Understanding of the molecular and neural bases of biological timing, especially of the timing of circadian (ca. 24 hours) and seasonal phenomena, have been rapidly advanced for the last decade. The mechanism responsible for these functions is the circadian system which is consisted of the central clock located in the suprachiasmatic nucleus (SCN) of the hypothalamus and the peripheral clocks in a variety of tissues. The SCN is directly innervated by the retinal ganglion cells and received photic information. The SCN is also innervated by the ascending fibers originated from the brain stem, but the functions of these neural pathways are not well understood. The SCN neurons send the fibers to various regions in the brain, and also secrete neuropepites, which are supposed to mediate the circadian information from the SCN to the peripheral clocks.

A unilateral SCN consists of ca.10,000 neurons, most of which show the circadian rhythms in neuronal activity. In rats and mice, the circadian periods of individual neurons in the cultured SCN are not identical but show a Gaussian distribution. The range of distribution is wider in the dispersed cell culture than in the slice, suggesting that the architecture or the neural network of the SCN is important to exhibit coherent circadian activity rhythms. However, a line of evidence indicates that the origin of circadian rhythm in the SCN neurons is not the neural network but the transcription-translation feedback of cellular molecules. Several clock genes and their protein products are involved in the feedback loop and it takes approximately 24 hours for one turn of this loop. Mutation or knockdown of one of these clock genes leads to substantial changes in the circadian period or abolishment of circadian rhythm in behavior.

The circadian behavior rhythm in mammals consists of an active period (wakefulness) and a rest period (sleep). The length of active or rest period has been well known as the subject of seasonal changes in photoperiod. These seasonal changes are ascribed to more than one regional pacemaker in the SCN circadian clock which responds independently to photic information from the retina. The mutual coupling among the regional pacemakers is of special importance in the control of behavior rhythm. Thus, the circadian clock in the SCN is understood as a hierarchical multi-oscillator system.

In the present session, in order to have a better understanding of biological timing in behavior, the neural basis and dynamics of clock work in the mammalian SCN will be discussed not only from viewpoints of biology but also of mathematical modeling.

Neuronal synchronization is considered to play an important role for the control of cognitive functions in physiological states (e.g. sensory transmission, information binding, regulation of sleep-wake cycles) as well as under pathophysiological conditions (e.g. as the cause of epileptic seizures or the tremor in Parkinson’s disease). Most synchronization studies focus on the neurons’ connectivity. The impact of intrinsic neuronal dynamics is often neglected although it is known that functionally most relevant changes of synchronization states are typically going along with alterations of neuronal impulse patterns and oscillations which, in turn, can significantly be influenced by noise. This session specifically aims to elucidate the physiologically and pathophysiologically relevant control parameters of neural synchronization by means of computer simulations to attain further insights into presumably complex interdependencies between neuronal networks’ connectivity, neural impulse pattern, ionic membrane processes, and noise effects.

The understanding of cognitive dynamics and brain related disorders requires considering the interaction of different systems at different functional levels, from subcellular mechanisms to neurons, synapses and neuronal networks up to cognition and behavior. In this situation, computer models can be of particular value especially in extracting the physiologically relevant mechanisms of the system dynamics and elucidating the pathophysiologically critical conditions at which the system goes into disordered state - in the expectation thereby also to find new targets for the treatment of the disease. The actual simulation approaches, however, exhibit an enormous heterogeneity. Bringing together different approaches from different levels on a common platform is a major challenge for large, integrative projects like “physiome” or “silicon cell”. Accordingly, the contributors shall be encouraged to illustrate how their approaches could be combined with others. The aim is to stimulate inspiring discussions about the most effective computational strategies for future simulations in cognitive neurodynmics.

Neural activity is characterized by great variability. Electrocortical recording of brain activity shows dynamic patterns of synchronization and desynchronization. These patterns occur at a variety of scales, durations, and frequencies. In humans, they are the neural correlate of cognitive processes. Cognitive processes need to be flexible, but are also subject to requirements of robustness and reliability. How can these, seemingly opposite, demands be met with the complexity of brain processes? Despite their variability, the dynamic patterns of brain activity also show regularities that offer cues on how information processes proceed in the brain. Traditionally, efforts have been concentrated on evoked brain potentials and efforts to localize static sources of brain activity. Such an approach, however, implies treating the complexity of brain activity as residual fluctuation. More recently, researchers have made addressed the complexity in different ways, based on the analysis of spatiotemporal patterns of phase synchronization, moving wave patches, standing and traveling waves. This session will be focusing on these analyses and their relevance to cognitive processing of information.

The aim of this session is to bring together various research fields in order to understand our intelligence which emerges from interactions with the environment. In our daily life, we behave adaptively depending upon ever-changing situation and also communicate with others. These intelligent behaviors involve various aspects of higher cognitive functions such as thought, reasoning, planning and integration of information in time. Although a lot of efforts have been made to investigate intelligence in the fields of neuroscience, cognitive science, engineering and robotics, underlying mechanisms are still open questions.

In the conventional studies, stimulus-response paradigms have often been used to investigate cognitive functions. However, such simple sensorimotor reactions may not lead us to understand adaptive and emergent aspects of our behavior. Here, we would like to emphasize the importance of information flow (throughout multiple brain regions) in a system-environment loop and dynamics of internal representations of the environment in order to understand mechanisms of intelligence. As the environment has huge degrees of freedom and is ever-changing, a predefined algorithm is not suitable to process information. We often apply heuristic approaches based on our past experiences to find a behavioral strategy under a current situation. In other words, our behavioral principle emerges from the interactions with the environment and the interactions also affect the internal representations of the environment. Such emergent properties would likely result in externally observed “intelligent behavior” in a complex situation.

