Awareness reports as inference in a higher-order state space

Awareness reports as inference in a higher-order state space

Stephen M. Fleming
Wellcome Centre for Human Neuroimaging
12 Queen Square
London WC1N 3AR
University College London
stephen.fleming@ucl.ac.uk
July 1, 2019
Abstract

Humans have the ability to report the contents of their subjective experience – we can say to each other, “I am aware of X”. However, the decision processes that support reports about mental contents remain poorly understood. In this article I propose a computational framework that characterises awareness reports as metacognitive decisions (inference) about a generative model of perceptual content. This account is motivated from the perspective of how flexible hierarchical state spaces are built during learning and decision-making. Internal states supporting awareness reports, unlike those covarying with perceptual contents, are simple and abstract, varying along a one-dimensional continuum from absent to present. A critical feature of this architecture is that it is both higher-order and asymmetric: there is a vast number of perceptual states is nested under “present”, but a much smaller number of possible states is nested under “absent”. Via simulations I show that this asymmetry provides a natural account of observations of “global ignition” in brain imaging studies of awareness reports.

 

Awareness reports as inference in a higher-order state space


  Stephen M. Fleming Wellcome Centre for Human Neuroimaging 12 Queen Square London WC1N 3AR University College London stephen.fleming@ucl.ac.uk

July 1, 2019

1 Introduction

Humans have the ability to report the contents of their subjective experience - we can say to each other, “I am aware of X”. Such reports are intended to convey meaning, and unlike other aspects of behaviour, have contents [1]. This unique property of reports makes them central to a science of consciousness, which has focused on measuring and quantifying differences in awareness while holding other aspects of stimuli and behaviour constant [2, 3].

However, the decision processes that support reports about mental contents remain poorly understood. In this article I propose a computational framework that characterises awareness reports as metacognitive decisions (inference) about a generative model of perceptual content [4]. This higher-order state space (HOSS) framework builds on Bayesian approaches to perception that invoke hierarchical probabilistic inference as a route towards efficiently modeling the external world [5, 6, 7, 8].

The outline of the paper is as follows. First, I start by describing the psychological processes that support reports with reference to experimental paradigms commonly used to study conscious awareness. Second, I outline the central hypothesis, that awareness reports reflect a decision about whether a generative model of perceptual contents is representing signal or noise. Third, I use a toy model of a simple perceptual decision to explicate aspects of the framework, and distinguish it from other, related approaches such as signal detection theory (SDT; [9, 10]). Finally, I briefly highlight empirical predictions that flow from the model, and how it relates to existing theories of consciousness such as global workspace and higher-order theories.

2 Psychological basis of awareness reports

Several authors have proposed that the psychological basis of a (visual) awareness report is an internal decision about the visibility of perceptual contents111The same computational considerations likely hold for awareness of other sensory modalities - a focus on visibility here reflects a historical bias towards vision in studies of conscious awareness. [11, 12, 10]. This implies that internal states supporting awareness reports, unlike those covarying with perceptual contents themselves, are both simple and abstract, varying along a one-dimensional continuum from unaware to aware. Note that a better terminology for “unaware” is really “absent”, “unseen” or “noise”, as participants of course remain aware of seeing nothing on trials on which they report “unaware”. Awareness reports also refer to different subsets of perceptual content: for instance, subjects may be asked “did you see the word?”, “did you see the number?” or “did you see anything at all?”. These two features imply that awareness reports are metacognitive decisions about a rich perceptual generative model, rather than a feature of this generative model. We will make this hypothesis more concrete in the next section.

A range of experimental paradigms have been developed to introduce variability in awareness reports while keeping other aspects of stimuli and behaviour fixed (see [13] for a review). For example, using backward masking, Dehaene and colleagues found that they could make words invisible while showing (via priming effects and brain imaging) that they were processed up to a semantic level [14]. When subjects reported consciously seeing the words, whole-brain fMRI showed elevated activations in the parietal and prefrontal cortex, which have become known as “global ignition” responses due to their non-linear response profile in relation to stimulation strength [3, 15].222Since these classic studies, alternative explanations of frontoparietal ignition have been put forward, including that it is involved in the act of reporting, but not conscious awareness, or that it reflects greater performance capacity on conscious trials [16, 17]. These debates are ongoing (see [18, 19], for recent arguments from both sides), but here we put them to one side given that we are content to seek an explanation of report itself.

