Anchor Divergences Enable Context-Dependent Semantic Similarity from Fixed Embeddings

Victor Veitch and collaborators published a paper introducing Anchor Divergences, a method that allows vector similarity to vary with context without changing the embeddings, enabling context-dependent semantic retrieval.

2026-10-19 ~ 2026-10-20 · 2 related posts

• Anchor Divergence paper: contextual semantic similarity on fixed embeddings — victorveitch · 2026-10-19
• Anchor Divergences: contextual semantic similarity for vector search without changing embeddings — victorveitch · 2026-10-20