Several approaches to implementing symbol-like representations in neurally plausible models have been proposed. These approaches include binding through synchrony, “mesh” binding, and conjunctive binding. Recent theoretical work has suggested that most of these methods will not scale well, that is, that they cannot encode structured representations using any of the tens of thousands of terms in the adult lexicon without making implausible resource assumptions. Here, we empirically demonstrate that the biologically plausible structured representations employed in the Semantic (...) Pointer Architecture approach to modeling cognition do scale appropriately. Specifically, we construct a spiking neural network of about 2.5 million neurons that employs semantic pointers to successfully encode and decode the main lexical relations in WordNet, which has over 100,000 terms. In addition, we show that the same representations can be employed to construct recursively structured sentences consisting of arbitrary WordNet concepts, while preserving the original lexical structure. We argue that these results suggest that semantic pointers are uniquely well-suited to providing a biologically plausible account of the structured representations that underwrite human cognition. (shrink)

We present techniques for integrating low‐level neurobiological constraints into high‐level, functional cognitive models. In particular, we use these techniques to construct a model of eyeblink conditioning in the cerebellum based on temporal representations in the recurrent Granule‐Golgi microcircuit.

A fundamental claim associated with parallel distributed processing theories of cognition is that knowledge is coded in a distributed manner in mind and brain. This approach rejects the claim that knowledge is coded in a localist fashion, with words, objects, and simple concepts, that is, coded with their own dedicated representations. One of the putative advantages of this approach is that the theories are biologically plausible. Indeed, advocates of the PDP approach often highlight the close parallels between distributed (...) representations learned in connectionist models and neural coding in brain and often dismiss localist theories as biologically implausible. The author reviews a range a data that strongly challenge this claim and shows that localist models provide a better account of single-cell recording studies. The author also contrast local and alternative distributed coding schemes and argues that common rejection of grandmother cell theories in neuroscience is due to a misunderstanding about how localist models behave. The author concludes that the localist representations embedded in theories of perception and cognition are consistent with neuroscience; biology only calls into question the distributed representations often learned in PDP models. (shrink)

This paper describes how Tononi’s information integration theory of consciousness was used to make detailed predictions about the distribution of phenomenal states in a spiking neural network. This network had approximately 18,000 neurons and 700,000 connections and it used models of emotion and imagination to control the eye movements of a virtual robot and avoid ‘negative’ stimuli. The first stage in the analysis was the development of a formal definition of Tononi’s theory of consciousness. The network was then (...) analysed for information integration and detailed predictions were made about the distribution of consciousness for each time step of recorded activity. This work demonstrates how an artificial system can be analysed for consciousness using a particular theory and in the future this approach could be used to make predictions about the phenomenal states associated with biological systems. (shrink)

Deep neural networks (DNNs) have had extraordinary successes in classifying photographic images of objects and are often described as the best models of biological vision. This conclusion is largely based on three sets of findings: (1) DNNs are more accurate than any other model in classifying images taken from various datasets, (2) DNNs do the best job in predicting the pattern of human errors in classifying objects taken from various behavioral datasets, and (3) DNNs do the best job (...) in predicting brain signals in response to images taken from various brain datasets (e.g., single cell responses or fMRI data). However, these behavioral and brain datasets do not test hypotheses regarding what features are contributing to good predictions and we show that the predictions may be mediated by DNNs that share little overlap with biological vision. More problematically, we show that DNNs account for almost no results from psychological research. This contradicts the common claim that DNNs are good, let alone the best, models of human object recognition. We argue that theorists interested in developing biologically plausible models of human vision need to direct their attention to explaining psychological findings. More generally, theorists need to build models that explain the results of experiments that manipulate independent variables designed to test hypotheses rather than compete on making the best predictions. We conclude by briefly summarizing various promising modeling approaches that focus on psychological data. (shrink)

A dynamic threshold, which controls the nature and course of learning, is a pivotal concept in Page's general localist framework. This commentary addresses various issues surrounding biologically plausible implementations for such thresholds. Relevant previous research is noted and the particular difficulties relating to the creation of so-called instance representations are highlighted. It is stressed that these issues also apply to distributed models.

Quilty-Dunn et al. argue that deep convolutional neural networks (DCNNs) optimized for image classification exemplify structural disanalogies to human vision. A different kind of artificial vision – found in reinforcement-learning agents navigating artificial three-dimensional environments – can be expected to be more human-like. Recent work suggests that language-like representations substantially improves these agents’ performance, lending some indirect support to the language-of-thought hypothesis (LoTH).

