Towards Better Understanding of Contrastive Learning

Towards Better Understanding of Contrastive Learning

DataLearning@ICL via YouTube Direct link

Assumptions

17 of 25

17 of 25

Assumptions

Class Central Classrooms beta

YouTube videos curated by Class Central.

Classroom Contents

Towards Better Understanding of Contrastive Learning

Automatically move to the next video in the Classroom when playback concludes

  1. 1 Intro
  2. 2 Great Empirical Success of Deep Models
  3. 3 Self-supervised Learning (SSL)
  4. 4 Contrastive Learning (CL)
  5. 5 Formulation of Contrastive Learning
  6. 6 Understanding of Contrastive Loss
  7. 7 What Deep Learning Brings?
  8. 8 Example: InfoNCE
  9. 9 Coordinate-wise Optimization
  10. 10 A Surprising Connection to Kernels
  11. 11 Overview of Nonlinear Analysis
  12. 12 Nonlinear Setting
  13. 13 Training Dynamics
  14. 14 1-layer 1-node nonlinear network
  15. 15 How to reduce the local roughness p(w)?
  16. 16 1-layer multiple node nonlinear network
  17. 17 Assumptions
  18. 18 Conditional Independence
  19. 19 What linear network cannot do
  20. 20 Global modulation
  21. 21 Feature Emergence
  22. 22 Experiment Setting
  23. 23 Model Architecture & Evaluation Metric
  24. 24 Visualization
  25. 25 Quadratic Loss versus InfoNCE

Never Stop Learning.

Get personalized course recommendations, track subjects and courses with reminders, and more.

Someone learning on their laptop while sitting on the floor.