paper-with-me

홈 › Papers

Lost in a Single Vector: Improving Long-Document Retrieval with Chunk Evidence Aggregation

2026-06-17 · Shanshan Lyu, Yiwei Wang, Yujun Cai, Jiafeng Guo, Shenghua Liu arxiv

Dense retrieval ranks one query vector against one document vector. On long documents, this interface can fail when a short but decisive span is weakened during document encoding before ranking. We study this failure mode as document-side early compression and introduce the Evidence Dilution Index (EDI) to measure how far a document-level representation falls below the strongest chunk-level evidence within the same gold document. Guided by this view, we propose DICE (Document Inference via Chunk Evidence), a training-free document-side strategy that splits documents into chunks, encodes them independently with a frozen model, and aggregates them back into a single vector while preserving the standard one-query-one-document interface. On LongEmbed, DICE improves retrieval across four backbones, with the largest gains on slices beyond 4k tokens: for Dream, Passkey >4k rises from 30.0 to 90.0 and Needle >4k from 23.3 to 74.0. Across 12,779 filtered samples, DICE yields lower EDI than the single-vector baseline in 92.8% of cases. These results establish document-level encoding as a practical and underexplored lever for long-document retrieval.

📄 PDF Abstract BibTeX arXiv:2606.18781

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Hierarchical Retrieval: The Geometry and a Pretrain-Finetune Recipe

2025-09-19 · Chong You, Rajesh Jayaram, Ananda Theertha Suresh, Robin Nittka 외 arxiv

Dual encoder (DE) models, where a pair of matching query and document are embedded into similar vector representations, are widely used in information retrieval due to their simplicity and scalability. However, the Eucli…

Information Retrieval

BERAG: Bayesian Ensemble Retrieval-Augmented Generation for Knowledge-based Visual Question Answering

2026-04-24 · Jinghong Chen, Jingbiao Mei, Guangyu Yang, Bill Byrne arxiv

A common approach to question answering with retrieval-augmented generation (RAG) is to concatenate documents into a single context and pass it to a language model to generate an answer. While simple, this strategy can o…

Visual Question Answering

P-SIF: Document Embeddings Using Partition Averaging

2020-05-18 · Vivek Gupta, Ankit Saw, Pegah Nokhiz, Praneeth Netrapalli 외

Simple weighted averaging of word vectors often yields effective representations for sentences which outperform sophisticated seq2seq neural models in many tasks. While it is desirable to use the same method to represent…

Retrieval Quality at Context Limit

2025-11-08 · Max McKinnon arxiv

The ability of large language models (LLMs) to recall and retrieve information from long contexts is critical for many real-world applications. Prior work (Liu et al., 2023) reported that LLMs suffer significant drops in…

Hybrid-Vector Retrieval for Visually Rich Documents: Combining Single-Vector Efficiency and Multi-Vector Accuracy

2025-10-25 · Juyeon Kim, Geon Lee, Dongwon Choi, Taeuk Kim 외 arxiv

Retrieval over visually rich documents is essential for tasks such as legal discovery, scientific search, and enterprise knowledge management. Existing approaches fall into two paradigms: single-vector retrieval, which i…