Saved in:
Bibliographic Details
Main Authors: Dantart, Alex, Kóvacs-Navarro, Marco
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2601.10215
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917205193523200
author Dantart, Alex
Kóvacs-Navarro, Marco
author_facet Dantart, Alex
Kóvacs-Navarro, Marco
contents In enterprise datasets, documents are rarely pure. They are not just text, nor just numbers; they are a complex amalgam of narrative and structure. Current Retrieval-Augmented Generation (RAG) systems have attempted to address this complexity with a blunt tool: linearization. We convert rich, multidimensional tables into simple Markdown-style text strings, hoping that an embedding model will capture the geometry of a spreadsheet in a single vector. But it has already been shown that this is mathematically insufficient. This work presents Topo-RAG, a framework that challenges the assumption that "everything is text". We propose a dual architecture that respects the topology of the data: we route fluid narrative through traditional dense retrievers, while tabular structures are processed by a Cell-Aware Late Interaction mechanism, preserving their spatial relationships. Evaluated on SEC-25, a synthetic enterprise corpus that mimics real-world complexity, Topo-RAG demonstrates an 18.4% improvement in nDCG@10 on hybrid queries compared to standard linearization approaches. It's not just about searching better; it's about understanding the shape of information.
format Preprint
id arxiv_https___arxiv_org_abs_2601_10215
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Topo-RAG: Topology-aware retrieval for hybrid text-table documents
Dantart, Alex
Kóvacs-Navarro, Marco
Artificial Intelligence
In enterprise datasets, documents are rarely pure. They are not just text, nor just numbers; they are a complex amalgam of narrative and structure. Current Retrieval-Augmented Generation (RAG) systems have attempted to address this complexity with a blunt tool: linearization. We convert rich, multidimensional tables into simple Markdown-style text strings, hoping that an embedding model will capture the geometry of a spreadsheet in a single vector. But it has already been shown that this is mathematically insufficient. This work presents Topo-RAG, a framework that challenges the assumption that "everything is text". We propose a dual architecture that respects the topology of the data: we route fluid narrative through traditional dense retrievers, while tabular structures are processed by a Cell-Aware Late Interaction mechanism, preserving their spatial relationships. Evaluated on SEC-25, a synthetic enterprise corpus that mimics real-world complexity, Topo-RAG demonstrates an 18.4% improvement in nDCG@10 on hybrid queries compared to standard linearization approaches. It's not just about searching better; it's about understanding the shape of information.
title Topo-RAG: Topology-aware retrieval for hybrid text-table documents
topic Artificial Intelligence
url https://arxiv.org/abs/2601.10215