Introduction to Deep Learning · HIT

Week 12   Part III · Architectures & Representation Learning

Representation Learning

Autoencoders and latent representations; contrastive and self-supervised methods.

Curated, free, canonical references for this week: a course or lecture, a book chapter, a video, and an authoritative blog post or official tutorial. Each opens in a new tab.

Course
Stanford CS231n Lecture: Self-Supervised Learning youtube.com

Stanford lecture on pretext tasks, contrastive learning, and self-supervised representations.

Book
Goodfellow, Bengio & Courville, Deep Learning, Chapter 14: Autoencoders deeplearningbook.org

The standard free chapter on autoencoders and their role in learning latent representations.

Video
SimCLR: Contrastive Learning of Visual Representations (illustrated) youtube.com

A focused walkthrough of SimCLR: augmentation views, the contrastive loss, and the projection head.

Blog / Docs
Lil'Log: Contrastive Representation Learning lilianweng.github.io

A deeply-cited survey of contrastive and self-supervised methods (SimCLR, MoCo, BYOL, SwAV).

← Back to the Week 12 lab

PreviousWeek 11: LSTMs, GRUs & Sequence TasksNextWeek 13: Integration & Transfer Learning