Neo4j

Where sparse (BM25/SPLADE) retrieval reliably beats dense embeddings

Contextual retrieval prepends generated context to each chunk before embedding — what chunking failure does that fix?

Contextual retrieval, in one move at ingestion time

Why does embedding quality plateau as you increase dimensionality?

Pick an embedding dimension for a 50M-doc index with a 10 GB RAM budget

Define an embedding in terms of what it IS, not what it does

When would you reach for a cross-encoder over a bi-encoder?

Compute the cost to embed 1 billion tokens with text-embedding-3-small

When do cosine and Euclidean rankings DIVERGE for the same query and corpus?

Why does contrastive training produce useful embedding geometry?