SemanticGuessr: a Wordle-style game teaching vector embeddings on a 43K-word map
Neon0asis · reddit · 2026-10-15
Isaacus ML engineer Abdur-Rahman Butler built an interactive write-up on vector embeddings, covering vectorization, cosine similarity, semantic search, hierarchical retrieval, clustering, zero-shot classification and deduplication, with Python snippets.
The centerpiece, SemanticGuessr (Wordle × GeoGuessr), has players triangulate a random target word on a 2D map of 43,339 common English words via vector math like london - english + french and 0.4 noodles + 0.6 pasta. He cut embeddings to 512 dims with 4-bit quantization, shrinking the payload from 311 MB to 12.3 MB with only 6% precision loss — all running statically in the browser.
More from Apps
- Boba, a 'social intelligence' AI, wins fans for actually being funny — gharik · 2026-10-23
- Eric Zelikman unveils social AI 'Boba', says it learned interaction 'from millennia of dreams' — gharik · 2026-10-23
- Boba team celebrates app launch, invites users to try it in group chats — gharik · 2026-10-23
- Account hacks surge: 6-step emergency checklist plus an AI audit prompt — petergyang · 2026-10-23
- 30 websites to download research papers for free, from arXiv to Sci-Hub — Faheem_uh · 2026-10-23
- Discord now renders Markdown file previews properly, a boon for AI workflows — The Verge AI · 2026-10-23