Seonglae Cho- Collect data
- Build co-occurrence matrix
- Sparsify the co-occurrence matrix into a k-nearest-neighbors graph
- Embed that into an intermediate high(er)-dimensional embedding with PyMDE
- Project that intermediate embedding down into 2D using UMAP or t-SNE
- Render the result using WebGL or WebGPU
What I've Learned Building Interactive Embedding Visualizations
Over the past few years, I've built several different interactive embedding visualizations. I have a big interest in exploring and understanding things in a free-form manner, and these kinds of tools feel to me like an extremely effective way of facilitating that sort of experience.
https://cprimozic.net/blog/building-embedding-visualizations-from-user-profiles/?utm_source=tldrnewsletter
Visualizing High-Dimensional Space by Daniel Smilkov, Fernanda Viégas, Martin Wattenberg & the Big Picture team at Google - Experiments with Google
Since 2009, coders have created thousands of amazing experiments using Chrome, Android, AI, WebVR, AR and more. We're showcasing projects here, along with helpful tools and resources, to inspire others to create new experiments.
https://experiments.withgoogle.com/visualizing-high-dimensional-space
Visualize your RAG Data — EDA for Retrieval-Augmented Generation
How to use UMAP dimensionality reduction for Embeddings to show Questions, Answers and their relationships to source documents with…
https://itnext.io/visualize-your-rag-data-eda-for-retrieval-augmented-generation-0701ee98768f