3 Hypothesis

In common with other predictive processing approaches, we assume that the brain is engaged in building a hierarchical, probabilistic generative model of the world, one in which inference and learning proceed using (approximations of) Bayes’ rule. The novel aspect of the current framework is its focus on report, and explaining how decisions to respond “I am aware of X” or “I am unaware of X” get made.333Note that here I am focusing on reportable states of awareness, and leaving aside the issue of whether non-reportable contents may be conscious [20, 21]. By adopting this stance, we frame a clear question that is answerable by cognitive science: what are the computational processes involved in a report of awareness? [22, 23]

The central hypothesis is:


Awareness reports are decisions about whether our perceptual generative model currently reflects presence (signal) or absence (noise).

These decisions are governed by a second-order (metacognitive) inference about the state of a first-order (perceptual) generative model [4]. One way of implementing this second-order inference is by adding an additional hierarchical state above the perceptual generative model, which we refer to as an “awareness state”. Paralleling the psychological simplicity of awareness reports, this awareness state is also simple, and signals a probability of whether there is signal or noise in the lower layers (corresponding to reports of “present” or “absent”). It is also part of the generative model, such that if the model is run forward, states of presence (vs. absence) lead to the generation of perceptual content in lower layers.

4 Model

We describe the model formally in terms of a probabilistic graphical model [24], where nodes correspond to unknown variables and the graph structure is used to indicate dependencies between variables. These graphs provide a concise description of how sensory data is generated (Figure 1A).

Figure 1: Figure 1. A) Probabilistic graphical model of awareness reports. Nodes represent random variables and the graph structure is used to indicate dependencies as indicated by directed arrows. The shaded node indicates that this variable is observed by the system as sensory input. B) Expanded version of the graphical model from panel (A) that makes explicit the asymmetry in the state space. Figures created using the Daft package in Python.

In this model, is a vector that encodes the relative probabilities of each of discrete perceptual states. is a scalar awareness state encoding the probability of reporting “presence”. is a vector defining the location (mean) of a multivariate Gaussian determined by the currently active state in in a feature space of dimensionality . is a covariance matrix which for simplicity we assume is independent of .

Note that each “perceptual” state here is discrete, but in reality this state space is likely to be multidimensional and also hierarchically organised. is included to simplify the notation, but is a redundant variable that inherits a copy of . When answering the query, “Present or absent?”, the model computes the posterior , marginalising over perceptual states :

(1)

In this architecture, awareness is a higher-order state in a generative model of perceptual contents (Figure 1). As in standard models of perceptual decision-making, inference on contents is also straightforward, and may allow the observer to jointly determine both awareness and contents in response to specific queries.

A critical feature of this architecture is that the state space nested under the awareness state is asymmetric. In the absence of awareness, there is (by definition) an absence of perceptual content (Figure 1B). In contrast, a vast number of potential perceptual states is nested under the awareness state of “presence”. This imposes an asymmetry in the model which we will leverage in the next section when seeking to account for global ignition responses.

4.1 Simulations

To simulate the model I build on previous work using a two-dimensional feature space to capture important features of multidimensional perceptual categorisation [10]. Each axis represents the strength of activation of one of two possible stimulus features, such as leftward or rightward tilted grating orientations (see Figure 2). The origin represents low activation on both features, consistent with no stimulus (or noise) being presented. As in the more general case described in the previous section, each stimulus category generates samples from a multivariate Gaussian whose mean is dominated by one or other feature. Thus if I receive a sample of , I can be confident that I was shown a left-tilted stimulus; if I receive a sample , I can be confident in seeing a right-tilted stimulus.

