Neural networks : the official journal of the International Neural Network Society
August 1, 2013
James A. Reggia
186 citations
Computational models of consciousness, known as artificial consciousness, have been developed over the last two decades with two main goals: to better understand human and animal consciousness and to create machines with conscious awareness. This review categorizes models into five types based on their central focus: global workspace, information integration, internal self-model, higher-level representations, or attention mechanisms. The review concludes that computational modeling is now an accepted scientific method for studying consciousness, and existing models have successfully simulated many neurobiological and cognitive correlates of conscious processing. However, no current approach has convincingly demonstrated phenomenal machine consciousness or provided clear evidence that it will eventually be possible.
Neural networks : the official journal of the International Neural Network Society
July 1, 2025
Rita Pizzi, Hao Quan, Matteo Matteucci et al.
A new variant of the Turing Test is proposed that relies on the emotional content of human cognition and its link with memory enhancement, a feature argued to be inaccessible to machines even in principle. EEG signals were recorded from 39 subjects who underwent a test where they recognized images previously associated with an emotional reaction more easily than those without such an association, a statistically significant difference. The authors suggest this distinction, rooted in human qualia, could make human and machine responses distinguishable regardless of advances in artificial intelligence.
Neural networks : the official journal of the International Neural Network Society
December 1, 2024
Michele Farisco, Kathinka Evers, Jean-Pierre Changeux
Analyzing the question of developing artificial consciousness from an evolutionary perspective, using the evolution of the human brain and its relation with consciousness as a benchmark, reveals several structural and functional features of the human brain that appear key for human-like complex conscious experience. Current AI research should take these into account. Even if AI is limited in emulating human consciousness for intrinsic and extrinsic reasons, taking inspiration from brain characteristics that make human-like conscious processing possible is a promising strategy.
Neural networks : the official journal of the International Neural Network Society
August 1, 2024
Giovanni Granato, Gianluca Baldassarre
Consciousness helps align mental representations with goals, enabling more flexible and effective behavior. A new theoretical framework integrates cognitive neuroscience and AI/robotics, extending a previously validated model of goal-directed flexible cognition tested in over 20 human populations. The theory bridges major accounts of consciousness and goal-directed action, offering implications for designing experimental tasks, interpreting clinical data, and improving machine learning and robotic systems, including applications in digital-twin healthcare and roboethics.
Neural networks : the official journal of the International Neural Network Society
March 1, 2024
Neil W. Bailey, Ben D Fulcher, Bridget Caldwell et al.
Over 7,000 time-series features extracted from resting EEG data distinguished meditators from non-meditators with 67% accuracy, but only in a central-parietal brain component. Meditators showed higher temporal stability (more consistent statistical properties across short time segments) and altered distribution of voltage values around the mean. Traditional band-power measures failed to differentiate the groups. The findings suggest that meditation is associated with greater attentional stability reflected in more stationary neural dynamics.
Neural networks : the official journal of the International Neural Network Society
June 1, 2021
Haruka Nakamura, Yoshimasa Tawatsuji, Siyuan Fang et al.
Mindfulness meditation's emotion regulation mechanism involves two opposing brain processes. A computational model based on anatomical brain networks shows that mindfulness increases output from the thalamus and sensory cortex, activating the insula and anterior cingulate cortex. This can trigger top-down inhibition of the amygdala by the orbitofrontal and dorsolateral prefrontal cortices. However, when bottom-up sensory processing dominates, amygdala activity increases through insula and ACC activation, revealing how previously reported neural patterns during mindfulness arise from information propagation across brain regions.
Neural networks : the official journal of the International Neural Network Society
March 1, 2017
Stephen Grossberg
Conscious experiences of seeing, hearing, feeling, and knowing arise from resonant states in the brain, according to Adaptive Resonance Theory (ART). ART explains how brains autonomously learn to attend, recognize, and predict objects and events. It specifies mechanistic links between consciousness, learning, expectation, attention, resonance, and synchrony. Not all resonances become conscious, and not all brain dynamics are resonant. The theory classifies brain resonances that support conscious experiences, clarifying psychological and neurobiological data in normal individuals and clinical patients. Complementary and laminar cortical processing figure prominently in explanations of conscious and unconscious processes.
Neural networks : the official journal of the International Neural Network Society
August 1, 2012
Nagendra Marupaka, Laxmi R Iyer, Ali A Minai
Thought arises from the combination of existing concepts through associations, and the structure of semantic networks—small-world, scale-free connectivity—influences the creativity of ideation. A new neural model represents semantic memory as a recurrent network with itinerant dynamics, where conceptual combinations emerge as co-active neural groups and persist as metastable attractors, recognized as ideas. Simulations show that networks with both small-world and scale-free characteristics significantly enhance the generation of unique conceptual combinations, linking the qualitative structure of associations to the effectiveness of spontaneous thought.
Neural networks : the official journal of the International Neural Network Society
November 1, 2007
Edmund T Rolls
Recordings from populations of neurons in the inferior temporal visual cortex indicate that most information about which stimulus is shown comes from the firing rate of each neuron, not from stimulus-dependent synchrony, making it unlikely that synchrony or oscillations are essential for visual object perception. Backward masking experiments show the threshold for conscious visual perception is set higher than the level at which small but significant information exists in neuronal firing, allowing guessing without conscious awareness; this may prevent noise in sensory pathways from interrupting conscious thought processes.
Neural networks : the official journal of the International Neural Network Society
November 1, 2007
Axel Cleeremans, Bert Timmermans, Antoine Pasquali
Consciousness involves not only being aware of something but also being aware that one is aware. This paper explores computational mechanisms that could implement this idea, starting from the observation that connectionist networks acquire knowledge that remains embedded in their task performance—knowledge in the network rather than knowledge for the network. The authors present simulations where a second-order network observes the internal states of a first-order network trained on a categorization task. The second-order network either reproduces those states or uses them as cues for a secondary task, achieving a limited form of metarepresentation. This mechanism offers a beginning for understanding mental attitudes and suggests that consciousness involves learned knowledge of one's own internal representations through interaction with self, world, and others.
Neural networks : the official journal of the International Neural Network Society
October 1, 1997
Sung Bae Cho, Bernard J. Baars, James Newman
A neurocognitive model of consciousness is presented, defining it as a global integration and dissemination system nested within a distributed array of specialized bioprocessors. This system controls the allocation of central nervous system processing resources via cortical gating of a strategic thalamic nucleus. The model integrates experimental data from cognitive psychology, artificial intelligence, and neuroscience, building on developments of Baars' Global Workspace theory. The basic circuitry of this neural system is reasonably well understood and can be approximated using neural network principles.