Advanced RL Snake Agent — Q‑Learning (Single Component)

A portfolio-ready, single-file React demo of tabular Q‑Learning training a Snake agent in real time. Includes canvas rendering, Tailwind UI, Lucide icons, Q‑value overlays, epsilon/phase stats, and performance metrics.

Run locally

npm install
npm run dev

Then open the printed local URL (usually http://localhost:5173).

Build

npm run build
npm run preview

Notes

• All RL state is in-memory only (no localStorage/sessionStorage).
• Game loop: 150 ms per step, auto-restart between episodes.
• Epsilon decays per episode; prevents immediate direction reversal; episode timeout to avoid loops.