Functional and spatial rewiring jointly generate convergent-divergent units in self-organizing networks
Self-organization through adaptive rewiring of random neural networks generates brain-like topologies comprising modular small-world structures with rich club effects, merely as the product of optimizing the network topology. In the nervous system, spatial organization is optimized no less by rewiring, through minimizing wiring distance and maximizing spatially aligned wiring layouts. We show that such spatial organization principles interact constructively with adaptive rewiring, contributing to establish the networks' connectedness and modular structures. We use an evolving neural network model with weighted and directed connections, in which neural traffic flow is based on consensus and advection dynamics, to show that wiring cost minimization supports adaptive rewiring in creating convergent-divergent unit structures. Convergent-divergent units consist of a convergent input-hub, connected to a divergent output-hub via subnetworks of intermediate nodes, which may function as the computational core of the unit. The prominence of minimizing wiring distance in the dynamic evolution of the network determines the extent to which the core is encapsulated from the rest of the network, i.e., the context-sensitivity of its computations. This corresponds to the central role convergent-divergent units play in establishing context-sensitivity in neuronal information processing.
Code (0)
등록된 구현이 없습니다.
Tasks
SensitivitySimilar Papers 제목 키워드 기반
Adaptive rewiring of random neural networks generates convergent-divergent units
Brain networks are adaptively rewired continually, adjusting their topology to bring about functionality and efficiency in sensory, motor and cognitive tasks. In model neural network architectures, adaptive rewiring gene…
Distributed ComputingOn higher order computations, rewiring the connectome, and non-von Neumann computer architecture
Structural plasticity in the brain (i.e. rewiring the connectome) may be viewed as mechanisms for dynamic reconfiguration of neural circuits. First order computations in the brain are done by static neural circuits, wher…
Locality-Aware Graph-Rewiring in GNNs
Graph Neural Networks (GNNs) are popular models for machine learning on graphs that typically follow the message-passing paradigm, whereby the feature of a node is updated recursively upon aggregating information over it…
Inductive BiasRewiring Neurons in Non-Stationary Environments
The human brain rewires itself for neuroplasticity in the presence of new tasks. We are inspired to harness this key process in continual reinforcement learning, prioritizing adaptation to non-stationary environments. In…
Continual Learningreinforcement-learningReinforcement LearningPANDA: Expanded Width-Aware Message Passing Beyond Rewiring
Recent research in the field of graph neural network (GNN) has identified a critical issue known as "over-squashing," resulting from the bottleneck phenomenon in graph structures, which impedes the propagation of long-ra…
Graph ClassificationGraph Neural NetworkGraph RegressionNode Classification