https://nalar.dev/tags/representation-geometry/ Recent content in Representation Geometry on Nalar Hugo en-us Wed, 16 Sep 2026 00:00:00 +0700 https://nalar.dev/measure-embedding-anisotropy-before-vector-retrieval/ Wed, 16 Sep 2026 00:00:00 +0700 https://nalar.dev/measure-embedding-anisotropy-before-vector-retrieval/ <p>Cosine similarity is often treated as a local comparison between one query embedding and one candidate. That interpretation becomes less informative when most vectors occupy a narrow set of directions. Unrelated items can then share a substantial common component, compressing the range of angles that retrieval uses to separate candidates.</p> <p>This directional concentration is commonly described as embedding anisotropy. It is a property of a vector distribution, not a defect implied by any single similarity score. For developers, the practical issue is that a fixed cosine value has no universal meaning. Its usefulness depends partly on the geometry of the embedding population in which it was produced.</p>