arXiv:2607. 14117v1 Announce Type: cross Abstract: Cross-version differencing of scientific documents is essential in scholarly publishing and technical documentation, but remains challenging because scientific documents are page-structured artifacts containing heterogeneous elements such as text, tables, formulas, figures, and layout cues.
arXiv:2606. 06242v1 Announce Type: cross Abstract: Institutional documents contain substantial amounts of operational and analytical information embedded within figures and tables.
SciDocBench is a workflow-centered benchmark for scientific document understanding that includes 124 expert-authored questions across seven capability groups and 19 subtasks in five scientific domains. Each question is evaluated under four conditions—English or Chinese, all-images-first or interleaved document representations—resulting in 496 evaluation instances. The benchmark is paired with SciDocIR, a typed evidence-graph representation, and SciDocDataset, a collection of 15K fine-tuning and 8K reinforcement-learning samples, forming an evaluation-to-training framework for scientific-document assistants.
arXiv:2608. 15064v1 Announce Type: new Abstract: Parsing visual documents into machine-readable representations is fundamental to document intelligence.
arXiv:2608. 05478v1 Announce Type: cross Abstract: Graphical Abstracts (GAs) visually summarize the key findings of academic papers, playing a crucial role in facilitating the understanding of research content.
Institutional documents contain substantial amounts of operational and analytical information embedded within figures and tables. Current approaches for extracting visual content from documents are largely built around generic document layout analysis, where figures and tables are treated as uniformly relevant document objects rather than semantically meaningful analytical artifacts.
Transforming table-form documents into machine-processable records requires recovering not only their visible content but also the multilevel structure that organizes it. However, existing benchmarks evaluate either holistic document outputs or conventional table grids, and their aggregate scores provide little insight into where structural failures occur.
arXiv:2607. 29058v1 Announce Type: new Abstract: Professional-document review is a constraint-checking problem in which decisions depend on relations among text, geometry, pages, and document revisions.
arXiv:2609.27784v1 Announce Type: cross Abstract: Semi-structured documents are ubiquitous in scientific reports, financial statements, and technical manuals. Question answering over such documents r...
The paper introduces a new task called compositional layout understanding, focusing on interpreting complex, multi‑layer document and UI designs. It presents CoDeLayout, a VQA dataset of about 20,000 real‑world layouts annotated with compositional element pairs and design intent. The authors identify semantic drift and structural ambiguity as key challenges for vision‑language models and propose MASON, a post‑training approach that combines multimodal alignment and structural perception to improve performance, achieving 91.66% accuracy with only 30% of the training data.
arXiv:2608. 14032v1 Announce Type: cross Abstract: Existing multimodal RAG methods often flatten structured documents into isolated text and image units, weakening the source organization and local text-image logic needed for faithful evidence selection and placement.
arXiv:2609.24220v1 Announce Type: new Abstract: Retrieval-Augmented Generation (RAG) systems over enterprise knowledge bases must ingest heterogeneous document formats -- PDFs, Word documents, presen...
The paper introduces a new task called compositional layout understanding, focusing on interpreting complex, multi-layer document and UI layouts that involve hierarchical relationships among visually entangled elements. It presents CoDeLayout, a VQA dataset of about 20,000 real-world layouts annotated with compositional element pairs and design intent, and identifies two main challenges for current vision‑language models: semantic drift between textual metadata and visual content, and structural ambiguity in hierarchical inter‑element relationships. To address these, the authors propose MASON, a post‑training paradigm that combines multimodal alignment and structural perception, achieving a 91.66% accuracy on CoDeLayout and outperforming full‑data direct fine‑tuning with only 30% of the training data.