Near-optimal vector quantization from Google's ICLR 2026 paper — 95% recall, 5x compression, zero preprocessing, pure Python FAISS replacement

High-performance database management system

Hybrid B+ Tree and HNSW index for efficient k-NN search with scalar filtering using probabilistic optimization

Optimizing multi-attribute filtering for Approximate Nearest Neighbor (ANN) search using HNSW. This project integrates bitsets and Roaring Bitmaps into HNSWLib to accelerate query performance by reducing evaluation costs for large attribute datasets.

PostgreSQL TurboQuant Index for PGVector

Edit repository details Description Website Topics (separate with spaces) Include in the home page Releases Packages Deployments

Lightweight HNSW for experimenting with approximate nearest neighbors.

Vector Databases: Use Cases, Algorithms and Key Features

Transform-domain representation enabling 3–4× storage reduction with direct ANN search and novel multi-resolution signals. UK patent application under accelerated examination (Green Channel).