证据链：面向迭代检索增强生成的像素级视觉归因
Chain of Evidence: Pixel-Level Visual Attribution for Iterative Retrieval-Augmented Generation
AI总结：
提出Chain of Evidence (CoE)框架，利用视觉语言模型直接对检索到的文档截图进行推理，输出精确边界框以可视化完整推理链，解决迭代检索增强生成中的粗粒度归因和视觉语义丢失问题。
摘要：
迭代检索增强生成（iRAG）已成为通过逐步检索和推理外部文档来回答复杂多跳问题的强大范式。然而，当前系统主要基于解析文本运行，这造成了两个关键瓶颈：（1）粗粒度归因，用户需要根据模糊的文本级引用在冗长文档中手动定位证据；（2）视觉语义丢失，将视觉丰富的文档（如幻灯片、带有图表的PDF）转换为文本会丢弃对推理至关重要的空间逻辑和布局线索。为弥合这一差距，我们提出了证据链（CoE），这是一个与检索器无关的视觉归因框架，利用视觉语言模型直接对检索到的文档候选截图进行推理。CoE消除了特定格式的解析，输出精确的边界框，可视化检索候选集中的完整推理链。我们在两个不同的基准上评估CoE：Wiki-CoE，一个源自2WikiMultiHopQA的大规模结构化网页数据集；以及SlideVQA，一个具有挑战性的演示幻灯片数据集，包含复杂图表和自由形式布局。实验表明，微调后的Qwen3-VL-8B-Instruct取得了稳健的性能，在需要视觉布局理解的场景中显著优于基于文本的基线，同时为像素级可解释的iRAG建立了与检索器无关的解决方案。我们的代码可在https://github.com/PeiYangLiu/CoE.git获取。
英文摘要：
Iterative Retrieval-Augmented Generation (iRAG) has emerged as a powerful paradigm for answering complex multi-hop questions by progressively retrieving and reasoning over external documents. However, current systems predominantly operate on parsed text, which creates two critical bottlenecks: (1) \textit{Coarse-grained attribution}, where users are burdened with manually locating evidence within lengthy documents based on vague text-level citations; and (2) \textit{Visual semantic loss}, where the conversion of visually rich documents (e.g., slides, PDFs with charts) into text discards spatial logic and layout cues essential for reasoning. To bridge this gap, we present \textbf{Chain of Evidence (CoE)}, a retriever-agnostic visual attribution framework that leverages Vision-Language Models to reason directly over screenshots of retrieved document candidates. CoE eliminates format-specific parsing and outputs precise bounding boxes, visualizing the complete reasoning chain within the retrieved candidate set. We evaluate CoE on two distinct benchmarks: \textbf{Wiki-CoE}, a large-scale dataset of structured web pages derived from 2WikiMultiHopQA, and \textbf{SlideVQA}, a challenging dataset of presentation slides featuring complex diagrams and free-form layouts. Experiments demonstrate that fine-tuned Qwen3-VL-8B-Instruct achieves robust performance, significantly outperforming text-based baselines in scenarios requiring visual layout understanding, while establishing a retriever-agnostic solution for pixel-level interpretable iRAG. Our code is available at https://github.com/PeiYangLiu/CoE.git.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
