A Gaussian membership function is a type of function used in fuzzy logic to represent the degree of truth of a variable, defined by a symmetric bell-shaped curve. It is characterized by its mean and standard deviation, which determine the center and width of the curve, respectively, allowing for smooth transitions between membership grades.