LILaC: Late Interacting in Layered Component Graph for Open-domain Multimodal Multihop Retrieval

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Yun, Joohyung, Lee, Doyup, Han, Wook-Shin
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866917247108251648
author Yun, Joohyung
Lee, Doyup
Han, Wook-Shin
author_facet Yun, Joohyung
Lee, Doyup
Han, Wook-Shin
contents Multimodal document retrieval aims to retrieve query-relevant components from documents composed of textual, tabular, and visual elements. An effective multimodal retriever needs to handle two main challenges: (1) mitigate the effect of irrelevant contents caused by fixed, single-granular retrieval units, and (2) support multihop reasoning by effectively capturing semantic relationships among components within and across documents. To address these challenges, we propose LILaC, a multimodal retrieval framework featuring two core innovations. First, we introduce a layered component graph, explicitly representing multimodal information at two layers - each representing coarse and fine granularity - facilitating efficient yet precise reasoning. Second, we develop a late-interaction-based subgraph retrieval method, an edge-based approach that initially identifies coarse-grained nodes for efficient candidate generation, then performs fine-grained reasoning via late interaction. Extensive experiments demonstrate that LILaC achieves state-of-the-art retrieval performance on all five benchmarks, notably without additional fine-tuning. We make the artifacts publicly available at github.com/joohyung00/lilac.
format Preprint
id arxiv_https___arxiv_org_abs_2602_04263
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle LILaC: Late Interacting in Layered Component Graph for Open-domain Multimodal Multihop Retrieval
Yun, Joohyung
Lee, Doyup
Han, Wook-Shin
Information Retrieval
Multimodal document retrieval aims to retrieve query-relevant components from documents composed of textual, tabular, and visual elements. An effective multimodal retriever needs to handle two main challenges: (1) mitigate the effect of irrelevant contents caused by fixed, single-granular retrieval units, and (2) support multihop reasoning by effectively capturing semantic relationships among components within and across documents. To address these challenges, we propose LILaC, a multimodal retrieval framework featuring two core innovations. First, we introduce a layered component graph, explicitly representing multimodal information at two layers - each representing coarse and fine granularity - facilitating efficient yet precise reasoning. Second, we develop a late-interaction-based subgraph retrieval method, an edge-based approach that initially identifies coarse-grained nodes for efficient candidate generation, then performs fine-grained reasoning via late interaction. Extensive experiments demonstrate that LILaC achieves state-of-the-art retrieval performance on all five benchmarks, notably without additional fine-tuning. We make the artifacts publicly available at github.com/joohyung00/lilac.
title LILaC: Late Interacting in Layered Component Graph for Open-domain Multimodal Multihop Retrieval
topic Information Retrieval
url https://arxiv.org/abs/2602.04263