Chaos in nervous system is a fascinating but controversial field of investigation. To approach the role of chaos in the real brain, we theoretically and numerically investigate the occurrence of chaos inartificial neural networks. Most of the time, recurrent networks (with feedbacks) are fully connected. This architecture being not biologically plausible, the occurrence of chaos is studied here for a randomly diluted architecture. By normalizing the variance of synaptic weights, we produce a bifurcation parameter, dependent (...) on this variance and on the slope of the transfer function, that allows a sustained activity and the occurrence of chaos when reaching a critical value. Even for weak connectivity and small size, we find numerical results in accordance with the theoretical ones previously established for fully connected infinite sized networks. The route towards chaos is numerically checked to be a quasi-periodic one, whatever the type of the first bifurcation is. Our results suggest that such high-dimensional networks behave like low-dimensional dynamical systems. (shrink)

Further tests of Amit's model are indicated. One strategy is to use the apparent coding sparseness of the model to make predictions about coding sparseness in Miyashita's network. A second approach is to use memory overload to induce false positive responses in modules and biological systems. In closing, the importance of temporal coding and timing requirements in developing biologically plausible attractor networks is mentioned.

The ability to improve in speed and accuracy as a result of repeating some task is an important hallmark of intelligent biological systems. Although gradual behavioral improvements from practice have been modeled in spiking neural networks, few such models have attempted to explain cognitive development of a task as complex as addition. In this work, we model the progression from a counting-based strategy for addition to a recall-based strategy. The model consists of two networks working in (...) parallel: a slower basal ganglia loop and a faster cortical network. The slow network methodically computes the count from one digit given another, corresponding to the addition of two digits, whereas the fast network gradually “memorizes” the output from the slow network. The faster network eventually learns how to add the same digits that initially drove the behavior of the slower network. Performance of this model is demonstrated by simulating a fully spiking neural network that includes basal ganglia, thalamus, and various cortical areas. Consequently, the model incorporates various neuroanatomical data, in terms of brain areas used for calculation and makes psychologically testable predictions related to frequency of rehearsal. Furthermore, the model replicates developmental progression through addition strategies in terms of reaction times and accuracy, and naturally explains observed symptoms of dyscalculia. (shrink)

Inductive reasoning is a fundamental and complex aspect of human intelligence. In particular, how do subjects, given a set of particular examples, generate general descriptions of the rules governing that set? We present a biologically plausible method for accomplishing this task and implement it in a spiking neuron model. We demonstrate the success of this model by applying it to the problem domain of Raven's Progressive Matrices, a widely used tool in the field of intelligence testing. The (...) model is able to generate the rules necessary to correctly solve Raven's items, as well as recreate many of the experimental effects observed in human subjects. (shrink)

Artificial neural networks (ANNs) are new mathematical techniques which can be used for modelling real neural networks, but also for data categorisation and inference tasks in any empirical science. This means that they have a twofold interest for the philosopher. First, ANN theory could help us to understand the nature of mental phenomena such as perceiving, thinking, remembering, inferring, knowing, wanting and acting. Second, because ANNs are such powerful instruments for data classification and inference, their use (...) also leads us into the problems of induction and probability. Ever since David Hume expressed his famous doubts about induction, the principles of scientific inference have been a central concern for philosophers. (shrink)

Although great progress in neuroanatomy and physiology has occurred lately, we still cannot go directly to those levels to discover the neural mechanisms of higher cognition and consciousness. But we can use neurocomputational methods based on these details to push this project forward. Here we describe vector subtraction as an operation that computes sequential paths through high-dimensional vector spaces. Vector-space interpretations of network activity patterns are a fruitful resource in recent computational neuroscience. Vector subtraction also appears to be implemented (...) neurally in primate frontal eye field activity, which computes dimensions of saccadic eye movements. We use this apparent neural implementation as a model and construct from it a general neurocomputational account of an important type of sequential cognitive and conscious process. We defend the biological plausibility of all components of the general model and show that it yields testable anatomical and physiological predictions. We close by suggesting some interesting consequences for consciousness if our model characterizes correctly the neural mechanisms producing a common type of episode in our conscious streams. (shrink)

