@inproceedings{dong-etal-2026-hierarchical,
    title = "Hierarchical Visual Agent: Managing Contexts in Joint Image-Text Space for Advanced Chart Reasoning",
    author = "Dong, Qihua  and
      He, Ruozhen  and
      Chen, Junwen  and
      Wang, Yizhou  and
      Ma, Xu  and
      Jiang, Songyao  and
      Fu, Yun",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.findings-acl.1914/",
    pages = "38390--38401",
    ISBN = "979-8-89176-395-1",
    abstract = "Advanced chart question answering requires both precise perception of small visual elements and multi-step reasoning across several subplots. While existing MLLMs are strong at understanding single plots, they often struggle with multi-step reasoning across multiple subplots. We propose HierVA, a \textbf{hierarchical visual agent} framework for chart reasoning that iteratively constructs and updates a working context in a joint image{--}text space. A high-level manager generates plans and maintains a compact context containing only key information, while specialized sub-agents perform reasoning, gather evidence, and return results. In particular, the agent maintains separate visual and textual contexts, using a \textit{zoom-in tool} to restrict the visual context. Experiments on the chart reasoning benchmarks demonstrate consistent improvements over strong multimodal baselines, and ablation studies verify that hierarchical architecture, limited visual context, and distilled context contribute complementary gains."
}