3D Embedding Visualizations

Text embeddings are dense vector representations that capture the semantic meaning of words and sentences. By mapping similar concepts to nearby points in high-dimensional space, they enable machines to understand relationships between different pieces of text.

This interactive visualization uses state-of-the-art transformer models to generate embeddings from your text, then applies UMAP (Uniform Manifold Approximation and Projection) to reduce the dimensionality down to 3D space while preserving the semantic relationships.

How it works:

• 🤖 AI-powered embeddings: Uses Snowflake’s Arctic Embed XS model to convert text into semantic vectors
• 📐 Dimensionality reduction: UMAP algorithm intelligently compresses embeddings to 3D while maintaining semantic clustering
• 🎮 Interactive 3D space: Navigate the embedding space with mouse controls - drag to rotate, scroll to zoom
• 🔍 Semantic exploration: Texts with similar meanings appear closer together in the 3D space
• 🔒 Private by design: The visualization loads entirely in your browser using WebGL and WebAssembly - no server required. All AI model inference happens locally, keeping your text private

Try adding your own texts to see how different concepts cluster together. You might discover interesting semantic relationships that weren’t immediately obvious!