Travelling waves crossing the nervous networks at mesoscopic/macroscopic scales have been correlated with different brain functions, from long-term memory to visual stimuli. Here we investigate a feasible relationship between wave generation/propagation in recurrent nervous networks and a physical/chemical model, namely the Belousov–Zhabotinsky reaction. Since BZ’s nonlinear, chaotic chemical process generates concentric/intersecting waves that closely resemble the diffusive nonlinear/chaotic oscillatory patterns crossing the nervous tissue, we aimed to investigate whether wave propagation of brain oscillations could be described in terms (...) of BZ features. We compared experimentally detected oscillations during the spontaneous activity of the brain with BZ-like concentric waves simulated by a recently introduced artificial network. The observed overlap and agreement between simulated and measured oscillatory patterns suggests that changes in cortical areas’ neural activity might be described in terms of a recognizable diffusion pattern. We describe biological plausibility, benefits and limits of our approach and discuss the relationship among BZ-like networks, Pandemonium-like architectures and the spontaneous activity of the brain. ER -. (shrink)

Endowing artificial systems with explanatory capacities about the reasons guiding their decisions, represents a crucial challenge and research objective in the current fields of Artificial Intelligence (AI) and Computational Cognitive Science [Langley et al., 2017]. Current mainstream AI systems, in fact, despite the enormous progresses reached in specific tasks, mostly fail to provide a transparent account of the reasons determining their behavior (both in cases of a successful or unsuccessful output). This is due to the fact that the classical problem (...) of opacity in artificial neural networks (ANNs) explodes with the adoption of current Deep Learning techniques [LeCun, Bengio, Hinton, 2015]. In this paper we argue that the explanatory deficit of such techniques represents an important problem, that limits their adoption in the cognitive modelling and computational cognitive science arena. In particular we will show how the current attempts of providing explanations of the deep nets behaviour (see e.g. [Ritter et al. 2017] are not satisfactory. As a possibile way out to this problem, we present two different research strategies. The first strategy aims at dealing with the opacity problem by providing a more abstract interpretation of neural mechanisms and representations. This approach is adopted, for example, by the biologically inspired SPAUN architecture [Eliasmith et al., 2012] and by other proposals suggesting, for example, the interpretation of neural networks in terms of the Conceptual Spaces framework [Gärdenfors 2000, Lieto, Chella and Frixione, 2017]. All such proposals presuppose that the neural level of representation can be considered somehow irrelevant for attacking the problem of explanation [Lieto, Lebiere and Oltramari, 2017]. In our opinion, pursuing this research direction can still preserve the use of deep learning techniques in artificial cognitive models provided that novel and additional results in terms of “transparency” are obtained. The second strategy is somehow at odds with respect to the previous one and tries to address the explanatory issue by avoiding to directly solve the “opacity” problem. In this case, the idea is that one of resorting to pre-compiled plausible explanatory models of the word used in combination with deep-nets (see e.g. [Augello et al. 2017]). We argue that this research agenda, even if does not directly fits the explanatory needs of Computational Cognitive Science, can still be useful to provide results in the area of applied AI aiming at shedding light on the models of interaction between low level and high level tasks (e.g. between perceptual categorization and explanantion) in artificial systems. (shrink)

This Neurocomputing special issue is based on selected, expanded and significantly revised versions of papers presented at the Second International Conference on Brain Inspired Cognitive Systems (BICS 2006) held at Lesvos, Greece, from 10 to 14 October 2006. The aim of BICS 2006, which followed the very successful first BICS 2004 held at Stirling, Scotland, was to bring together leading scientists and engineers who use analytic, syntactic and computational methods both to understand the prodigious processing properties of biological systems and, (...) specifically, of the brain, and to exploit such knowledge to advance computational methods towards ever higher levels of cognitive competence. The biennial BICS Conference Series (with BICS 2008 recently held in Sao Luis, Brazil, 24–27 June, and BICS 2010 due to be held in Madrid, Spain) aims to become a major point of contact for research scientists, engineers and practitioners throughout the world in the fields of cognitive and computational systems inspired by the brain and biology. The first paper in this special issue is by Carnell who presents an analysis of the use of Hebbian and Anti-Hebbian spike timedependent plasticity (STDP) learning functions within the context of recurrent spiking neural networks. He shows that under specific conditions Hebbian and Anti-Hebbian learning can be considered approximately equivalent. Finally, the author demonstrates that such a network habituates to a given stimulus and is capable of detecting subtle variations in the structure of the stimuli itself. Hodge, O’Keefe and Austin present a binary neural shape matcher using Johnson counters and chain codes. They show that images may be matched as whole images or using shape matching. Finally, they demonstrate shape matching using a binary associative-memory neural network to index and match chain codes where the chain code elements are represented by Johnson codes. (shrink)

