Adaptive Retrieval and Scalable Indexing for k-NN Search with Cross-Encoders

arXiv:2405.03651v1 Announce Type: cross
Abstract: Cross-encoder (CE) models which compute similarity by jointly encoding a query-item pair perform better than embedding-based models (dual-encoders) at estimating query-item relevance. Existing approaches perform k-NN search with CE by approximating the CE similarity with a vector embedding space fit either with dual-encoders (DE) or CUR matrix factorization. DE-based retrieve-and-rerank approaches suffer from poor recall on new domains and the retrieval with DE is decoupled from the CE. While CUR-based approaches can be more accurate …

abstract arxiv compute cs.ir cs.lg embedding encoder encoding indexing query retrieval scalable search space type vector