For much of the history of modern neuroscience, it has been a assumed that the neuron is the functional unit of the brain.  But now there is increasing evidence that ensembles of neurons, not individuals, are the functional units.  One line of evidence is that many neurons in higher cortical areas have “mixed selectivity” , responses to diverse combinations of variables; they don’t signal one “message”.  Thus, their activity only makes sense when simultaneously considering the activity of other neurons.  In fact, we (Rigotti et al., 2013; Fusi et al., 2016) have shown that mixed selectivity gives the brain the computational horsepower needed for complex behavior.

In this paper, Dehaqani et al show that simultaneously recorded prefrontal cortex neurons have high-dimensional, mixed-selectivity, representations and convey more information as a population than even individuals.  This was especially true for parts of visual space that were weakly encoded by single neurons.  Less-informative neurons were recruited into ensemble to fully encode visual space.

Prefrontal neurons expand their representation of space by increase in dimensionality and decrease in noise correlation.  Mohammad-Reza Dehaqani, Abdol-Hossein Vahabie, Mohammadbagher Parsa, Behrad Noudoost, Alireza Soltani
doi: http://dx.doi.org/10.1101/065581

Further reading:
Fusi, S., Miller, E.K., and Rigotti, M. (2016) Why neurons mix: High dimensionality for higher cognition.  Current Opinion in Neurobiology. 37:66-74  doi:10.1016/j.conb.2016.01.010. View PDF »

Rigotti, M., Barak, O., Warden, M.R., Wang, X., Daw, N.D., Miller, E.K., & Fusi, S. (2013) The importance of mixed selectivity in complex cognitive tasks. Nature, 497, 585-590, doi:10.1038/nature12160. View PDF »

Yuste, Rafael. “From the neuron doctrine to neural networks.” Nature Reviews Neuroscience 16.8 (2015): 487-497.

About the Author


The Miller Lab uses experimental and theoretical approaches to study the neural basis of the high-level cognitive functions that underlie complex goal-directed behavior. ekmillerlab.mit.edu