rosikand / ml-concepts

📓 Short, atomic notes on concepts in machine learning.

📓 ml-concepts

Short, atomic notes on concepts in machine learning.

Contents

• Jax vmap
• Neural tangents
• multi-task-learning

To-do

• Domain adaption
• Self-supervised representation learning
• Gaussian processes
• Embedding space (how to learn one for e.g., both images and text)
• spurious correlations (robustness to distribution shifts)