Clark Hash: Stateless Sparse Johnson-Lindenstrauss Quantization for Neural Embeddings

ArXi:2605.28034v1 Announce Type: new Clark Hash is a small method for storing neural embeddings in less space. It normalizes each database vector, applies a deterministic sparse signed Johnson-Lindenstrauss projection, clips the result, and s a fixed-width scalar-quantized code. Queries stay in floating point and are scored against the d sketches. In the default 384-dimensional sentence-embedding setting, Clark Hash s a cosine-search vector in 48 bytes instead of 1536 bytes for dense f32 storage. This is 32x smaller. The method does not need a.