Skip to content

Page-Aware Retrieval-Augmented Generation for EvalLLM 2026: A Five-Variant Study on French PDFs

Sep 2026 · arXiv.org · Vol abs/2609.34776 · 0 citations · 13 references
Computer Science

TL;DR

Correct page selection, rather than semantic retrieval in isolation, is identified as the main opportunity for improvement in this setting of retrieval-augmented generation, with higher latency and more detected output artifacts.

Abstract

We study retrieval-augmented generation (RAG) for questions about French PDF documents when both the answer and its supporting document pages are evaluated. Five system variants add dense retrieval, rank fusion, reranking, and query decomposition to a BM25 baseline. On 595 challenge questions, the complete system scores 0.4450 MRR@10 and 0.4013 Recall@10, compared with 0.3430 and 0.2994 for BM25. Dense retrieval alone and a simple lexical--dense fusion both underperform BM25. Reranking improves the hybrid system, whereas adding query decomposition produces the largest further gain, with higher latency and more detected output artifacts. The complete system slightly exceeds the reported anonymous overall mean on two answer metrics but falls below it on most page-retrieval metrics. These results identify accurate page selection, rather than semantic retrieval in isolation, as the main opportunity for improvement in this setting.

View source

Similar papers

#small language model Preprint Aug 2026

Query Expansion Is More Than Generation: Improving Dense Retrieval through Better Integration

This work introduces AnchorQE, a training-free method that separately encodes the original query and its expansion before interpolating them, and shows that AnchorQE improves retrieval effectiveness by up to 12.89% when compared to widely-used expansion-only or text-level concatenation baselines across TREC-DL, LoTTE,...

Sixia Sun, Mihai Surdeanu · 1 citation
Conference Aug 2026

Resource-Efficient Semantic Retrieval Optimization for Retrieval-Augmented Generation Using FAISS, Milvus, HNSW, and Product Quantization

Retrieval-Augmented Generation (RAG) enhances large language models (LLMs) by integrating external knowledge, but its performance depends heavily on efficient semantic vector search. This paper presents a deployment-oriented empirical study that systematically compares established backends and ANN configurations under...

Achmad Rizal, Jui-Tang Wang, Gerald Wijaya Tanubrata et al. · 0 citations
Preprint Aug 2026

What the Reranker Sees: Multi-Aspect Page Annotation for Long-Document Multimodal Question Answering

This work proposes Trident, with two complementary components: Trident-R, a retriever-agnostic LLM reranker that converts each candidate into an LLM-readable semantic record, then performs a single adaptive-K rerank call; and Trident-S, a generation-side module that prompts the VLM under topical, entity, and structural...

Guanchen Wu, Jia-Yuan Ding, Subhabrata Mukherjee et al. · 0 citations
Aug 2026

STaR: a soft-labeling and triplet-aware retriever for efficient retrieval-augmented QA

This study proposes STaR, a novel retriever fine-tuning framework that integrates BM25 similarity graph-based soft labeling with a triplet similarity learning strategy based on Sentence-BERT (SBERT), and introduces a triplet-aware SBERT training architecture that explicitly models relative semantic distances between qu...

Jiali Jiang, Chih-Yung Chang, Youxi Li et al. · 0 citations

Related blog posts

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.