Log-Likelihood From First Principles
April 13, 2026
Building log-likelihood from joint probability through likelihood, cross-entropy, softmax, and LLM training — plus numerical stability and why the same objective can drive overconfidence.
Blog
Writing, notes, and technical references I keep coming back to.
Writing Gradient Descent From Scratch
March 16, 2026
A ground-up look at gradient descent — what it is, why it matters, and how to implement it in plain Python without any libraries.