ClassroomThe classroom
Back to the lesson libraryMECHANISM · 11 MIN
M11.4 CONNECT THE MECHANISM

How an error reaches earlier weights

A network misses its target by 2. Which of its weights is to blame, and by how much? Trace the backward sweep that answers that for every weight at once.

LESSON OVERVIEW11 min lesson

Lesson overview

A network misses its target by 2. Which of its weights is to blame, and by how much? Trace the backward sweep that answers that for every weight at once.

What you’ll explore

  • Trace a forward loss, multiply local derivatives backward, and add contributions when a parameter affects more than one branch.
GO TO THE SOURCE

Original explanations, connected to the research.

Dive into Deep Learning — forward/backward computationDive into Deep Learning — calculusRumelhart, Hinton & Williams — Learning representations by back-propagating errors (Nature, 1986)
Suggest a correction

A precise note can make an explanation better.

Choose the scene and describe what needs attention. Download a feedback file to share through a channel you already use. This page does not send feedback or connect you with a reviewer.

The file includes this note, the scene title, and lesson metadata. Your saved progress and quiz responses are excluded. Download before leaving or reloading to keep your note.