Human decision making shows systematic simplifications and deviations from the tenets of rationality (‘heuristics’) that may lead to suboptimal decisional outcomes (‘cognitive biases’). There are currently three prevailing theoretical perspectives on the origin of heuristics and cognitive biases: a cognitive-psychological, an ecological and an evolutionary perspective. However, these perspectives are mainly descriptive and none of them provides an overall explanatory framework for the underlying mechanisms of cognitive biases. To enhance our understanding of cognitive heuristics and biases we propose a (...) class='Hi'>neural network framework for cognitive biases, which explains why our brain systematically tends to default to heuristic (‘Type 1’) decision making. We argue that many cognitive biases arise from intrinsic brain mechanisms that are fundamental for the working of biological neural networks. In order to substantiate our viewpoint, we discern and explain four basic neural network principles: (1) Association, (2) Compatibility (3) Retainment, and (4) Focus. These principles are inherent to (all) neural networks which were originally optimized to perform concrete biological, perceptual, and motor functions. They form the basis for our inclinations to associate and combine (unrelated) information, to prioritize information that is compatible with our present state (such as knowledge, opinions and expectations), to retain given information that sometimes could better be ignored, and to focus on dominant information while ignoring relevant information that is not directly activated. The supposed mechanisms are complementary and not mutually exclusive. For different cognitive biases they may all contribute in varying degrees to distortion of information. The present viewpoint not only complements the earlier three viewpoints, but also provides a unifying and binding framework for many cognitive bias phenomena. (shrink)

Head direction (HD) cells, abundant in the rat postsubiculum and anterior thalamic nuclei, fire maximally when the rat’s head is facing a particular direction. The activity of a population of these cells forms a distributed representation of the animal’s current heading. We describe a neural network model that creates a stable, distributed representation of head direction and updates that representation in response to angular velocity information. In contrast to earlier models, our model of the head direction system accurately tracks (...) a series of actual rat head rotations, and, using biologically plausible neurons, it fits the single-cell tuning curves of real HD cells recorded from rats executing those same rotations. The model makes neurophysiological predictions that can be tested using current technologies. (shrink)

Numerous species use different forms of communication in order to successfully interact in their respective environment. This article seeks to elucidate limitations of the classical conduit metaphor by investigating communication from the perspectives of biology and artificial neural networks. First, communication is a biological natural phenomenon, found to be fruitfully grounded in an organism’s embodied structures and memory system, where specific abilities are tied to procedural, semantic, and episodic long-term memory as well as to working memory. Second, the (...) account explicates differences between non-verbal and verbal communication and shows how artificial neural networks can communicate by means of ontologically non-committal modelling. This approach enables new perspectives of communication to emerge regarding both sender and receiver. It is further shown that communication features gradient properties that are plausibly divided into a reflexive and a reflective form, parallel to knowledge and reflection. (shrink)

Eye movement desensitization and reprocessing therapy is a well-established therapeutic method to treat post-traumatic stress disorder. However, how EMDR exerts its therapeutic action has been studied in many types of research but still needs to be completely understood. This is in part due to limited knowledge of the neurobiological mechanisms underlying EMDR, and in part to our incomplete understanding of PTSD. In order to model PTSD, we used a biologically inspired computational model based on firing rate units, encompassing the (...) cortex, hippocampus, and amygdala. Through the modulation of its parameters, we fitted real data from patients treated with EMDR or classical exposure therapy. This allowed us to gain insights into PTSD mechanisms and to investigate how EMDR achieves trauma remission. (shrink)

On the basis of recent advancements in both neuroscience and archaeology, we propose a plausible biocultural mechanism at the basis of cultural evolution. The proposed mechanism, which relies on the notions of cultural exaptation and cultural neural reuse, may account for the asynchronous, discontinuous, and patchy emergence of innovations around the globe. Cultural exaptation refers to the reuse of previously devised cultural features for new purposes. Cultural neural reuse refers to cases in which exposure to cultural practices (...) induces the formation, activation, and stabilization of new functional and/or structural brain networks during the individual lifespan. The invention of writing is interpreted as a case of cultural exaptation of previous devices to record information, in use since at least the Early Later Stone Age and the beginning of the Upper Paleolithic. The measurable changes in brain structure and functioning caused by learning to read are proposed as an exemplar case of cultural neural reuse. It is argued that repeated cycles of cultural exaptation, development of appropriate strategies of cultural transmission, and ensuing cultural neural reuse represent the fundamental mechanism that has regulated the cultural evolution of our lineage. A general predictive model of when and under which circumstances the proposed mechanism should be expected to occur is proposed, and the relationship of our mechanism with gene-culture coevolutionary models is discussed. (shrink)

