---
title: Embedding Explorer
emoji: 📐
colorFrom: purple
colorTo: indigo
sdk: gradio
sdk_version: "6.5.1"
app_file: app.py
pinned: false
---

# Embedding Explorer

Interactive word vector visualization for the Responsible AI course at Huston-Tillotson University.

## Features

- **Explore** — Type words, see them in 3D space. Similar words cluster together. Click a word to see its nearest neighbors.
- **Vector Math** — Word arithmetic like `king - man + woman ≈ queen`. Results reveal hidden relationships encoded in the vectors.

## Configuration

All examples are configurable via HuggingFace Space environment variables (no code changes needed):

| Variable | Default | Description |
|----------|---------|-------------|
| `EXPLORE_EXAMPLES` | `["dog cat fish car truck", ...]` | JSON list of word groups for Explore tab |
| `ARITHMETIC_EXAMPLES` | `["king - man + woman", ...]` | JSON list of expressions for Vector Math tab |
| `N_NEIGHBORS` | `8` | Number of nearest neighbors to show on click |

## Model

Uses [GloVe](https://nlp.stanford.edu/projects/glove/) (Global Vectors for Word Representation) trained on Wikipedia + Gigaword — 400K words, 300 dimensions. Downloaded automatically on first startup via gensim (~376 MB, cached for subsequent runs).

## Course

Responsible AI: Technology, Power, and Justice (COSC-2300A)
Huston-Tillotson University, Spring 2026
Instructor: Chris Hyams