📝 Notes

Short writings on research ideas, concepts, and reflections.

🧠 Representation Learning

Representation Learning: A Gentle Introduction

What does it mean for a model to "learn" a representation? This note explores the intuition behind representation learning.

📊 Log, Softmax & Likelihood

Why Log, Softmax & Likelihood — The Language of Every Loss Function

Every loss function in deep learning is secretly the same idea. This note covers probability, likelihood, softmax, cross-entropy, KL divergence, and the elegant gradient that ties them together.