Dogs display vast phenotypic diversity, including differences in height, skull shape, tail, etc. Yet, humans are almost always able to quickly recognize a dog, despite no single feature or group of features are critical to distinguish dogs from other objects/animals. In search of the mental activities leading human individuals to state “I see a dog”, we hypothesize that the brain might extract meaningful information from the environment using Ramsey sentences-like procedures. To turn the proposition “I see a dog” in a (...) Ramsey sentence, the term dog must be replaced by a long and complex assertion consisting only of observational terms, existential quantifiers and operational rules. The Ramsey sentence for “I see a dog” sounds: “There is at least an entity called dog which satisfies the following conditions: it is an animal, it has four legs, …, etc, …, and is something that I have in my sight”. We discuss the biological plausibility and the putative neural correlates of a Ramsey-like mechanism in the central nervous system. We accomplish a brain-inspired, theoretical neural architecture consisting of a parallel network that requires virtually no memory, is devoid of probabilistic choices and can analyze huge but finite amounts of unique visual details, combining them into a single conceptual output. In sum, Ramsey sentence stands for a versatile tool that can be used not just as a methodological device to cope with biophysical affairs, but also for a model to describe the real functioning of cognitive operations such as sensation and perception. (shrink)

As an imitation of the biological nervous systems, neural networks, which have been characterized as powerful learning tools, are employed in a wide range of applications, such as control of complex nonlinear systems, optimization, system identification, and patterns recognition. This article aims to bring a brief review of the state-of-the-art NNs for the complex nonlinear systems by summarizing recent progress of NNs in both theory and practical applications. Specifically, this survey also reviews a number of NN based robot (...) control algorithms, including NN based manipulator control, NN based human-robot interaction, and NN based cognitive control. (shrink)

For millions of years, biological creatures have dealt with the world without being able to see it; however, the change in the atmospheric condition during the Cambrian period and the subsequent increase of light, triggered the sudden evolution of vision and the consequent evolutionary benefits. Nevertheless, how from simple organisms to more complex animals have been able to generate meaning from the light who fell in their eyes and successfully engage the visual world remains unknown. As shown by many psychophysical (...) experiments, biological visual systems cannot measure the physical properties of the world. The light projected onto the retina is, in fact, unable to specify the physical properties of the world in which humans and other visually ‘intelligent’ animals behave; however, visual behaviours are habitually successful. Through psychophysical evidence, examples of the functioning of Artificial Neural Networks and a reflection upon visual appreciation in the cultural and artistic context, this paper shows how vision emerged by random trial and error during evolution and lifetime learning; how the functioning of ANNs may provide evidence and insights on how machine and human vision works; and how rethinking vision theory in terms of trial and error may offer a new approach to better understand vision—biological and artificial—and reveal new insights into why we like what we like. (shrink)

Purely parallel neural networks can model object recognition in brief displays – the same conditions under which illusory conjunctions have been demonstrated empirically. Correcting errors of illusory conjunction is the “tag-assignment” problem for a purely parallel processor: the problem of assigning a spatial tag to nonspatial features, feature combinations, and objects. This problem must be solved to model human object recognition over a longer time scale. Our model simulates both the parallel processes that may underlie illusory conjunctions and (...) the serial processes that may solve the tag-assignment problem in normal perception. One component of the model extracts pooled features and another provides attentional tags that correct illusory conjunctions. Our approach addresses two questions: How can objects be identified from simultaneously attended features in a parallel, distributed representation? How can the spatial selectional requirements of such an attentional process be met by a separation of pathways for spatial and nonspatial processing? Our analysis of these questions yields a neurally plausible simulation of tag assignment based on synchronizing feature processing activity in a spatial focus of attention. (shrink)

Bowers et al. propose to use controlled behavioral experiments when evaluating deep neural networks as models of biological vision. We agree with the sentiment and draw parallels to the notion that “neuroscience needs behavior.” As a promising path forward, we suggest complementing image recognition tasks with increasingly realistic and well-controlled task environments that engage real-world object recognition behavior.

