Open access
Jun 2026
Aker: Density-Aware Approximate Caching for Vector Search
Aker is presented, an approximate cache for disk-based ANNS that implements a low-overhead refresh mechanism that bounds cache staleness while preserving search accuracy and integrates into pgvector and is evaluated on representative workloads.
Sukjoon Oh, Minki Kang, Dohyun Kim et al.
· Proceedings of the VLDB Endo... · 0 citations