SLM Engineering — Interactive Visualizations

Twenty interactive visualizations on engineering Small Language Models (under 5B parameters) for edge and domain deployment — every million parameters, every megabyte, every inference-joule counts. Each demo recomputes live in the browser from the identical formula in its source.

RESEARCH.md · the 20-bundle master plan (20/20) ↑ all tutorials

Every card opens a live .html demo recomputed in your browser; the md chip opens the guide and the py chip the source on GitHub. Sibling of ../llm/ (large models) and ../local-llm/ (local inference). Tailwind Play CDN needs internet to style.

The front door of the slm-engineering/ visualization suite. Each card click opens the static .html demo (runs live in the browser); the md chip links the write-up guide and the py chip the canonical source on GitHub. Progress persists locally under localStorage["slm-done"].