In the evolution of complex organs, such as the brain, phases of gradual modification are repeatedly interrupted by discrete innovations that mark the evolutionary "invention" of novel architectures or information processing principles. Capturing the basic principles of such key neurobiological innovations and explaining their sequence is one of the major challenges for theoretical and evolutionary neuroscience. Prof. Wolf's group is currently working on two problem areas in this field: the reorganization of the visual cortex network at the origin of the primate brain and the optimization of the neural code for population coding in forebrain evolution.

In the cerebral cortex, relevant information is usually represented and processed by the activity of populations of thousands of neurons. Nevertheless, the performance and processing speed of biological neuronal networks depend critically on the specialized molecular architecture and biophysics of individual neurons. In a number of projects, Prof. Wolf's group is engaged in linking aspects of the molecular architecture of specialized compartments of the neuron to performance at the level of large networks in a quantitative and theoretically sound manner.

Humans, animals, and artificial agents are often faced with the task of using information about their environment in a social context to control behavior, where the behavior of other actors both determines the evaluation of possible actions and can be used as their own source of information. Traditionally, optimal strategies for social actors have been studied using game theoretic models and their evolutionary optimization. However, such game-theoretic models have so far neglected the fact that the processing of social and sensory information, action planning and initiation usually proceed in parallel. Thus, Prof. Wolf has introduced a new class of game theoretic models, "transparent games", which allows to take this adequately into account.

The networks of the cerebral cortex are able to restructure their architecture through learning and self-organization processes enabling nerve cell populations to take on new tasks or improve their performance. This neuroplasticity, which is particularly highly developed in the young brain, is the crucial basis of our ability to acquire new skills throughout life. Prof. Löwel's group is addressing the questions of which molecular processes underlie the high plasticity of the juvenile brain, how it is transferred into an adult form, and whether a learning capacity similar to that of the young brain can be restored for therapeutic purposes. For example, the group was able to discover a novel mechanism for the regulation of plasticity in the visual cortex. These studies are complemented by work in the CIDBN Neurophysics Laboratory on conditioned in vitro models of network plasticity.

The ability of the cerebral cortex networks to reorganize in an experience-driven manner is not determined by age alone, but can be influenced by environmental conditions and physical activity. In now classical work, Prof. Löwel was able to show in the rodent model that growing up in an enriched environment is sufficient to maintain functional plasticity of the same extent in the adult state as in young age. Most recently, Prof. Löwel's group has been able to further investigate the extent and limits of adult plasticity and determine ways to induce enhanced plasticity for therapeutic purposes.

The architecture of biological networks essentially serves to mediate, guide, and integrate the information flows between their elements. Quantitatively capturing these information flows is a central challenge of data-driven analysis of biological networks. Prof. Wibral's group has further developed methods for measuring transfer information and information measures of specific and synergistic components of information flows and successfully applied them in magnetoencephalography experiments with neurotypical and psychiatric subjects.

Biological networks fulfill their function through the organized interaction of their elements, which is determined by working points of the network and can thus be modulated depending on the state. Examples include the variable lifetime of working memory traces, but also the nature and intensity of an infection event in an interacting population of organisms. To determine core parameters of such network working points, Prof. Wibral's group has developed and applied methods of a previously unattained resolving power during the reporting period.

The analysis of the dynamics of biological networks as well as their technological use and therapeutic interventions are currently progressing very dynamically due to the development of biological interface technologies. In this context, high performance interfaces require the optimization of the effectors used, tuned to the intrinsic dynamics of the specific system and the capture of the coupled systems of interface and biological network in mathematical models. Within the CIDBN, the Neurophysics Laboratory headed by Dr. Neef is a competence center and platform for neural interface technologies and their applications.

Complex fluids can self-organize spatially by separation of different fluid phases. For some years now, more and more functional compartments have been detected in the field of cell biology that are formed by analogous phase separation processes. In contrast to classical physical models of phase separation, many of these systems are maintained in thermodynamic disequilibrium by energy input and can thereby express novel and biologically functional properties. Dr. Zwicker's group has made a number of contributions to this young field of living systems dynamics.

The dynamics of evolutionary adaptation of living systems is the fundamental process of biological evolution. However, a large number of living subsystems are subject to mutual adaptation, coevolution, which need not result in a steady-state equilibrium. For this reason, the dynamics of co-evolving systems, such as pathogens and their host's immune response, is rich and still poorly understood. Dr. Nourmohammad and her group use methods from statistical mechanics of disequilibrium, optimization theory, and systems theory to study paradigmatic examples of coevolutionary dynamics.