Deep Gated Hebbian Predictive Coding Accounts for Emergence of Complex Neural Response Properties Along the Visual Cortical Hierarchy

An integrative, multiscale view on neural theories of consciousness

How is conscious experience related to material brain processes? A variety of theories aiming to answer this age-old question have emerged from the recent surge in consciousness research, and some are now hotly debated. Although most researchers have so far focused on the development and validation of their preferred theory in relative isolation, this article, written by a group of scientists representing different theories, takes an alternative approach. Noting that various theories often try to explain different aspects or mechanistic levels of consciousness, we argue that the theories do not necessarily contradict each other. Instead, several of them may converge on fundamental neuronal mechanisms and be partly compatible and complementary, so that multiple theories can simultaneously contribute to our understanding. Here, we consider unifying, integration-oriented approaches that have so far been largely neglected, seeking to combine valuable elements from various theories.

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.

Spontaneously emerging internal models of visual sequences combine abstract and event-specific information in the prefrontal cortex

When exposed to sensory sequences, do macaque monkeys spontaneously form abstract internal models that generalize to novel experiences? Here, we show that neuronal populations in macaque ventrolateral prefrontal cortex jointly encode visual sequences by separate codes for the specific pictures presented and for their abstract sequential structure. We recorded prefrontal neurons while macaque monkeys passively viewed visual sequences and sequence mismatches in the local-global paradigm. Even without any overt task or response requirements, prefrontal populations spontaneously form representations of sequence structure, serial order, and image identity within distinct but superimposed neuronal subspaces. Representations of sequence structure rapidly update following single exposure to a mismatch sequence, while distinct populations represent mismatches for sequences of different complexity. Finally, those representations generalize across sequences following the same repetition structure but comprising different images. These results suggest that prefrontal populations spontaneously encode rich internal models of visual sequences reflecting both content-specific and abstract information.

Assessing the commensurability of theories of consciousness: On the usefulness of common denominators in differentiating, integrating and testing hypotheses

How deep is the current diversity in the panoply of theories to define consciousness, and to what extent do these theories share common denominators? Here we first examine to what extent different theories are commensurable (or comparable) along particular dimensions. We posit logical (and, when applicable, empirical) commensurability as a necessary condition for identifying common denominators among different theories. By consequence, dimensions for inclusion in a set of logically and empirically commensurable theories of consciousness can be proposed. Next, we compare a limited subset of neuroscience-based theories in terms of commensurability. This analysis does not yield a denominator that might serve to define a minimally unifying model of consciousness. Theories that seem to be akin by one denominator can be remote by another. We suggest a methodology of comparing different theories via multiple probing questions, allowing to discern overall (dis)similarities between theories. Despite very different background definitions of consciousness, we conclude that, if attention is paid to the search for a common methological approach to brain-consciousness relationships, it should be possible in principle to overcome the current Babylonian confusion of tongues and eventually integrate and merge different theories.

How deep is the brain? The shallow brain hypothesis

Deep learning and predictive coding architectures commonly assume that inference in neural networks is hierarchical. However, largely neglected in deep learning and predictive coding architectures is the neurobiological evidence that all hierarchical cortical areas, higher or lower, project to and receive signals directly from subcortical areas. Given these neuroanatomical facts, today's dominance of cortico-centric, hierarchical architectures in deep learning and predictive coding networks is highly questionable; such architectures are likely to be missing essential computational principles the brain uses. In this Perspective, we present the shallow brain hypothesis: hierarchical cortical processing is integrated with a massively parallel process to which subcortical areas substantially contribute. This shallow architecture exploits the computational capacity of cortical microcircuits and thalamo-cortical loops that are not included in typical hierarchical deep learning and predictive coding networks. We argue that the shallow brain architecture provides several critical benefits over deep hierarchical structures and a more complete depiction of how mammalian brains achieve fast and flexible computational capabilities.

How 'visual' is the visual cortex? The interactions between the visual cortex and other sensory, motivational and motor systems as enabling factors for visual perception

The definition of the visual cortex is primarily based on the evidence that lesions of this area impair visual perception. However, this does not exclude that the visual cortex may process more information than of retinal origin alone, or that other brain structures contribute to vision. Indeed, research across the past decades has shown that non-visual information, such as neural activity related to reward expectation and value, locomotion, working memory and other sensory modalities, can modulate primary visual cortical responses to retinal inputs. Nevertheless, the function of this non-visual information is poorly understood. Here we review recent evidence, coming primarily from studies in rodents, arguing that non-visual and motor effects in visual cortex play a role in visual processing itself, for instance disentangling direct auditory effects on visual cortex from effects of sound-evoked orofacial movement. These findings are placed in a broader framework casting vision in terms of predictive processing under control of frontal, reward- and motor-related systems. In contrast to the prevalent notion that vision is exclusively constructed by the visual cortical system, we propose that visual percepts are generated by a larger network-the extended visual system-spanning other sensory cortices, supramodal areas and frontal systems. This article is part of the theme issue 'Decision and control processes in multisensory perception'.