King and Dehaene [10] showed that by placing different types of decision criteria onto this space, multiple empirical relationships between discrimination performance, confidence and visibility could be simulated. In their model, visibility was modeled as the distance from the origin, and stimulus awareness reflected a first-order (flat) perceptual categorisation in which “absent” was one of several potential stimulus classes (Figure 2A). Our model builds closely on theirs and inherits the benefits of being able to accommodate dissociations between forced-choice responding and subjective reports. However it differs in proposing that awareness is not inherent to perceptual categorisation; instead, perceptual categorisation is nested under a generative model of awareness (Figure 2B)). In other words, unlike in SDT, deciding that a stimulus is “absent” in the HOSS model is governed by a more abstract state than deciding a stimulus is tilted to the left or right. We will see that this seemingly minor change in architecture leads to important consequences for the relationship between awareness and global ignition.

Figure 2: Figure 2. A) Two-dimensional feature space for a toy perceptual decision problem involving classifying two possible stimuli (e.g. left- and right-tilted Gabors). Each Gaussian indicates the likelihood of observing a pair of features (e.g. orientation) given each stimulus class. The right-tilted stimuli occupy the righthand side of the grid; left-tilted stimuli occupy lefthand side of the grid. The absence of stimulation is represented by a distribution in which activation of each feature is low, towards the origin. In two-dimensional signal detection theory (SDT), there are three stimulus classes organised in a flat (non-hierarchical) structure. B) The same two-dimensional feature space from (A), modified to make explicit the hierarchical aspect of the higher-order state space (HOSS) model. A higher-order awareness state () nests perceptual states and .

To explore the properties of the model we simulate inference at different levels of the hierarchy for the two-class stimulus discrimination problem described in Figure 2B (where , or and is the identity matrix). We first simulate, for a variety of two-dimensional inputs (’s), the probability of saying “aware” or “seen” (). Figure 3A shows that this probability rises in a graded manner from the lower left corner of the graph (low activation of any feature) to the upper right (high activation of both features). In contrast, confidence in making a discrimination response (e.g. rightward vs. leftward) increases away from the major diagonal (Figure 3B), as the model becomes sure that the sample was generated by either a leftward or rightward tilted stimulus. As in [10], these changes in discrimination confidence may still occur in the absence of reporting “seen”.

Figure 3: Figure 3. Simulations of inference on A) awareness state and B) perceptual states , as a function of sensory input . In panel (A) the posterior probability of a report of “presence” rises from the lower left to the upper right of the grid. In panel (B) confidence in stimulus identify (e.g. left- or right-tilted Gabor) increases towards the corners of the grid. Overlaid in white is the 0.5 contour from panel (A) showing that graded changes in confidence in identity still occur on trials that have a high likelihood of being classed as “unseen” by the model.

We next simulate a proxy for prediction error at each layer in the model – in other words, how much belief change was induced by the sensory sample. We use the Kullback-Leibler (K-L) divergence as a compact summary of how far the posterior probability distribution at each level in the network differs from the prior. The K-L divergence is a measure of Bayesian surprise at each level in the network, which under predictive coding accounts is linked to neural activation at each level in a hierarchical network [5, 25]. Thus computing K-L divergence in the network provides a rough proxy for the amount of “activation” we would expect as a function of different types of decision.

At the level of perceptual states , there is substantial asymmetry in the K-L divergence expected when the model says “seen” vs. “unseen” (Figure 4A). This is due to the large belief updates invoked in the perceptual layer by samples that deviate from the origin. In contrast, when we compute K-L divergence for the awareness state (Figure 4B), the level of prediction error is symmetric across seen and unseen decisions. This is because at this level there is minimal asymmetry between inference on presence and absence. When simulating these belief updates over a range of precisions to mimic increasing stimulus-onset asynchrony in a typical backward-masking experiment, we see that the asymmetry in K-L divergence of the states increases with SOA, producing an ignition-like pattern when the stimulus is “seen” (Figure 4C).

