All Topics
Neural Nets & Embeddings
How neural networks represent meaning as embeddings, how backpropagation adjusts millions of weights, and why training needs far more memory than inference.
AI FoundationsAI
Videos (2)
Podcasts (1)
Slide decks (1)
Infographics (1)
Sources (9)
- web page Calculus on Computational Graphs: Backpropagation -- colah's blog
- web page Feature Visualization
- web page The Illustrated Word2vec – Jay Alammar – Visualizing machine learning one concept at a time.
- web page Understanding LSTM Networks -- colah's blog
- web page Visual Information Theory -- colah's blog
- web page Why Momentum Really Works
- web page [1301.3781] Efficient Estimation of Word Representations in Vector Space
- web page [1512.03385] Deep Residual Learning for Image Recognition
- web page [1607.06450] Layer Normalization