A Representation Sharpening Framework for Zero Shot Dense Retrieval

TL;DR

This paper introduces a training-free representation sharpening framework that enhances zero-shot dense retrieval by improving document differentiation, achieving state-of-the-art results across multiple datasets without additional inference costs.

Contribution

The paper presents a novel, training-free representation sharpening method that improves zero-shot dense retrieval performance and is compatible with existing approaches, setting new benchmarks.

Findings

Consistently outperforms traditional retrieval on over twenty datasets.

Achieves state-of-the-art results on the BRIGHT benchmark.

Provides an indexing-time approximation with no additional inference cost.

Abstract

Zero-shot dense retrieval is a challenging setting where a document corpus is provided without relevant queries, necessitating a reliance on pretrained dense retrievers (DRs). However, since these DRs are not trained on the target corpus, they struggle to represent semantic differences between similar documents. To address this failing, we introduce a training-free representation sharpening framework that augments a document's representation with information that helps differentiate it from similar documents in the corpus. On over twenty datasets spanning multiple languages, the representation sharpening framework proves consistently superior to traditional retrieval, setting a new state-of-the-art on the BRIGHT benchmark. We show that representation sharpening is compatible with prior approaches to zero-shot dense retrieval and consistently improves their performance. Finally, we address the performance-cost tradeoff presented by our framework and devise an indexing-time approximation that preserves the majority of our performance gains over traditional retrieval, yet suffers no additional inference-time cost.