# Query Embedding Pruning for Dense Retrieval

arXiv: https://arxiv.org/abs/2108.10341 · HTML: https://mlsanity.com/paper/2108.10341/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 23 Aug 2021.

## Built on by (most cited first, 5 of 5)

- [PLAID: An Efficient Engine for Late Interaction Retrieval](https://mlsanity.com/paper/2205.09707/) (2022)
- [Introducing Neural Bag of Whole-Words with ColBERTer: Contextualized Late Interactions using Enhanced Reduction](https://mlsanity.com/paper/2203.13088/) (2022)
- [Towards Sparsification of Graph Neural Networks](https://mlsanity.com/paper/2209.04766/) (2022)
- [A Study on Token Pruning for ColBERT](https://mlsanity.com/paper/2112.06540/) (2021)
- [CAME: Competitively Learning a Mixture-of-Experts Model for First-stage Retrieval](https://mlsanity.com/paper/2311.02834/) (2023)

## Abstract (arXiv)

Recent advances in dense retrieval techniques have offered the promise of being able not just to re-rank documents using contextualised language models such as BERT, but also to use such models to identify documents from the collection in the first place. However, when using dense retrieval approaches that use multiple embedded representations for each query, a large number of documents can be retrieved for each query, hindering the efficiency of the method. Hence, this work is the first to consider efficiency improvements in the context of a dense retrieval approach (namely ColBERT), by pruning query term embeddings that are estimated not to be useful for retrieving relevant documents. Our proposed query embeddings pruning reduces the cost of the dense retrieval operation, as well as reducing the number of documents that are retrieved and hence require to be fully scored. Experiments conducted on the MSMARCO passage ranking corpus demonstrate that, when reducing the number of query embeddings used from 32 to 3 based on the collection frequency of the corresponding tokens, query embedding pruning results in no statistically significant differences in effectiveness, while reducing the number of documents retrieved by 70%. In terms of mean response time for the end-to-end to end system, this results in a 2.65x speedup.