AttentionRetriever: Attention Layers are Secretly Long Document Retrievers

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Fu, David Jiahao, Do, Lam Thanh, Li, Jiayu, Chang, Kevin Chen-Chuan
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911443857702912
author Fu, David Jiahao
Do, Lam Thanh
Li, Jiayu
Chang, Kevin Chen-Chuan
author_facet Fu, David Jiahao
Do, Lam Thanh
Li, Jiayu
Chang, Kevin Chen-Chuan
contents Retrieval augmented generation (RAG) has been widely adopted to help Large Language Models (LLMs) to process tasks involving long documents. However, existing retrieval models are not designed for long document retrieval and fail to address several key challenges of long document retrieval, including context-awareness, causal dependence, and scope of retrieval. In this paper, we proposed AttentionRetriever, a novel long document retrieval model that leverages attention mechanism and entity-based retrieval to build context-aware embeddings for long document and determine the scope of retrieval. With extensive experiments, we found AttentionRetriever is able to outperform existing retrieval models on long document retrieval datasets by a large margin while remaining as efficient as dense retrieval models.
format Preprint
id arxiv_https___arxiv_org_abs_2602_12278
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle AttentionRetriever: Attention Layers are Secretly Long Document Retrievers
Fu, David Jiahao
Do, Lam Thanh
Li, Jiayu
Chang, Kevin Chen-Chuan
Information Retrieval
Artificial Intelligence
Retrieval augmented generation (RAG) has been widely adopted to help Large Language Models (LLMs) to process tasks involving long documents. However, existing retrieval models are not designed for long document retrieval and fail to address several key challenges of long document retrieval, including context-awareness, causal dependence, and scope of retrieval. In this paper, we proposed AttentionRetriever, a novel long document retrieval model that leverages attention mechanism and entity-based retrieval to build context-aware embeddings for long document and determine the scope of retrieval. With extensive experiments, we found AttentionRetriever is able to outperform existing retrieval models on long document retrieval datasets by a large margin while remaining as efficient as dense retrieval models.
title AttentionRetriever: Attention Layers are Secretly Long Document Retrievers
topic Information Retrieval
Artificial Intelligence
url https://arxiv.org/abs/2602.12278