Chain of Evidence: Pixel-Level Visual Attribution for Iterative Retrieval-Augmented Generation

arXiv:2605.01284v2 Announce Type: replace-cross
Abstract: 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) textitCoarse-grained attribution, where users are burdened with manually locating evidence within lengthy documents based on vague text-level citations; and (2) textitVisual 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 textbfChain 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: textbfWiki-CoE, a large-scale dataset of structured web pages derived from 2WikiMultiHopQA, and textbfSlideVQA, 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.