Two universal functional principles of Grossberg’s Adaptive Resonance Theory decipher the brain code of all biological learning and adaptive intelligence. Low-level representations of multisensory stimuli in their immediate environmental context are formed on the basis of bottom-up activation and under the control of top-down matching rules that integrate high-level, long-term traces of contextual configuration. These universal coding principles lead to the establishment of lasting brain signatures of perceptual experience in all living species, from aplysiae to primates. They are re-visited in (...) this concept paper on the basis of examples drawn from the original code and from some of the most recent related empirical findings on contextual modulation in the brain, highlighting the potential of Grossberg’s pioneering insights and groundbreaking theoretical work for intelligent solutions in the domain of developmental and cognitive robotics. (shrink)

In Natural Ethical Facts William Casebeer argues that we can articulate a fully naturalized ethical theory using concepts from evolutionary biology and cognitive science, and that we can study moral cognition just as we study other forms of cognition. His goal is to show that we have "softly fixed" human natures, that these natures are evolved, and that our lives go well or badly depending on how we satisfy the functional demands of these natures. Natural Ethical Facts is a comprehensive (...) examination of what a plausible moral science would look like.Casebeer begins by discussing the nature of ethics and the possible relationship between science and ethics. He then addresses David Hume's naturalistic fallacy and G. E. Moore's open-question argument, drawing on the work of John Dewey and W. V. O. Quine. He then proposes a functional account of ethics, offering corresponding biological and moral descriptions. Discussing in detail the neural correlates of moral cognition, he argues that neural networks can be used to model ethical function. He then discusses the impact his views of moral epistemology and ontology will have on traditional ethical theory and moral education, concluding that there is room for other moral theories as long as they take into consideration the functional aspect of ethics; the pragmatic neo-Aristotelian virtue theory he proposes thus serves as a moral "big tent." Finally, he addresses objections to ethical naturalism that may arise, and calls for a reconciliation of the sciences and the humanities. "Living well," Casebeer writes, "depends upon reweaving our ethical theories into the warp and woof of our scientific heritage, attending to the myriad consequences such a project will have for the way we live our lives and the manner in which we structure our collective moral institutions.". (shrink)

In Natural Ethical Facts William Casebeer argues that we can articulate a fully naturalized ethical theory using concepts from evolutionary biology and cognitive science, and that we can study moral cognition just as we study other forms of cognition. His goal is to show that we have "softly fixed" human natures, that these natures are evolved, and that our lives go well or badly depending on how we satisfy the functional demands of these natures. Natural Ethical Facts is a comprehensive (...) examination of what a plausible moral science would look like.Casebeer begins by discussing the nature of ethics and the possible relationship between science and ethics. He then addresses David Hume's naturalistic fallacy and G. E. Moore's open-question argument, drawing on the work of John Dewey and W. V. O. Quine. He then proposes a functional account of ethics, offering corresponding biological and moral descriptions. Discussing in detail the neural correlates of moral cognition, he argues that neural networks can be used to model ethical function. He then discusses the impact his views of moral epistemology and ontology will have on traditional ethical theory and moral education, concluding that there is room for other moral theories as long as they take into consideration the functional aspect of ethics; the pragmatic neo-Aristotelian virtue theory he proposes thus serves as a moral "big tent." Finally, he addresses objections to ethical naturalism that may arise, and calls for a reconciliation of the sciences and the humanities. "Living well," Casebeer writes, "depends upon reweaving our ethical theories into the warp and woof of our scientific heritage, attending to the myriad consequences such a project will have for the way we live our lives and the manner in which we structure our collective moral institutions.". (shrink)

We recognize today's deep neural network (DNN) models of language behaviors as engineering achievements. However, what we know intuitively and scientifically about language shows that what DNNs are and how they are trained on bare texts, makes them poor models of mind and brain for language organization, as it interacts with infant biology, maturation, experience, unique principles, and natural law.

Two universal functional principles of Grossberg’s Adaptive Resonance Theory [19] decipher the brain code of all biological learning and adaptive intelligence. Low-level representations of multisensory stimuli in their immediate environmental context are formed on the basis of bottom-up activation and under the control of top-down matching rules that integrate high-level long-term traces of contextual configuration. These universal coding principles lead to the establishment of lasting brain signatures of perceptual experience in all living species, from aplysiae to primates. They are re-visited (...) in this concept paper here on the basis of examples drawn from the original code and from some of the most recent related empirical findings on contextual modulation in the brain, highlighting the potential of Grossberg’s pioneering insights and groundbreaking theoretical work for intelligent solutions in the domain of developmental and cognitive robotics. (shrink)

��This article reviews a theory of explanatory coherence that provides a psychologically plausible account of how people evaluate competing explanations. The theory is implemented in a computational model that uses simple artificial neural networks to simulate many important cases of scientific and legal reasoning. Current research directions include extensions to emotional thinking and implementation in more biologically realistic neural networks.

