// ⬡ Deep Learning Foundations

Backpropagation

Computes every parameter's gradient by applying the chain rule backward.

反向传播 1986 Foundational core

IN PLAIN WORDS

After the forward pass, propagate the error from the last layer backward with the chain rule, getting all gradients in one sweep. PyTorch autograd is its general implementation.

Used in: Autograd in every deep learning framework

TIMELINE · prerequisites → this → what builds on it

1951 Stochastic Gr…1958 Perceptron1986 Backpropagation1997 Long Short-Te…1998 Convolutional…2006 Autoencoder2010 Weight Initia…2012 Dropout

01Original paper

Learning representations by back-propagating errors

Rumelhart et al. · Nature · 1986

DOI ↗ ✓ Verified · Crossref API doi:10.1038/323533a0

03Learning resources

Official docs, explainers, courses and reference code. All links checked.

04Where the course teaches it

U12 Calculus & Optimization Derivatives, partial derivatives, the chain rule and gradients — “walking downhill along the gradient”. 12 h2 algorithms1 papers U14 PyTorch & Autograd Tensors, autograd, data loading and the training loop. 10 h2 algorithms0 papers U30 Neural Networks & Backpropagation From neurons and multilayer perceptrons to backpropagation and activation functions. 10 h4 algorithms2 paperslesson ready

On learning paths:Zero to Deep Learning

For AI / data

Verification: verified, checked 2026-10-21.
JSON: /api/learn/algorithms.json (id: backprop)

Backpropagation

01Original paper

02Where it sits

Learn first

Then you can learn · 14

03Learning resources

Explainers

Reference code

Video

Courses

04Where the course teaches it