The implications of categorical and category-free mixed selectivity on representational geometries - ScienceDirect

• Mixed selectivity occurs when a neuron's activity is modulated by multiple task variables. • Mixed selectivity is frequently identified, but rarely characterized fully. • Nonlinear selectivity enables more powerful coding at the cost of generalization and noise sensitivity. • Neurons can exhibit mixed selectivity while forming categories or a continuum. • Representational geometry provides a powerful toolkit for understanding mixed selectivity. Mixed selectivity occurs when a neuron's activity is modulated by multiple task variables. Mixed selectivity is frequently identified, but rarely characterized fully. Nonlinear selectivity enables more powerful coding at the cost of generalization and noise sensitivity. Neurons can exhibit mixed selectivity while forming categories or a continuum. Representational geometry provides a powerful toolkit for understanding mixed selectivity. The firing rates of individual neurons displaying mixed selectivity are modulated by multiple task

• Mixed selectivity occurs when a neuron's activity is modulated by multiple task variables. • Mixed selectivity is frequently identified, but rarely characterized fully. • Nonlinear selectivity enables more powerful coding at the cost of generalization and noise sensitivity. • Neurons can exhibit mixed selectivity while forming categories or a continuum. • Representational geometry provides a powerful toolkit for understanding mixed selectivity. Mixed selectivity occurs when a neuron's activity is modulated by multiple task variables. Mixed selectivity is frequently identified, but rarely cha