ColBERTSaR: Sparsified ColBERT Index via Product Quantization

• URL: http://arxiv.org/abs/2606.05568v1
• Date: Thu, 04 Jun 2026 01:28:45 GMT
• Title: ColBERTSaR: Sparsified ColBERT Index via Product Quantization
• Abstract summary: ColBERT is an effective neural retrieval architecture, but it requires a heavy index structure to support candidate set retrieval.<n>We propose an embedding quantization approach that turns a ColBERT index into a true inverted index.
• Score: 23.59576017606267
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: While ColBERT is an effective neural retrieval architecture, it requires a heavy index structure to support candidate set retrieval based on approximated token embeddings, gathering and decompressing document token embeddings, and applying the MaxSim operation. Indexes in PLAID and similar ColBERT implementations require five to ten times the disk storage of the original raw text, which limits their scalability. Furthermore, prior work has identified that the gathering and decompression stages are the primary inefficiencies at query time. Limiting the number of document tokens that must be gathered by thresholding and score approximation does not eliminate the need for the entire index to support ad hoc queries. In this work, we propose an embedding quantization approach that turns a ColBERT index into a true inverted index. We show that, theoretically, ColBERT with embedding quantization is equivalent to learned-sparse retrieval except for the scoring mechanism. Empirically, we demonstrate that our index is 50-70% smaller than a one-bit PLAID index while retaining retrieval effectiveness.

Related papers

• A Parametric Memory Head for Continual Generative Retrieval [52.66674234249913]
  Generative information retrieval (GenIR) consolidates retrieval into a single neural model that decodes document identifiers (docids) directly from queries.<n>We show that sequential adaptation improves retrieval on newly added documents but substantially degrades performance on earlier slices.<n>We propose post-adaptation memory tuning (PAMT), a memory-only stabilization stage that augments an adapted model with a modular parametric memory head.
  arXiv  Detail & Related papers  (2026-04-25T17:38:51Z)
• HISA: Efficient Hierarchical Indexing for Fine-Grained Sparse Attention [62.79085204939384]
  HISA (Hierarchical Indexed Sparse Attention) rewrites the search path from a flat token scan into a two-stage hierarchical procedure.<n>On kernel-level benchmarks, HISA achieves up to speedup at 64K context.
  arXiv  Detail & Related papers  (2026-03-30T13:59:51Z)
• Multi-Vector Index Compression in Any Modality [73.7330345057813]
  Late interaction has emerged as a dominant paradigm for information retrieval in text, images, visual documents, and videos.<n>We introduce four approaches for index compression: sequence resizing, memory tokens, hierarchical pooling, and a novel attention-guided clustering (AGC)<n>AGC uses an attention-guided mechanism to identify the most semantically salient regions of a document as cluster centroids and to weight token aggregation.
  arXiv  Detail & Related papers  (2026-02-24T18:57:33Z)
• Forward Index Compression for Learned Sparse Retrieval [15.629655228398567]
  We focus on the size of a data structure that is common to all algorithmic flavors and that constitutes a substantial fraction of the overall index size: the forward index.<n>In particular, we seek compression techniques to reduce the storage footprint of the forward index without compromising search quality or inner product computation latency.
  arXiv  Detail & Related papers  (2026-02-05T08:35:17Z)
• Adaptive Retrieval and Scalable Indexing for k-NN Search with Cross-Encoders [77.84801537608651]
  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. We propose a sparse-matrix factorization based method that efficiently computes latent query and item embeddings to approximate CE scores and performs k-NN search with the approximate CE similarity.
  arXiv  Detail & Related papers  (2024-05-06T17:14:34Z)
• Semi-Parametric Retrieval via Binary Bag-of-Tokens Index [71.78109794895065]
  SemI-parametric Disentangled Retrieval (SiDR) is a bi-encoder retrieval framework that decouples retrieval index from neural parameters.<n>SiDR supports a non-parametric tokenization index for search, achieving BM25-like indexing complexity with significantly better effectiveness.
  arXiv  Detail & Related papers  (2024-05-03T08:34:13Z)
• Faster Learned Sparse Retrieval with Block-Max Pruning [11.080810272211906]
  This paper introduces Block-Max Pruning (BMP), an innovative dynamic pruning strategy tailored for indexes arising in learned sparse retrieval environments. BMP substantially outperforms existing dynamic pruning strategies, offering unparalleled efficiency in safe retrieval contexts.
  arXiv  Detail & Related papers  (2024-05-02T09:26:30Z)
• A Learned Index for Exact Similarity Search in Metric Spaces [25.330353637669386]
  LIMS is proposed to use data clustering and pivot-based data transformation techniques to build learned indexes. Machine learning models are developed to approximate the position of each data record on the disk. Extensive experiments on real-world and synthetic datasets demonstrate the superiority of LIMS compared with traditional indexes.
  arXiv  Detail & Related papers  (2022-04-21T11:24:55Z)
• A Study on Token Pruning for ColBERT [0.7646713951724011]
  The ColBERT model has recently been proposed as an effective BERT based ranker. The big downside of the model is the index size, which scales linearly with the number of tokens in the collection. In this paper, we study various designs for ColBERT models in order to attack this problem.
  arXiv  Detail & Related papers  (2021-12-13T10:24:54Z)
• Progressively Pretrained Dense Corpus Index for Open-Domain Question Answering [87.32442219333046]
  We propose a simple and resource-efficient method to pretrain the paragraph encoder. Our method outperforms an existing dense retrieval method that uses 7 times more computational resources for pretraining.
  arXiv  Detail & Related papers  (2020-04-30T18:09:50Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.