The article "Mixed Selectivity: Cellular Computations for Complexity" by Kay M. Tye, Earl K. Miller, Felix H. Taschbach, Marcus K. Benna, Mattia Rigotti, and Stefano Fusi explores the computational and biological mechanisms underlying mixed selectivity in neural circuits. Mixed selectivity refers to the ability of neurons to respond to multiple, statistically independent variables, which enhances the brain's computational power and flexibility. The authors define pure, linear, and nonlinear mixed selectivity and discuss how these response properties can be achieved in simple neural circuits. Linear mixed selectivity involves a weighted sum of responses to different variables, while nonlinear mixed selectivity involves a nonlinearity in the response function. Nonlinear mixed selectivity allows for high-dimensional representations, providing significant flexibility to downstream readouts. However, the brain cannot encode all possible mixtures of variables simultaneously due to combinatorial complexity, so gating mechanisms like oscillations and neuromodulation are crucial for dynamically selecting which variables are mixed and transmitted to the readout. The article also highlights the importance of mixed selectivity in various brain regions and functions, from high-level cognition to sensorimotor processes, and discusses the role of neuromodulation in tuning ensemble volume and orchestrating mixed selectivity.The article "Mixed Selectivity: Cellular Computations for Complexity" by Kay M. Tye, Earl K. Miller, Felix H. Taschbach, Marcus K. Benna, Mattia Rigotti, and Stefano Fusi explores the computational and biological mechanisms underlying mixed selectivity in neural circuits. Mixed selectivity refers to the ability of neurons to respond to multiple, statistically independent variables, which enhances the brain's computational power and flexibility. The authors define pure, linear, and nonlinear mixed selectivity and discuss how these response properties can be achieved in simple neural circuits. Linear mixed selectivity involves a weighted sum of responses to different variables, while nonlinear mixed selectivity involves a nonlinearity in the response function. Nonlinear mixed selectivity allows for high-dimensional representations, providing significant flexibility to downstream readouts. However, the brain cannot encode all possible mixtures of variables simultaneously due to combinatorial complexity, so gating mechanisms like oscillations and neuromodulation are crucial for dynamically selecting which variables are mixed and transmitted to the readout. The article also highlights the importance of mixed selectivity in various brain regions and functions, from high-level cognition to sensorimotor processes, and discusses the role of neuromodulation in tuning ensemble volume and orchestrating mixed selectivity.