SOURCE-LINKED INTELLIGENCE

When Can We Work in Embedding Space? What Text Embeddings Preserve

arXiv · AI, language, vision and robotics · article · Aug 31, 2026 · UTC

When do text embeddings work as inputs to empirical analysis? Their use rests on an assumption: that we can trade text for its low-dimensional embedding, and lose little in doing so. I make that assumption precise under a generative model in which documents are mixtures of latent topics. I study two uses---clustering units in embedding space and controlling for high-dimensional text. A cluster of embeddings is a set of documents with similar topic mixtures; controlling for the embedding is equivalent to controlling for the topic mixture, so validity reduces to whether that mixture captures the

Read original source ↗ Open in workspace

recordType

paper

region

Global