paulopontesm/how-does-a-neural-net-really-work

docs(notebook): move BCE/ROC walkthrough to section 4

Clarify notebook narrative and make results commentary run-agnostic

Fix validation leakage and sync regenerated notebook HTML

Use trained linear weights for evaluation and rebuild HTML

Migrate project workflow from pip/venv to uv

add nn part

pin dependency versions and rever plotly version to fix render

render some widgets on html

add 3d plot for error surface

add widget state

change html template

add first version of notebook

add gitignore

initial commit