We would welcome papers describing theoretical and experimental works that bridge between multidisciplinary fields relevant to the aim, and we encourage submissions dealing with new paradigms for designing an intelligent robot and brain-inspired artificial systems.

People are overwhelmed by information flow coming from the physical and social environment. While developmental psychology knows from the works of Piaget that infants intertwine the two kinds of information, it is only recently that cognitive neuroscience starts to understand their tight connexion and functioning within the brain, and that developmental robotics' community starts to consider their interplay as a design principle for robot interactions. Converging viewpoints agree on the importance of the perception-action loop (that is, physical embodiment) – i.e., the temporal integration across different modalities (contingency detection, timing, rhythm and synchrony) – to construct the neural circuits necessary for physical and social interactions. These shared circuits are likely to appear early during infant development when babies start to interact with their surrounding environment and begin to communicate non-verbally with caregivers. In cognitive neuroscience, this idea challenges the view how the brain represents oneself and others, their relation as well as their segregation, giving a bigger role to perception in contrast to a more motor-centric representation of the brain; e.g., the mirror neurons-like cells in the parietal lobe or in the pre-motor cortex. Furthermore, to understand the neural mechanisms involved, it is important that robotics propose computational models of the underlying cognitive neurodynamics.

This session aims at presenting recent researches in the fields of cognitive neuroscience, developmental psychology and developmental robotics, which study the interplays between physical and social cognition at the brain level and/or at the body level and the construction of sensorimotor circuits through physical and social interactions.

The workshop is intended for an inter-disciplinary audience and attempts to provide an equal exposure between experimental observations and computational neuroscience models or theories: How physical and social interactions shape progressively the neural dynamics for embodied cognition? Reversely, how sensorimotor defects/impairments may induce developmental disorder such as autism?

Relevant topics for this session include but are not limited to the experiments, observations, modeling and theoretical framework about; motor and social development, imitation, parieto-motor networks, the mirror neurons system, developmental and neural disorders linked to embodiment impairing, social resonance/interaction, joint attention, spatial and social representation of the body, self-other distinction/representation, self-perception and agency, body-ownership.

The increasing affordability of computer power has produced a recent trend in neural network modeling towards large-scale and supercomputational approaches, taking into account detailed biophysical properties of the individual network elements (neurons or neuronal populations). However, there are a number of network phenomena that can also be replicated with much simpler models. An example is the modularity of functional connectivity that can be observed in models ranging in complexity from large populations of coupled oscillators (Zhou et al. PRL 2007) to networks formed by discrete excitable nodes (Mueller-Linow et al. PLoS CB 2008).

These observations pose the question: How intricate does a neural model have to be in order to produce a particular network phenomenon, such as irregular sustained activity, bursting, neural avalanches or slow-frequency coupling of high-frequency oscillators? What are the minimal models for these phenomena, what features (e.g., noise, delays, heterogeneity of connections) do they need to include? While analytical answers for all these uestions may be still out of reach, we seek an improved practical understanding of essential parameters for neural network modeling.

The symposium thus proposes to bring together researchers working on structure-dynamics relationship in neural systems with models of various complexity to share their experience on neural network modeling.

Neural networks are highly complex systems, where rich dynamical phenomena take place, including self-organization and learning. They are utilized for information processing. In order to understand these rich phenomena, we need mathematical modeling and analysis of various aspects of dynamics of neural networks. They include complex dynamics including attractors, synchronizations and chaos, self-organization of dynamical structures and dynamics of learning, statistical and stochastic analysis of spikes and neural networks, etc. The present session welcomes any mathematical aspects of neural networks and neurodynamics, aiming at enriching mathematical methods to establish mathematical neuroscience.

Our goal is to examine deliberative decision-making from neurophysiological, cognitive, and mathematical perspectives. Deliberative decision-making entails behaviors in which agents explicitly consider multiple possibilities before acting. Early experiments identified behaviors in which a rat paused, looking back and forth, before acting. This behavior occurred during learning and after changes in experimental contingencies, and was suggested to create expectancies of the consequences of the available choices, but the computational and experimental techniques needed to address this question were not available at the time. An interest in this behavior has been revived in the past few years based on new computational models of decision-making systems, on the availability of new behavioral measures, and on recent discoveries of prospective encoding in the hippocampus and other structures. This endeavor brings together expertise in behavioral neuroscience, neurophysiology, computational neuroscience, cognitive psychology, and mathematics. Experiments will examine prediction, evaluation, and action-selection mechanisms in rats making deliberative decisions. Computational projects will examine mechanisms by which these systems can shift suddenly from consideration of one possibility to the other, based on models of chaotic attractors.

A further aim is to establish a new cross-disciplinary research field, comprised of “mathematical systems theory,” “interactions among hetero brain systems,” and “interactions among individuals.” The theory of describing brain dynamics for communication will be established via studies on evolutional systems, random dynamic systems, hybrid systems that can simultaneously treat discrete and continuous variables, and extended dynamic systems based on neuronal entrainment or chaotic itinerancy. By measuring the brain activity during communication, it is possible to investigate the entrainment of neuronal activities in distributed cortices, and the underlying mechanism of dynamical emergence of the mnemonic function. Finally, several research projects target the strategy for communication between humans, humans and non-human primates, humans and robots, or humans and the environment by measuring behavioral and brain activities.