Neural decoding of visual information in the zebrafish

A combination of theory and experiment has revealed some key principles of neural coding. Several aspects of neural development can be interpreted in terms of optimizing neural coding. Spontaneous activity may represent prior distributions, and prior distributions may be updated over development to facilitate Bayesian computations.In addition to simply refining, coding strategies may qualitatively change over development. Developing animals must begin to interact with the world before their neural development is complete. This means they must build neural codes appropriate for turning sensory inputs into motor outputs adaptively as their neural hardware matures. We review some recent progress in the understanding of the relationship between neural coding and neural circuit development. We focus particularly on neural coding in the context of topographic maps and spontaneous activity, as well as receptive field and circuit development, drawing on examples from both mammalian visual cortex and fish optic tectum. Overall we suggest that neural coding strategies during development may be highly dynamic. https://doi.org/10.1016/j.tins.2018.05.011 https://www.cell.com/trends/neurosciences/fulltext/S0166-2236(18)30158-9?_returnURL=https%3A%2F%2Flinkinghub.elsevier.com%2Fretrieve%2Fpii%2FS0166223618301589%3Fshowall%3Dtrue