Figure 4: Figure 4. A, B) Kullback-Leibler (K-L) divergence for A) perceptual states and B) awareness state as a function of sensory input . K-L divergence quantifies the change from prior to posterior after seeing the stimulus , and provides a metric for the magnitude of belief update at different levels of the network. The lower panels show the averaged K-L divergence for both and as a function of whether the model reports presence () or absence. The network nodes correspond to those in Figure 1A and the orange node indicates the node for which the K-L divergence is calculated. C) Behaviour of the network in a simulated masking experiment at various levels of stimulus-onset asynchrony (SOA, modeled as increasing sensory precision) in which sensory evidence was sampled from the three stimulus classes shown in Figure 1B. The lefthand panel shows that the model is more likely to report “seen” as SOA increases. The middle panel shows the K-L divergence at the level of perceptual states as a function of whether the model reports presence () or absence. The expected K-L divergence is asymmetric, with a bigger average belief update following “seen” decisions which may be a computational correlate of global ignition. The righthand panel shows the average KL divergence of awareness state as a function of whether the model reports presence () or absence. At this level the expected K-L divergence is relatively symmetric for “seen” and “unseen” decisions.

5 Empirical predictions

The model is currently situated at a computational level and remains agnostic about temporal dynamics and neural implementation444For recent work translating probabilistic graphical models into models of neuronal message passing see [26, 27].. Here I instead focus on coarser-scale predictions about the neural correlates of awareness reports in typical consciousness experiments.

First, as hinted above, an asymmetric state space for presence and absence suggests there will be greater summed prediction error in the entire network on presence decisions (as summarised by K-L divergence at each node of ). This may be a computational correlate of the global ignition responses often found to track awareness reports [15, 3].

Second, the model predicts that awareness reports (but not discrimination performance, which relies on lower-order inference on ) will depend on higher-order states. These may be instantiated in neural populations in prefrontal and parietal cortex [28]. Thus it may be possible to silence or otherwise inactivate the neural substrates of an awareness state without affecting performance – a type of blindsight [29, 30]. However, to the extent that this network is flexible in its functional contribution to higher cognition, showing both “multiple demand” characteristics [31] and mixed selectivity [32], we should also not be surprised by null results, given that single lesions may belie redundancy in its contribution to awareness [19].

Third, for the uppermost awareness state, we expect symmetry – decisions in favour of both presence and absence will lead to belief updates of similar magnitude. There has been limited focus on examining decisions about stimulus absence (as these decisions are often used as a baseline or control condition in studies of perceptual awareness). However, existing data are compatible with symmetric encoding of presence and absence at the upper level of the visual hierarchy, in primate lateral prefrontal cortex (LPFC; [33]). Merten and Nieder [34] trained monkeys to report the presence or absence of a variety of low-contrast shapes presented near to visual threshold. Neural activity tracking the decision (present or absent) was decorrelated from that involved in planning a motor response by use of a post-stimulus cue that varied from trial to trial (Figure 5). Distinct neural populations tracked the decision to report “seen” vs. “unseen”. Importantly the magnitude of activation of these populations was similar in timing and strength, suggesting a symmetric encoding of awareness in LPFC. Using fMRI, Christensen et al. also observed symmetric activation for judgments of presence and absence (compared to an intermediate visibility rating) in anterior prefrontal cortex, whereas a global ignition response was seen for presence (compared to absence) in a widespread frontoparietal/striatal network [35].

More broadly, the current framework suggests that focusing on inference about absence will be particularly fruitful for understanding the neural and computational basis of awareness reports [36, 37, 38, 34, 39].

Figure 5: Figure 5. Experimental paradigm and sample results reproduced from Merten and Nieder [34]. The left-hand panel shows the experimental paradigm used to study decisions about stimulus absence and presence after controlling for sensory and motor features of the response. The monkeys initiated each trial by grasping a lever and fixating a central fixation target. A low-contrast stimulus was then displayed for 100ms (on 50% of trials) or a blank screen was maintained (on the other 50% of trials). After a short delay, the response mappings for that trial were revealed (on some trials a present decision would require a lever release, whereas on other trials the same decision would require a lever hold). The right-hand panel shows that firing rates of neural populations in LPFC tracked abstract decisions about presence or absence before the motor mapping was known, and did so independently of stimulus properties (similar activations were seen for hits and false alarms, and for misses and correct rejections).

