The information content of receptive fields

Thomas L. Adelman, William Bialek, Robert M. Olberg

Research output: Contribution to journalArticlepeer-review

50 Scopus citations


The nervous system must observe a complex world and produce appropriate, sometimes complex, behavioral responses. In contrast to this complexity, neural responses are often characterized through very simple descriptions such as receptive fields or tuning curves. Do these characterizations adequately reflect the true dimensionality reduction that takes place in the nervous system, or are they merely convenient oversimplifications? Here we address this question for the target-selective descending neurons (TSDNs) of the dragonfly. Using extracellular multielectrode recordings of a population of TSDNs, we quantify the completeness of the receptive field description of these cells and conclude that the information in independent instantaneous position and velocity receptive fields accounts for 70%-90% of the total information in single spikes. Thus, we demonstrate that this simple receptive field model is close to a complete description of the features in the stimulus that evoke TSDN response.

Original languageEnglish (US)
Pages (from-to)823-833
Number of pages11
Issue number4
StatePublished - Nov 13 2003

All Science Journal Classification (ASJC) codes

  • General Neuroscience


Dive into the research topics of 'The information content of receptive fields'. Together they form a unique fingerprint.

Cite this