Symmetry in biological and physical systems is a product of self-organization driven by evolutionary processes, or mechanical systems under constraints. Symmetry-based feature extraction or representation by neural networks may unravel the most informative contents in large image databases. Despite significant achievements of artificial intelligence in recognition and classification of regular patterns, the problem of uncertainty remains a major challenge in ambiguous data. In this study, we present an artificial neural network that detects symmetry uncertainty states in human (...) observers. To this end, we exploit a neural network metric in the output of a biologically inspired Self- Organizing Map Quantization Error (SOM-QE). Shape pairs with perfect geometry mirror symmetry but a non-homogenous appearance, caused by local variations in hue, saturation, or lightness within and/or across the shapes in a given pair produce, as shown here, a longer choice response time (RT) for “yes” responses relative to symmetry. These data are consistently mirrored by the variations in the SOM-QE from unsupervised neural network analysis of the same stimulus images. The neural network metric is thus capable of detecting and scaling human symmetry uncertainty in response to patterns. Such capacity is tightly linked to the metric’s proven selectivity to local contrast and color variations in large and highly complex image data. (shrink)

In their account of learning and behavior, the authors define an interactor as emitted behavior that operates on the environment, which excludes Pavlovian learning. A unified neural-network account of the operant-Pavlovian dichotomy favors interpreting neurons as interactors and synaptic efficacies as replicators. The latter interpretation implies that single-synapse change is inherently Lamarckian.

It was recently shown that the Madelung equations, that is, a hydrodynamic form of the Schrödinger equation, can be derived from a canonical ensemble of neural networks where the quantum phase was identified with the free energy of hidden variables. We consider instead a grand canonical ensemble of neural networks, by allowing an exchange of neurons with an auxiliary subsystem, to show that the free energy must also be multivalued. By imposing the multivaluedness condition on the (...) free energy we derive the Schrödinger equation with “Planck’s constant” determined by the chemical potential of hidden variables. This shows that quantum mechanics provides a correct statistical description of the dynamics of the grand canonical ensemble of neural networks at the learning equilibrium. We also discuss implications of the results for machine learning, fundamental physics and, in a more speculative way, evolutionary biology. (shrink)

In recent years, neural networks have attracted much attention in the machine learning and the deep learning technologies. Bio-inspired functions and intelligence are also expected to process efficiently and improve existing technologies. In the visual pathway, the prominent features consist of nonlinear characteristics of squaring and rectification functions observed in the retinal and visual cortex networks, respectively. Further, adaptation is an important feature to activate the biological systems, efficiently. Recently, to overcome short-comings of the deep learning techniques, (...) orthogonality for the weights in the networks has been developed for the signal propagation and the efficient optimization of the learning. In this paper, bio-inspired asymmetric networks with nonlinear characteristics are proposed, which are derived from the retinal networks in the biological visual pathway. The asymmetric network proposed here was verified to detect the movement of the object, efficiently in our previous studies. This paper shows a new characteristic of the adaptive orthogonality in the asymmetric networks. First, it is shown that the asymmetric network with nonlinear characteristics is effective for generating orthogonality. Second, the proposed asymmetric network with Gabor filters is compared with the conventional energy model from the point of the orthogonality characteristics. Finally, the asymmetric networks with nonlinear characteristics can generate the extended orthogonal bases in independent subspaces, which are useful for classification and efficient learning. (shrink)

Current cognitive science models of perception and action assume that the objects that we move toward and perceive are represented as determinate in our experience of them. A proper phenomenology of perception and action, however, shows that we experience objects indeterminately when we are perceiving them or moving toward them. This indeterminacy, as it relates to simple movement and perception, is captured in the proposed phenomenologically based recurrent network models of brain function. These models provide a possible foundation from which (...) predicative structures may arise as an emergent phenomenon without the positing of a representing subject. These models go some way in addressing the dual constraints of phenomenological accuracy and neurophysiological plausibility that ought to guide all projects devoted to discovering the physical basis of human experience. (shrink)

