Sign inSign in
ConceptAI & ML

Weight decay

A regularization method that adds the squared norm of the weights to the in-sample error to penalize large weights and prevent overfitting.
as Yaser Abu-Mostafa defines itCaltechThe wording closest to all 3 definitions
Watch it explained

Know first

Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.

Opens up

Concepts that name this one as a prerequisite.

Lecturers

3

Each in their own terms, from the easiest level.

  • Phillip Isola

    MIT · Undergraduate

    An implicit or explicit regularizer that shrinks network weights toward zero during optimization.
  • Andrej Karpathy

    Graduate

    A regularization technique that adds a small penalty proportional to the weights' squared magnitude to the loss function to prevent overfitting.
    • 2:28:56Let's Reproduce GPT-2 (124M): Distributed PyTorch, FlashAttention & FineWeb

Lectures that assume it

1

Connections

14

The lectures that define it or assume it, and the concepts next to it.

3 lecturers. Definitions are written by AI from each lecture’s transcript. A time with a play mark starts the lecture at the moment the lecturer says it.