Episode
CiteVQA: Benchmarking Evidence Attribution for Trustworthy Document Intelligence
- Podcast
- Daily Paper Cast
- Published
- May 19, 2026
- Duration seconds
- 1390
- Processing state
not_requested- Canonical source
- https://share.transistor.fm/s/95291574
Actions
POST https://stenobird.com/v1/public/podcasts/daily-paper-cast-7079649/episodes/citevqa-benchmarking-evidence-attribution-for-trustworthy-document-intelligence/transcription-requests
Idempotently request low-priority transcript generation for this episode.GET https://stenobird.com/podcast/daily-paper-cast-7079649/citevqa-benchmarking-evidence-attribution-for-trustworthy-document-intelligence.md
Read the agent-friendly Markdown representation of this episode resource.
Summary
🤗 Upvotes: 164 | cs.CL, cs.CV Authors: Dongsheng Ma, Jiayu Li, Zhengren Wang, Yijie Wang, Jiahao Kong, Weijun Zeng, Jutao Xiao, Jie Yang, Wentao Zhang, Bin Wang, Conghui He Title: CiteVQA: Benchmarking Evidence Attribution for Trustworthy Document Intelligence Arxiv: http://arxiv.org/abs/2605.12882v1 Abstract: Multimodal Large Language Models (MLLMs) have significantly advanced document understanding, yet current Doc-VQA evaluations score only the final answer and leave the supporting evidence unchecked. This answer-only approach masks a critical failure mode: a model can land on the correct answer while grounding it in the wrong passage -- a critical risk in high-stakes domains like law, finance, and medicine, where every conclusion must be traceable to a specific source region. To address this, we introduce CiteVQA, a benchmark that requires models to return element-level bounding-box citations alongside each answer, evaluating both jointly. CiteVQA comprises 1,897 questions across 711 PDFs spanning seven domains and two languages, averaging 40.6 pages per document. To ensure fidelity and scalability, the ground-truth citations are generated by an automated pipeline-which identifies crucial evidence via masking ablation-and are subsequently validated through expert review. At the core of our evaluation is Strict Attributed Accuracy (SAA), which credits a prediction only when the answer and the cited region are both correct. Auditing 20 MLLMs reveals a pervasive Attribution Hallucination: models frequently produce the right answer while citing the wrong region. The strongest system (Gemini-3.1-Pro-Preview) achieves an SAA of only 76.0, and the strongest open-source MLLM reaches just 22.5. Ultimately, towards trustworthy document intelligence, CiteVQA exposes a reliabi…