KnowraBackpropagationLinked fromLinked fromThe 13 pages that link to Backpropagation, each with the reason it gives.All 13Broader topic 1Related 10Compared with 2Jacobian matrixRelated: Layerwise Jacobians combine through the chain rule to calculate neural-network gradients.Chain ruleBroader topic: Each layer passes gradients to earlier layers using local chain-rule derivatives.ConnectionismRelated: It made multilayer networks trainable for tasks used to model cognition.Gradient descentRelated: It supplies neural-network gradients that gradient descent uses to update weights.Neural networkRelated: It supplies gradients used to adjust network parameters during training.Deep learningRelated: Its adoption made gradient-based training of multilayer networks practical.Convolutional neural networkRelated: It supplies gradients used to adjust convolutional filters during training.Large language modelRelated: Training uses it to adjust model parameters after measuring prediction errors.Total derivativeRelated: It efficiently applies chain-rule derivatives through many dependent computational variables.Adjoint operatorRelated: Reverse-mode differentiation propagates covectors through transposed Jacobians, the finite-dimensional adjoint action.Hebbian theoryCompared with: It illustrates a powerful error-driven alternative to local correlation-based learning.Geoffrey HintonRelated: Hinton helped make this method central to training multilayer neural networks.Donald O. HebbCompared with: Unlike local Hebbian rules, it uses error signals propagated through a network.