5.1 Role of disambiguating cues in resolving states of awareness

A higher-order awareness state is both partially observable (with respect to sensory input) and highly abstract [40]. Consider the following thought experiment in which we set up two conditions in a dark room, one in which the subject has their eyes open and one in which they have their eyes closed. Now imagine that we have arranged for neural activity in early visual areas to be identical in the two cases (the ’s are the same), and that in both cases the subject is told (for instance via an auditory cue) that there might have been a faint flash of light. Despite the visual activity being identical, the subject can be sure that they didn’t see anything when their eyes were closed compared to when they were open. In other words, whether our eyes are open or not provides disambiguating information as to the current state of awareness. Other disambiguating cues such as beliefs about the state of attention or other properties of the sensory system presumably also provide important, low-dimensional cues as to the state of awareness [41]. Computationally this may be implemented as beliefs about precision (priors on in the model in Figure 2A), where precision refers to the inverse of the noise (variability) we expect from a particular sensory channel.

One straightforward way of introducing this relationship is to allow precision itself to depend on awareness (a connection between and in Figure 1A. Such a modification implies that an awareness state may be two-dimensional, encoding the distinction between whether something has the potential to be seen (high vs. low expected precision) as well as whether something is seen (present vs. absent; [42, 43]). This aspect of the HOSS model is also in keeping with Graziano’s attention schema model of consciousness, in which awareness is equated to a model of attention [23]. However, in contrast to the attention schema, in HOSS a model of attention would provide a critical input into resolving ambiguity about whether we are aware or not (by affecting beliefs about precision), rather than determining awareness itself.

6 Relationship to other theories of consciousness

The goal of the higher-order state-space (HOSS) approach outlined here is modest - to delineate computations supporting metacognitive reports about awareness. This is reasonable given that report (or the potential for report) is the starting point for a scientific study of consciousness.

A stronger reading of the model is that conscious experience and metacognitive reports depend on shared mechanisms [44, 45]. This stronger version shares many similarities with higher-order theories of consciousness, particularly Lau’s proposal that consciousness involves “signal detection on the mind” [41]. Notably, a process of hierarchical inference may take place via passive message-passing without any strategic, cognitive access to this information e.g. in working memory [46], and is therefore compatible with higher-order representational accounts of phenomenal consciousness [47].

In HOSS, the higher-order awareness state is simple and low-dimensional. Lower-order states clearly must make a contribution to perceptual content under this arrangement – a variant of the “joint determination” view advocated by Lau and Brown [48]. However it seems to us an empirical question as to the relative granularity of higher-order and first-order representations in terms of their contribution to conscious experience, and a range of intermediate views are plausible. The more important point is that the state space is factorised to allow two separate causes of the sensory data – what it is, and whether I have seen it. In other words, becoming aware of a red, tilted object may depend on learning an abstract, factorised state of presence/absence that is not bound up with the states of being red or tilted.

HOSS also provides a new perspective on global workspace (GWS) architectures. GWS proposes that consciousness occurs when information is “globally broadcast” throughout the brain. As a result of global broadcast, cognitive and linguistic machinery have access to information about a particular stimulus or subpersonal mental state. GWS theory accounts for ignition responses on present vs. absent trials by positing that workspace neurons with long-range connections are only activated during global broadcast [49].

HOSS retains the “global” aspect of GWS, in that an awareness state is hierarchically higher with respect to the range of possible perceptual states, and therefore has a wider conceptual (and presumably temporal) purview. However, HOSS recasts ignition-like activations as asymmetric inference about stimulus presence rather than a consequence of stimulus content being “broadcast”. In any case, it is arguable whether such global broadcast is able to directly account for how a system claims to be conscious of a stimulus without positing additional machinery. Global access to the workspace would allow the system to say “there is an X”, but not endow it with the capacity to report awareness of X. This point is made concisely by Graziano [50]:

Consider asking ‘Are you aware of the apple?’ The search engine searches the internal model and finds no answer. It finds information about an apple, but no information about what ‘awareness’ is, or whether it has any of it… It cannot answer the question. It does not compute in this domain.

