Language-AI Education Game
Raimi's AI Language Lab
A game-style web app where visitors find out how language AI learns by playing. With the museum character Raimi, they go through five touch mini-games, each covering one NLP concept: tokenization (slicing a sentence into pieces), embeddings (sorting words by meaning), word order, attention (spotting the word that matters), and next-word prediction with live probability bars.
It ran on 3 exhibition tablets from June to October 8, 2026, as an offline-capable PWA in Korean and English. Staff tracked per-device usage in a PIN-protected panel and exported reports to Excel.
Highlights
- 5 touch mini-games, one per NLP concept: tokenization, embeddings, word order, attention, and next-word prediction with live probability bars.
- Visitors play with the museum character Raimi through a stamp-based lesson menu.
- Runs entirely in the browser on plain JavaScript, with no framework or server.
- Works in Korean and English.
- Ran on 3 exhibition tablets from June to October 8, 2026, as an offline-capable PWA that reset to the start screen after 2 idle minutes.
- Staff tracked per-device usage in a PIN-protected panel and exported it to Excel.
Architecture
How the app is structured: screens, roles, and data flow.
This site keeps anonymous visit statistics. Privacy