Choice Outstanding Academic Title, 1996. In hundreds of articles by experts from around the world, and in overviews and "road maps" prepared by the editor, The Handbook of Brain Theory and Neural Networks charts the immense progress made in recent years in many specific areas related to great questions: How does the brain work? How can we build intelligent machines? While many books discuss limited aspects of one subfield or another of brain theory and neural networks, (...) the Handbook covers the entire sweep of topics—from detailed models of single neurons, analyses of a wide variety of biological neural networks, and connectionist studies of psychology and language, to mathematical analyses of a variety of abstract neural networks, and technological applications of adaptive, artificial neural networks. Expository material makes the book accessible to readers with varied backgrounds while still offering a clear view of the recent, specialized research on specific topics. (shrink)

The present commentary addresses the Quartz & Sejnowski (Q&S) target article from the point of view of the dynamical learning algorithm for neural networks. These techniques implicitly adopt Q&S's neural constructivist paradigm. Their approach hence receives support from the biological and psychological evidence. Limitations of constructive learning for neural networks are discussed with an emphasis on grammar learning.

By pointing to deep philosophical confusions endemic to cognitive science, Wittgenstein might seem an enemy of computational approaches. We agree that while Wittgenstein would reject the classicist’s symbols and rules approach, his observations align well with connectionist or neural network approaches. While many connectionisms that dominated the later twentieth century could fall prey to criticisms of biological, pedagogical, and linguistic implausibility, current connectionist approaches can resolve those problems in a Wittgenstein-friendly manner. We present the basics of a Vector Symbolic (...) Architecture formalism, inspired by Smolensky, and indicate how high-dimensional vectors can operate in a context-sensitive and object-independent manner in biologically plausible time scales, reflecting Wittgenstein’s notions of language-games and family resemblance; we show how “soft” symbols for such a formalism can be formed with plausible learning cycles using Sparse Distributed Memory, resolving disputes surrounding Wittgenstein’s private language argument; and show how connectionist networks can extrapolate meaningful patterns to solve problems, providing “ways to go on” without explicit rules, which indicates linguistic plausibility. Connectionism thus provides a systematicity and productivity that is more than a mere implementation of a classical approach, and provides Wittgenstein-friendly and Wittgenstein-illuminating models of mind and language for cognitive science. (shrink)

Page has done connectionist researchers a valuable service in this target article. He points out that connectionist models using localized representations often work as well or better than models using distributed representations. I point out that models using distributed representations are difficult to understand and often lack parsimony and plausibility. In conclusion, I give an example – the case of the missing fundamental in music – that can easily be explained by a model using localist representations but can be explained (...) only with great difficulty and implausibility by a model using distributed representations. (shrink)

Psychologically faithful deep neural networks (DNNs) could be constructed by training with psychophysics data. Moreover, conventional DNNs are mostly monocular vision based, whereas the human brain relies mainly on binocular vision. DNNs developed as smaller vision agent networks associated with fundamental and less intelligent visual activities, can be combined to simulate more intelligent visual activities done by the biological brain.

The action observation networks (AON) (or the mirror neuron system) are the neural underpinnings of visuomotor integration and play an important role in motor control. Besides, one of the main functions of the human mirror neuron system is recognition of observed actions and the prediction of its outcome through the comparison with the internal mental motor representation. Previous studies focused on the human mirror neurons (MNs) activation during object-oriented movements observation, therefore intransitive movements observation effects on MNs activity (...) remains relatively little-studied. Moreover, the dependence of MNs activation on the biomechanical characteristics of observed movement and their biological plausibility remained highly underexplored. In this study we proposed that naturalness of observed intransitive movement can modulate the MNs activity. Event-related desynchronization (ERD) of sensorimotor electroencephalography (EEG) rhythms, N400 event-related potentials (ERPs) component and corticospinal excitability were investigated in twenty healthy volunteers during observation of simple non-transitive finger flexion that might be either biomechanically natural or unnatural when finger wriggled out toward the dorsal side of palm. We showed that both natural and unnatural movements caused mu/beta-desynchronization, which gradually increased during the flexion phase and returned to baseline while observation of extension. Desynchronization of the mu-rhythm was significantly higher during observation of the natural movements. At the same time, beta-rhythm was not found to be sensitive to the action naturalness. Also, observation of unnatural movements caused an increased amplitude of the N400 component registered in the centro-parietal regions. We suggest that the sensitivity of N400 to intransitive action observation with no explicit semantic context might imply the broader role of N400 sources within AON. Surprisingly, no changes in corticospinal excitability were found. This lack of excitability modulation by action observation could be related with dependence of the M1 activity on the observed movement phase. (shrink)