Local minimization of prediction errors drives learning of invariant object representations in a generative network model of visual perception

The ventral visual processing hierarchy of the cortex needs to fulfill at least two key functions: perceived objects must be mapped to high-level representations invariantly of the precise viewing conditions, and a generative model must be learned that allows, for instance, to fill in occluded information guided by visual experience. Here, we show how a multilayered predictive coding network can learn to recognize objects from the bottom up and to generate specific representations via a top-down pathway through a single learning rule: the local minimization of prediction errors. Trained on sequences of continuously transformed objects, neurons in the highest network area become tuned to object identity invariant of precise position, comparable to inferotemporal neurons in macaques. Drawing on this, the dynamic properties of invariant object representations reproduce experimentally observed hierarchies of timescales from low to high levels of the ventral processing stream. The predicted faster decorrelation of error-neuron activity compared to representation neurons is of relevance for the experimental search for neural correlates of prediction errors. Lastly, the generative capacity of the network is confirmed by reconstructing specific object images, robust to partial occlusion of the inputs. By learning invariance from temporal continuity within a generative model, the approach generalizes the predictive coding framework to dynamic inputs in a more biologically plausible way than self-supervised networks with non-local error-backpropagation. This was achieved simply by shifting the training paradigm to dynamic inputs, with little change in architecture and learning rule from static input-reconstructing Hebbian predictive coding networks.

What is neurorepresentationalism? From neural activity and predictive processing to multi-level representations and consciousness

This review provides an update on Neurorepresentationalism, a theoretical framework that defines conscious experience as multimodal, situational survey and explains its neural basis from brain systems constructing best-guess representations of sensations originating in our environment and body [1]. It posits that conscious experience is characterized by five essential hallmarks: (i) multimodal richness, (ii) situatedness and immersion, (iii) unity and integration, (iv) dynamics and stability, and (v) intentionality. Consciousness is furthermore proposed to have a biological function, framed by the contrast between reflexes and habits (not requiring consciousness) versus goal-directed, planned behavior (requiring multimodal, situational survey). Conscious experience is therefore understood as a sensorily rich, spatially encompassing representation of body and environment, while we nevertheless have the impression of experiencing external reality directly. Contributions to understanding neural mechanisms underlying consciousness are derived from models for predictive processing, which are trained in an unsupervised manner, do not necessarily require overt action, and have been extended to deep neural networks. Even with predictive processing in place, however, the question remains why this type of neural network activity would give rise to phenomenal experience. Here, I propose to tackle the Hard Problem with the concept of multi-level representations which emergently give rise to multimodal, spatially wide superinferences corresponding to phenomenal experiences. Finally, Neurorepresentationalism is compared to other neural theories of consciousness, and its implications for defining indicators of consciousness in animals, artificial intelligence devices and immobile or unresponsive patients with disorders of consciousness are discussed.

Linking Brain Structure, Activity, and Cognitive Function through Computation

Understanding the human brain is a "Grand Challenge" for 21st century research. Computational approaches enable large and complex datasets to be addressed efficiently, supported by artificial neural networks, modeling and simulation. Dynamic generative multiscale models, which enable the investigation of causation across scales and are guided by principles and theories of brain function, are instrumental for linking brain structure and function. An example of a resource enabling such an integrated approach to neuroscientific discovery is the BigBrain, which spatially anchors tissue models and data across different scales and ensures that multiscale models are supported by the data, making the bridge to both basic neuroscience and medicine. Research at the intersection of neuroscience, computing and robotics has the potential to advance neuro-inspired technologies by taking advantage of a growing body of insights into perception, plasticity and learning. To render data, tools and methods, theories, basic principles and concepts interoperable, the Human Brain Project (HBP) has launched EBRAINS, a digital neuroscience research infrastructure, which brings together a transdisciplinary community of researchers united by the quest to understand the brain, with fascinating insights and perspectives for societal benefits.