Paths as Units of Evolution in the Brain

publication Evidence: moderate

Author Information

Author(s): Fernando Chrisantha, Vasas Vera, Szathmáry Eörs, Husbands Phil

Primary Institution: Department of Informatics, University of Sussex, Brighton, United Kingdom

Can paths of activity through neuronal networks serve as a basis for unlimited hereditary variation?

The study demonstrates that paths in neuronal networks can act as hereditary substrates for evolution without explicit replication.

Paths in a network can exhibit unlimited hereditary variation.
Path evolution algorithms can outperform traditional genetic algorithms in certain optimization tasks.
Paths can maintain memory of past environments, aiding in adaptation.

This study shows that the brain can use paths of neuron activity to evolve and adapt, just like living things do.

The study uses a path evolution algorithm to demonstrate how paths in a neuronal network can evolve through natural selection.

The model may not fully represent the complexity of real neuronal networks and lacks recurrent connections.

Access the complete publication on the publisher's website