Phillip Isola, Sara Beery and Jeremy Bernstein
Deep Learning
Deep Learning — twenty-three lectures from MIT, Fall 2024.Deep Learning. Twenty-three lectures from MIT’s Fall 2024 course: training, architectures, generalization, representation learning, generative models, transfer and language models.
Lectures
Deep Learning · 23 lectures
1: Introduction to Deep Learning
2: How to Train a Neural Net
3: Approximation Theory
4: Architectures: Grids
5: Architectures: Graphs
6: Generalization Theory
7: Scaling Rules for Optimization
8: Architectures: Transformers
9: Hacker's Guide to Deep Learning
10: Architectures: Memory
11: Representation Learning: Reconstruction-Based
12: Representation Learning: Similarity-Based
13: Representation Learning: Theory
14: Generative Models: Basics
15: Generative Models and Representation Learning
16: Generative Models: Conditional Models
17: Generalization: Out-of-Distribution (OOD)
18: Transfer Learning: Models
19: Transfer Learning: Data
20: Scaling Laws
21: Language Models
23: Metrized Deep Learning
24: Inference Methods for Deep Learning