Abstract: Human agents draw a variety of inferences effortlessly, spontaneously, and with remarkable efficiency — as though these inferences are a reflex response of their cognitive apparatus. Furthermore, these inferences are drawn with reference to a large body of background knowledge. This remarkable human ability seems paradoxical given the results about the complexity of reasoning reported by researchers in artificial intelligence. It also poses a challenge for cognitive science and computational neuroscience: How can a system of simple and slow neuron-like elements represent a large body of systematic knowledge and perform a range of inferences with such speed? We describe a computational model that is a step toward addressing the cognitive science challenge and resolving the artificial intelligence paradox. We show how a connectionist network can encode millions of facts and rules involving n-ary predicates and variables, and perform a class of inferences in a few hundred msec. Efficient reasoning requires the rapid representation and propagation of dynamic bindings. Our model achieves this by i) representing dynamic bindings as the synchronous firing of appropriate nodes, ii) rules as interconnection patterns

Consciousness poses the most baffling problems in the science of the mind. There is nothing that we know more intimately than conscious experience, but there is nothing that is harder to explain. All sorts of mental phenomena have yielded to scientific investigation in recent years, but consciousness has stubbornly resisted. Many have tried to explain it, but the

Abstract. Networks of compartmental model neurons were used to investigate the biophysical basis of the synchronization observed between sparsely-connected neurons in neocortex. A model of a single column in layer 5 consisted of 100 model neurons: 80 pyramidal and 20 inhibitory. The pyramidal cells had conductances that caused intrinsic repetitive bursting at different frequencies when driven with the same input. When connected randomly with a connection density of lo%, a single model column displayed synchronous oscillatory action potentials in response to stationary, uncorrelated Poisson spike-train inputs. Synchrony required a high ratio of inhibitory to excitatory synaptic strength; the optimal ratio was 4: 1, within the range observed in cortex. The synchrony was insensitive to variation in amplitudes of postsynaptic potentials and synaptic delay times, even when the mean synaptic delay times were varied over the range 1 to 7 ms. Synchrony was found to be sensitive to the strength of reciprocal inhibition between the inhibitory neurons in one column: Too weak or too strong reciprocal inhibition degraded intra-columnar synchrony. The only parameter that affected the oscillation frequency of the network was the strength of the external driving input which could shift the frequency between 35 to 60 Hz. The same results were obtained using a model column of 1000 neurons with a connection density of 5%, except that the oscillation became more regular. Synchronization between cortical columns was studied in a model consisting of two columns with 100 model

How does the brain group together different parts of an object into a coherent visual object representation? Different parts of an object may be processed by the brain at different rates and may thus become desynchronized. Perceptual framing is a process that resynchronizes cortical activities corresponding to the same retinal object. A neural network model is presented that is able to rapidly resynchronize desynchronized neural activities. The model provides a link between perceptual and brain data. Model properties quantitatively simulate perceptual framing data, including psychophysical data about temporal order judgments and the reduction of threshold contrast as a function of stimulus length. Such a model has earlier been used to explain data about illusory contour formation, texture segregation, shape-from-shading, 3-D vision, and cortical receptive fields. The model hereby shows how many data may be understood as manifestations of a cortical grouping process that can rapidly res...

In this paper, a novel model of object-based visual attention extending Duncan's Integrated Competition Hypothesis [24] is presented. In contrast to the attention mechanisms used in most previous machine vision systems which drive attention based on the spatial location hypothesis, the mechanisms which direct visual attention in our system are object-driven as well as feature-driven. The competition to gain visual attention occurs not only within an object but also between objects. For this purpose, two new mechanisms in the proposed model are described and analyzed in detail. The first mechanism computes the visual salience of objects and groupings; the second one implements the hierarchical selectivity of attentional shifts. The results of the new approach on synthetic and natural images are reported.

We propose a computational theory of consciousness and model data from three experiments in visual perception. The central idea of our theory is that the contents of consciousness correspond to temporally stable states in an interconnected network of specialized computational modules. Each module incorporates a relaxation search that is concerned with achieving semantically well-formed states. We claim that being an attractor of the relaxation search is a necessary condition for awareness. We show that the model provides sensible explanations for the results of three experiments, and makes testable predictions. The first experiment (Marcel, 1980) found that masked, ambiguous prime words facilitate lexical decision for targets related to either prime meaning, whereas consciously perceived primes facilitate only the meaning that is consistent with prior context. The second experiment (Fehrer and Raab, 1962) found that subjects can make detection responses in constant time to simple visual stimuli regardless of whether they are consciously perceived or masked by metacontrast and not consciously perceived. The third experiment (Levy and Pashler, 1996) found that visual word recognition accuracy is lower than baseline when an earlier speeded response was incorrect, and higher than baseline when the early response was correct, consistent with a causal relationship between conscious perception and subsequent processing.

Abstract not found

d together in a specific order, because different words can be written with the same letters. The present research had two aims: first, to clarify the cerebral stages of processing that lead to invariant word recognition, and second, to examine whether those stages can proceed in the absence of consciousness. In literate adults, an extended strip of the left fusiform gyrus activates whenever visual words are presented (Cohen et al., 2000; Cohen et al., 2002; Nobre, Allison, & McCarthy, 1994). This region, which has been termed the Visual Word Form Area (VWFA), is responsive only to written words, not to spoken words (Dehaene, Le Clec'H, Poline, Le Bihan, & Cohen, 2002). Its lesioning results in a severe word identification impairment, pure alexia, which is restricted to the visual modality (Leff et al., 2001). Thus, it is a plausible candidate for the neural basis of invariant visual word recognition. To further specify the nature of word coding in the VWFA, we used the priming metho

The search for neural correlates of consciousness (or NCCs) is arguably the cornerstone in the recent resurgence of the science of consciousness. The search poses many difficult empirical problems, but it seems to be tractable in principle, and some ingenious studies in recent years have led to considerable progress. A number of proposals have been put forward

This article reviews the limitations of the standard computing paradigm and sketches the concept of quantum neural computing. Implications of this idea for the understanding of biological information processing and design of new kinds of computing machines are described. Arguments are presented in support of the thesis that brains are to be viewed as quantum systems with their neural structures representing the classical measurement hardware. From a performance point of view, a quantum neural computer may be viewed as a collection of many conventional computers that are designed to solve different problems. A quantum neural computer is a single machine that reorganizes itself, in response to a stimulus, to perform a useful computation. Selectivity offered by such a reorganization appears to be at the basis of the gestalt style of biological information processing. Clearly, a quantum neural computer is more versatile than the conventional computing machine.