# Hello Deep Learning: Automatic differentiation, autograd

DevFeed: [Hello Deep Learning: Automatic differentiation, autograd](<https://devfeed.tech/articles/hello-deep-learning-automatic-differentiation-autograd-36257.md>)

Original publisher: [Read original article](<https://berthub.eu/articles/posts/autograd/>)

Published: 2023-03-30T10:00:03Z

Content type: tutorial

Language: en

Sources: [Bert Hubert's writings](<https://devfeed.tech/sources/bert-hubert-s-writings.md>)

Topics: [Deep learning](<https://devfeed.tech/topics/deep-learning.md>), [Neural Network](<https://devfeed.tech/topics/neural-network.md>), [Algorithm](<https://devfeed.tech/topics/algorithm.md>), [Code](<https://devfeed.tech/topics/code.md>)

Tags: [algorithm](<https://devfeed.tech/tags/algorithm.md>), [deep-learning](<https://devfeed.tech/tags/deep-learning.md>), [matrices](<https://devfeed.tech/tags/matrices.md>), [model](<https://devfeed.tech/tags/model.md>), [neural-network](<https://devfeed.tech/tags/neural-network.md>), [pixels](<https://devfeed.tech/tags/pixels.md>), [train](<https://devfeed.tech/tags/train.md>)

## AI overview

This tutorial chapter explains how automatic differentiation and gradient descent train a five-layer neural network to recognize all ten handwritten digits. It introduces the network's layers, parameters, weights, and biases, then connects parameter updates to derivatives of the error.

## Source excerpt

This page is part of the Hello Deep Learning series of blog posts. You are very welcome to improve this page via GitHub! In the previous chapter we configured a neural network and made it learn to distinguish between the digits 3 and 7. The learning turned out to consist of "twisting the knobs in the right direction". Although simplistic, the results were pretty impressive. But, you might still be a bit underwhelmed - the network only distinguished between two digits.