构建可审计的本地部署金融图表问答系统，兼顾准确性与可信度。
AgentFinVQA: A Deployable Multi-Agent Pipeline for Auditable Financial Chart QA
- 分步多智能体架构：规划、识别、图例对齐、视觉检查与验证，全程留痕。
- 在FinMME上比基线提升7.68个百分点，本地运行开源模型仍保持高精度。
- 验证结果可作置信度信号，助力人工审核，适合监管严格场景使用。
在受监管环境下，金融图表问答不仅需高准确率，还需可审计性与数据本地化。现有图表问答系统多关注准确率，缺乏透明性，且依赖外部API。本文提出AgentFinVQA，一种多智能体流水线，将每条查询分解为规划、OCR、图例对齐、视觉检查与验证，并为每个样本生成可追溯的模型评估包（MEP）。在FinMME数据集上，该系统相较于采用专用骨干网络（Gemini-3 Flash）的零样本基线提升7.68个百分点（71.24% vs. 63.56%，McNemar检验p ≈ 1.1×10⁻¹⁶），使用本地部署的开源模型Qwen3.6-27B-FP8也实现4.84个百分点的提升。验证结果作为置信度信号，确认答案准确率为68.2%，修订答案为55.6%。错误分析显示，问题误解、图例混淆和提取错误占近三分之二失败案例，且最难被验证器检测，指向未来改进方向。结果表明，可审计、本地部署的金融图表问答是可行的，开源系统在保障数据主权的同时保留了主要性能优势。代码已公开以支持复现评估。
原文摘要 · Abstract (English)
Financial chart question answering in regulated settings demands more than accuracy: practitioners must know which answers to trust before acting on them, and many institutions cannot send client data to external model providers. Yet existing chart-QA agents are accuracy-focused and opaque, and most assume proprietary API access; to our knowledge, none combines auditability with on-premise deployability without significant accuracy compromise. We present AgentFinVQA, a multi-agent pipeline that decomposes each query into planning, OCR, legend grounding, visual inspection, and verification, recording every step in a traceable Model Evaluation Packet (MEP) per sample. On FinMME, AgentFinVQA improves $+7.68$ pp over a primary-backbone matched zero-shot baseline with a proprietary backbone (Gemini-3 Flash; 71.24% vs. 63.56%, McNemar $p \approx 1.1 \times 10^{-16}$), and $+4.84$ pp with open-weights Qwen3.6-27B-FP8 served locally. The verifier's verdict also serves as a useful confidence signal (68.2% vs. 55.6% exact accuracy on confirmed vs. revised answers), enabling human-in-the-loop review routing. Error analysis shows that question misunderstanding, legend confusion and extraction error account for nearly two-thirds of failures and are the categories least detected by the verifier, identifying clear directions for future work. Together these results show that auditable, on-premise financial chart QA is practical and that the open-weights system keeps most of the accuracy gains while enabling full data residency. We release our code to support reproducible evaluation.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。