Blog

A conceptual guide to linear algebra built around vector spaces, linear transformations, matrices, bases, and coordinate systems.

1

A visual explanation of cross-attention and how it connects encoder and decoder stacks in the original Transformer architecture.

2

A first-principles explanation of why magnetic field lines run south to north inside a bar magnet and form closed loops.

3

A visual guide to encoder-only Transformers and BERT, covering tokenization, embeddings, attention, and classification heads.

4

Why implicit truthy and falsy checks can obscure intent, hide type errors, and make production code harder to reason about.

5

A visual and intuitive introduction to decoder-only Transformer architecture, including embeddings, attention, and language modeling.

6

How to build meaningful evaluations for AI app-building agents using prompt sets, code-based scorers, and automated analysis.

7