to indicate that predictions about linguistic prosodic information are utilized by the brain as early as 50 ms.

Bayesian encoding and decoding as distinct perspectives on neural coding

The Bayesian brain hypothesis is one of the most influential ideas in neuroscience. However, unstated differences in how Bayesian ideas are operationalized make it difficult to draw general conclusions about how Bayesian computations map onto neural circuits. Here, we identify one such unstated difference: some theories ask how neural circuits could recover information about the world from sensory neural activity (Bayesian decoding), whereas others ask how neural circuits could implement inference in an internal model (Bayesian encoding). These two approaches require profoundly different assumptions and lead to different interpretations of empirical data. We contrast them in terms of motivations, empirical support and relationship to neural data. We also use a simple model to argue that encoding and decoding models are complementary rather than competing. Appreciating the distinction between Bayesian encoding and Bayesian decoding will help to organize future work and enable stronger empirical tests about the nature of inference in the brain.

Probabilistic and rich individual working memories revealed by a betting game

When asked to remember a color, do people remember a point estimate (e.g., a particular shade of red), a point estimate plus an uncertainty estimate, or are memory representations rich probabilistic distributions over feature space? We asked participants to report the color of a circle held in working memory. Rather than collecting a single report per trial, we had participants place multiple bets to create trialwise uncertainty distributions. Bet dispersion correlated with performance, indicating that internal uncertainty guided bet placement. While the first bet was on average the most precisely placed, the later bets systematically shifted the distribution closer to the target, resulting in asymmetrical distributions about the first bet. This resulted in memory performance improvements when averaging across bets, and overall suggests that memory representations contain more information than can be conveyed by a single response. The later bets contained target information even when the first response would generally be classified as a guess or report of an incorrect item, suggesting that such failures are not all-or-none. This paradigm provides multiple pieces of evidence that memory representations are rich and probabilistic. Crucially, standard discrete response paradigms underestimate the amount of information in memory representations.

Hierarchical VAEs provide a normative account of motion processing in the primate brain

The relationship between perception and inference, as postulated by Helmholtz in the 19th century, is paralleled in modern machine learning by generative models like Variational Autoencoders (VAEs) and their hierarchical variants. Here, we evaluate the role of hierarchical inference and its alignment with brain function in the domain of motion perception. We first introduce a novel synthetic data framework, Retinal Optic Flow Learning (ROFL), which enables control over motion statistics and their causes. We then present a new hierarchical VAE and test it against alternative models on two downstream tasks: (i) predicting ground truth causes of retinal optic flow (e.g., self-motion); and (ii) predicting the responses of neurons in the motion processing pathway of primates. We manipulate the model architectures (hierarchical versus non-hierarchical), loss functions, and the causal structure of the motion stimuli. We find that hierarchical latent structure in the model leads to several improvements. First, it improves the linear decodability of ground truth factors and does so in a sparse and disentangled manner. Second, our hierarchical VAE outperforms previous state-of-the-art models in predicting neuronal responses and exhibits sparse latent-to-neuron relationships. These results depend on the causal structure of the world, indicating that alignment between brains and artificial neural networks depends not only on architecture but also on matching ecologically relevant stimulus statistics. Taken together, our results suggest that hierarchical Bayesian inference underlines the brain's understanding of the world, and hierarchical VAEs can effectively model this understanding.

Sampling-based Bayesian inference in recurrent circuits of stochastic spiking neurons

Two facts about cortex are widely accepted: neuronal responses show large spiking variability with near Poisson statistics and cortical circuits feature abundant recurrent connections between neurons. How these spiking and circuit properties combine to support sensory representation and information processing is not well understood. We build a theoretical framework showing that these two ubiquitous features of cortex combine to produce optimal sampling-based Bayesian inference. Recurrent connections store an internal model of the external world, and Poissonian variability of spike responses drives flexible sampling from the posterior stimulus distributions obtained by combining feedforward and recurrent neuronal inputs. We illustrate how this framework for sampling-based inference can be used by cortex to represent latent multivariate stimuli organized either hierarchically or in parallel. A neural signature of such network sampling are internally generated differential correlations whose amplitude is determined by the prior stored in the circuit, which provides an experimentally testable prediction for our framework.

Target motion misjudgments reflect a misperception of the background; revealed using continuous psychophysics

Determining the velocities of target objects as we navigate complex environments is made more difficult by the fact that our own motion adds systematic motion signals to the visual scene. The flow-parsing hypothesis asserts that the background motion is subtracted from visual scenes in such cases as a way for the visual system to determine target motions relative to the scene. Here, we address the question of why backgrounds are only partially subtracted in lab settings. At the same time, we probe a much-neglected aspect of scene perception in flow-parsing studies, that is, the perception of the background itself. Here, we present results from three experienced psychophysical participants and one inexperienced participant who took part in three continuous psychophysics experiments. We show that, when the background optic flow pattern is composed of local elements whose motions are congruent with the global optic flow pattern, the incompleteness of the background subtraction can be entirely accounted for by a misperception of the background. When the local velocities comprising the background are randomly dispersed around the average global velocity, an additional factor is needed to explain the subtraction incompleteness. We show that a model where background perception is a result of the brain attempting to infer scene motion due to self-motion can account for these results.

Bayesian accounts of perceptual decisions in the nonclinical continuum of psychosis: Greater imprecision in both top-down and bottom-up processes

Neurocomputational accounts of psychosis propose mechanisms for how information is integrated into a predictive model of the world, in attempts to understand the occurrence of altered perceptual experiences. Conflicting Bayesian theories postulate aberrations in either top-down or bottom-up processing. The top-down theory predicts an overreliance on prior beliefs or expectations resulting in aberrant perceptual experiences, whereas the bottom-up theory predicts an overreliance on current sensory information, as aberrant salience is directed towards objectively uninformative stimuli. This study empirically adjudicates between these models. We use a perceptual decision-making task in a neurotypical population with varying degrees of psychotic-like experiences. Bayesian modelling was used to compute individuals' reliance on prior relative to sensory information. Across two datasets (discovery dataset n = 363; independent replication in validation dataset n = 782) we showed that psychotic-like experiences were associated with an overweighting of sensory information relative to prior expectations, which seem to be driven by decreased precision afforded to prior information. However, when prior information was more uncertain, participants with greater psychotic-like experiences encoded sensory information with greater noise. Greater psychotic-like experiences were associated with aberrant precision in the encoding both prior and likelihood information, which we suggest may be related to generally heightened perceptions of task instability. Our study lends empirical support to notions of both weaker bottom-up and weaker (rather than stronger) top-down perceptual processes, as well as aberrancies in belief updating that extend into the non-clinical continuum of psychosis.