The state space approach outlined here is designed explicitly to compute in this domain, and therefore does not suffer from the same problem. Another critical difference between GWS and HOSS is that HOSS predicts prefrontal involvement for active decisions about stimulus absence, whereas GWS predicts that PFC remains quiescent on such trials due to a failure of the stimulus to gain access to the workspace.

Finally, to the extent that abstract awareness states need to be learnt, or constructed, they may emerge during a protracted period of development. Such development would begin with creating a perceptual generative model () before a more general property (awareness) could be abstracted from these perceptual states. This is consistent with Cleeremans’ “radical plasticity thesis” in which consciousness is underpinned by learning abstract representations of both ourselves and the world [51].

7 Research questions

I close with questions for future research motivated by the current computational sketch:

  1. How are awareness states represented in neural activity? Are presence and absence encoded symmetrically?

  2. Is a (neural) representation of awareness encoded separately from other aspects of perceptual content?

  3. How are awareness states learned?

Acknowledgements

I am grateful to the Metacognition and Theoretical Neurobiology groups at the Wellcome Centre for Human Neuroimaging and members of the University of London Institute of Philosophy for helpful discussions. I thank Matan Mazor, Chris Frith, Nicholas Shea and Karl Friston for comments on previous drafts of this manuscript. This work was supported by a Wellcome / Royal Society Sir Henry Dale Fellowship (206648/Z/17/Z).

