A multilayer perceptron (MLP) is a feedforward neural network composed of an input layer, one or more hidden layers of fully connected neurons with nonlinear activations, and an output layer, trained by backpropagation. It is a canonical universal function approximator underlying deeper architectures, and is frequently used as a lightweight decoder or coordinate-based function in implicit neural representations. NeRF and related implicit neural representation methods use an MLP to map spatial coordinates directly to colour and density values.