Skip to content

DocAttriBench: Benchmarking Answer Grounding in Document Visual Question Answering

Sep 2026 · arXiv.org · Vol abs/2609.20574 · 1 citation · 54 references
Computer Science

TL;DR

DocAttriBench (DAB) is introduced, a large-scale benchmark for fine-grained, element-level source attribution in Document VQA, grounding answers to specific layout elements such as text blocks, tables, and images.

Abstract

Answer grounding in document visual question answering remains an open challenge: most benchmarks lack grounding annotations or provide limited-quality labels, while constructing grounded datasets still requires costly manual effort. We introduce DocAttriBench (DAB), a large-scale benchmark for fine-grained, element-level source attribution in Document VQA, grounding answers to specific layout elements such as text blocks, tables, and images. To build DAB, we propose a Mask-based Perplexity-Derived Attribution method (MAPPET) that combines document layout and language modeling to identify the most informative element for each answer. MAPPET measures the increase in perplexity after masking candidate elements and attributes the answer to the element contributing most to model confidence. Applying MAPPET to multiple existing Document VQA datasets yields DAB, with 237k documents and 296k question-answer pairs with element-level grounding. We benchmark grounding-capable multimodal LLMs on DAB, evaluating answer accuracy, attribution accuracy, and overall answer quality. Results show that while larger models generally achieve higher answer accuracy, even the strongest models often fail to localize the supporting elements. DAB provides a scalable benchmark for developing grounded, verifiable, and trustworthy Document VQA models. Dataset and code are available at https://aimagelab.github.io/DocAttriBench/.

View source

Similar papers

#computer vision Preprint Oct 2026

VELD-QA: A Large-Scale Dataset for Visual Evidence Grounding and Question Answering in Long Documents

Existing long-document QA benchmarks rarely combine large-scale human-authored questions with detailed visual evidence annotations, limiting diagnosis of evidence localization and interpretation failures. We present VELD-QA, a large-scale benchmark for Visual Evidence grounding and question answering in Long Documents....

Zhivar Sourati, Marianne Menglin Liu, Rongguang Wang et al. · 0 citations
#machine learning Preprint Oct 2026

Towards a Learnable Retriever for Textual Graph Question Answering

Textual graph question answering (GraphQA) and graph-based retrieval-augmented generation (GraphRAG) have gained increasing attention as a way to ground large language models (LLMs) on structured knowledge. Despite extensive efforts, existing methods still face two key gaps. The first is an architectural gap: existing...

Kai-Feng Jin, Yu-Chen Yan, Zhichen Zeng et al. · 0 citations
Preprint Oct 2026

Benchmarking Answer-Grounding Consistency in Remote Sensing Visual Question Answering

In grounded remote sensing visual question answering, a model produces a textual answer and identifies relevant objects in the image as grounding evidence. Existing evaluations measure answer correctness and localization quality, but do not directly assess answer-grounding consistency. We propose the Answer-Evidence Su...

Ping Guo, Zhi-Qi Huang, Xin-Ran Li et al. · 0 citations
Open access Oct 2026

MangoVQA: A Multi-Answer Visual Question Answering System for Grading SNI-Aligned Mango Quality using Deep Learning and Text Processing

Visual Question Answering (VQA), a multimodal AI task combining computer vision and language understanding, answers text-based questions about visual content. Although VQA has developed rapidly, most existing approaches assume a single correct answer, whereas real-world images often require several valid answers simult...

Nida Hasanati, Taufik Djatna, I. S. Sitanggang et al. · 0 citations
#artificial intelligence Preprint Nov 2026

VisProve: Learning Typed Visual Provenance for Scientific Document Question Answering

Visual citations show where an answer is supported, but not how the cited content contributes to it. In scientific document question answering, a model may read a value directly, compress several observations, or combine evidence to infer a conclusion. We study typed visual provenance, where a system generates an answe...

Jing-Xuan Wei, Zhen-Yang Wang, Zheng Guo et al. · 0 citations
Book Open access Nov 2026

GQA-Q2Q: A Large-scale Benchmark for Entity Ambiguity Resolution in Visual Question Answering

Vision-language models achieve remarkable results on various visual question-answering benchmarks. However, they struggle with ambiguous questions where the target entity is unclear. To address this issue, it is essential to construct a benchmark dataset that pairs ambiguous questions with entity-grounded clarifying su...

Gyu-Min Park, Seong-Bae Park · 0 citations

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