References

  • [1] Chris Frith, Richard Perry, and Eric Lumer. The neural correlates of conscious experience: an experimental framework. Trends in Cognitive Sciences, 3(3):105–114, 1999.
  • [2] Bernard J Baars. A Cognitive Theory of Consciousness. Cambridge University Press, 1993.
  • [3] Stanislas Dehaene and Jean-Pierre Changeux. Experimental and Theoretical Approaches to Conscious Processing. Neuron, 70(2):200–227, 2011.
  • [4] S M Fleming and Nathaniel D Daw. Self-evaluation of decision-making: A general Bayesian framework for metacognitive computation. Psychological Review, 124(1):91–114, 2017.
  • [5] K. Friston. A theory of cortical responses. Philos.Trans.R.Soc.Lond B Biol.Sci., 360(1456):815–836, 2005.
  • [6] H von Helmholtz. Treatise on Physiological Optics. 1860.
  • [7] Jakob Hohwy. The Predictive Mind. Oxford University Press, 2013.
  • [8] Daniel Kersten, Pascal Mamassian, and Alan Yuille. Object perception as Bayesian inference. Annual Review of Psychology, 55:271–304, 2004.
  • [9] D.M. Green and J.A. Swets. Signal detection theory and psychophysics. Wiley, New York, 1966.
  • [10] J-R King and S Dehaene. A model of subjective report and objective discrimination as categorical decisions in a vast representational space. Philos.Trans.R.Soc.Lond B Biol.Sci., 369(1641):20130204, 2014.
  • [11] Claire Sergent and Stanislas Dehaene. Is consciousness a gradual phenomenon? evidence for an all-or-none bifurcation during the attentional blink. Psychological Science, 15(11):720–728, 2004.
  • [12] Thomas Zoëga Ramsøy and Morten Overgaard. Introspection and subliminal perception. Phenomenology and the Cognitive Sciences, 3(1):1–23, 2004.
  • [13] Chai-Youn Kim and Randolph Blake. Psychophysical magic: rendering the visible ’invisible’. Trends in Cognitive Sciences, 9(8):381–388, 2005.
  • [14] Stanislas Dehaene, Lionel Naccache, Laurent Cohen, Denis Le Bihan, Jean-François Mangin, Jean-Baptiste Poline, and Denis Rivière. Cerebral mechanisms of word masking and unconscious repetition priming. Nature Neuroscience, 4(7):752, 2001.
  • [15] A. Del Cul, S. Baillet, and S. Dehaene. Brain dynamics underlying the nonlinear threshold for access to consciousness. PLoS.Biol., 5(10):e260, 2007.
  • [16] Jaan Aru, Talis Bachmann, Wolf Singer, and Lucia Melloni. Distilling the neural correlates of consciousness. Neuroscience and Biobehavioral Reviews, 36(2):737–746, 2012.
  • [17] H.C. Lau and R.E. Passingham. Relative blindsight in normal observers and the neural correlate of visual consciousness. Proc.Natl.Acad.Sci.U.S.A, 103(49):18763–18768, 2006.
  • [18] Naotsugu Tsuchiya, Melanie Wilke, Stefan Frässle, and Victor A. F. Lamme. No-Report Paradigms: Extracting the True Neural Correlates of Consciousness. Trends in Cognitive Sciences, 19(12):757–770, 2015.
  • [19] Matthias Michel and Jorge Morales. Minority Reports: Consciousness and the Prefrontal Cortex. Mind and Language, 2019.
  • [20] Ned Block. On a Confusion About a Function of Consciousness. Brain and Behavioral Sciences, 18(2):227–247, 1995.
  • [21] Ned Block. Perceptual consciousness overflows cognitive access. Trends in Cognitive Sciences, 15(12):567–575, 2011.
  • [22] Daniel C. Dennett. Consciousness Explained. Penguin UK, 1993.
  • [23] Michael S. A. Graziano. Consciousness and the Social Brain. Oxford University Press, 2013.
  • [24] Judea Pearl. Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference. Morgan Kaufmann Publishers, 1988.
  • [25] Christopher Summerfield and Floris P de Lange. Expectation in perceptual decision making: neural and computational mechanisms. Nature Reviews Neuroscience, 15(11):745–756, 2014.
  • [26] Dileep George, Wolfgang Lehrach, Ken Kansky, Miguel Lázaro-Gredilla, Christopher Laan, Bhaskara Marthi, Xinghua Lou, Zhaoshi Meng, Yi Liu, Huayan Wang, Alex Lavin, and D. Scott Phoenix. A generative vision model that trains with high data efficiency and breaks text-based CAPTCHAs. Science (New York, N.Y.), 358(6368), 2017.
  • [27] Karl J. Friston, Thomas Parr, and Bert de Vries. The graphical brain: Belief propagation and active inference. Network Neuroscience, 1(4):381–414, 2017.
  • [28] Hakwan Lau and David Rosenthal. Empirical support for higher-order theories of conscious awareness. Trends in Cognitive Sciences, 15(8):365–373, 2011.
  • [29] Antoine Del Cul, Stanislas Dehaene, P Reyes, E Bravo, and A Slachevsky. Causal role of prefrontal cortex in the threshold for access to consciousness. Brain, 132(9):2531–2540, 2009.
  • [30] Lawrence Weiskrantz. Consciousness lost and found: A neuropsychological exploration. OUP Oxford, 1999.
  • [31] John Duncan. The multiple-demand (md) system of the primate brain: mental programs for intelligent behaviour. Trends in cognitive sciences, 14(4):172–179, 2010.
  • [32] Valerio Mante, David Sussillo, Krishna V Shenoy, and William T Newsome. Context-dependent computation by recurrent dynamics in prefrontal cortex. Nature, 503(7474):78, 2013.
  • [33] Theofanis I. Panagiotaropoulos, Gustavo Deco, Vishal Kapoor, and Nikos K. Logothetis. Neuronal discharges and gamma oscillations explicitly reflect visual consciousness in the lateral prefrontal cortex. Neuron, 74(5):924–935, 2012.
  • [34] Katharina Merten and Andreas Nieder. Active encoding of decisions about stimulus absence in primate prefrontal cortex neurons. Proceedings of the National Academy of Sciences, 2012.
  • [35] Mark S Christensen, Thomas Z Ramsøy, Torben E Lund, Kristoffer H Madsen, and James B Rowe. An fmri study of the neural correlates of graded visual perception. Neuroimage, 31(4):1711–1725, 2006.
  • [36] Anna Farennikova. Seeing Absence. Philosophical Studies, 166(3):429–454, 2013.
  • [37] Ryota Kanai, Vincent Walsh, and Chia-huei Tseng. Subjective discriminability of invisibility: a framework for distinguishing perceptual and attentional failures of awareness. Consciousness and Cognition, 19(4):1045–1057, 2010.
  • [38] Jean-Rémy Martin and Jérôme Dokic. Seeing Absence or Absence of Seeing? Thought: A Journal of Philosophy, 2(1):117–125, 2013.
  • [39] Katharina Merten and Andreas Nieder. Comparison of abstract decision encoding in the monkey prefrontal cortex, the presupplementary, and cingulate motor areas. Journal of Neurophysiology, 110(1):19–32, 2013.
  • [40] Nicolas W. Schuck, Robert Wilson, and Yael Niv. A State Representation for Reinforcement Learning and Decision-Making in the Orbitofrontal Cortex. bioRxiv, 2018.
  • [41] Hakwan C. Lau. A higher order Bayesian decision theory of consciousness. Progress in Brain Research, 168:35–48, 2008.
  • [42] Thomas Metzinger. How does the brain encode epistemic reliability? Perceptual presence, phenomenal transparency, and counterfactual richness. Cognitive Neuroscience, 5(2):122–124, 2014.
  • [43] Jakub Limanowski and Karl Friston. ’seeing the dark’: Grounding phenomenal transparency and opacity in precision estimation for active inference. Frontiers in Psychology, 9, 2018.
  • [44] E. Shaver, B Maniscalco, and HC Lau. Awareness as confidence. Anthropology and Philosophy, 9:58–65, 2008.
  • [45] Richard Brown, Hakwan Lau, and Joseph LeDoux. The Misunderstood Higher-Order Approach to Consciousness. preprint, PsyArXiv, 2019.
  • [46] Peter Carruthers. Block’s Overflow Argument. Pacific Philosophical Quarterly, 98(S1):65–70, 2017.
  • [47] Richard Brown. The HOROR Theory of Phenomenal Consciousness. Philosophical Studies, 172(7):1783–1794, 2015.
  • [48] Hakwan Lau and Richard Brown. The Emperor’s New Phenomenology? The Empirical Case for Conscious Experience Without First-Order Representations. In Adam Pautz and Daniel Stoljar, editors, Blockheads! Essays on Ned Block’s Philosophy of Mind and Consciousness. MIT Press, 2019.
  • [49] S Dehaene, M Kerszberg, and J P Changeux. A neuronal model of a global workspace in effortful cognitive tasks. Proceedings of the National Academy of Sciences of the United States of America, 95(24):14529–14534, 1998.
  • [50] Michael S. A. Graziano. Consciousness engineered. Journal of Consciousness Studies, 23(11-12):98–115, 2016.
  • [51] Axel Cleeremans. The Radical Plasticity Thesis: How the Brain Learns to be Conscious. Frontiers in Psychology, 2:1–12, 2011.
Comments 0
Request Comment
You are adding the first comment!
How to quickly get a good reply:
  • Give credit where it’s due by listing out the positive aspects of a paper before getting into which changes should be made.
  • Be specific in your critique, and provide supporting evidence with appropriate references to substantiate general statements.
  • Your comment should inspire ideas to flow and help the author improves the paper.

The better we are at sharing our knowledge with each other, the faster we move forward.
""
The feedback must be of minimum 40 characters and the title a minimum of 5 characters
   
Add comment
Cancel
Loading ...
371365
This is a comment super asjknd jkasnjk adsnkj
Upvote
Downvote
""
The feedback must be of minumum 40 characters
The feedback must be of minumum 40 characters
Submit
Cancel

You are asking your first question!
How to quickly get a good answer:
  • Keep your question short and to the point
  • Check for grammar or spelling errors.
  • Phrase it like a question
Test
Test description