摘要：视觉语言模型（VLM）已展现出根据文本或视觉规格生成可视化代码的强大能力。然而，现实中的可视化创作本质上是一个迭代过程：用户经常需要修改现有可视化，以修复存在缺陷的图表，或将其调整为所需风格。现有基准主要评估从零开始的生成能力，而基于多模态反馈的可视化代码编辑在很大程度上尚未得到探索。我们提出 VisEditBench，一个包含 1,395 个人工标注可视化代码编辑任务的基准，其任务基于真实的可视化工作流和故障案例构建。VisEditBench 覆盖两类实际场景：反馈引导修复，即模型结合存在错误或被标记的图表及文本反馈来修改可视化代码；以及参考引导重塑，即模型修改代码以匹配目标图表图像。对 20 个最先进 VLM 的评估表明，可视化代码编辑仍然具有挑战性：Claude-4.6-Sonnet 以 74.46% 的总体通过率取得最佳成绩，而多数开源模型的通过率仍低于 50%。在以视觉为基础的样式适配任务上，模型表现尤其较弱，Claude-4.6-Sonnet 的通过率也仅为 55.71%。为建立强有力的基线，我们进一步提出 VisEditAgent，这是一个以渲染结果为基础的编辑框架，可迭代地生成、执行、验证和优化候选编辑方案。VisEditAgent 基于 GPT-4o 构建，将总体通过率从 55.75% 提升至 67.99%，证明渲染驱动的反馈对于忠实地编辑可视化内容至关重要。我们将在此 HTTPS 链接发布 VisEditBench。

Abstract:Vision-language models (VLMs) have shown strong capabilities in generating visualization code from textual or visual specifications. However, real-world visualization authoring is inherently iterative: users frequently revise existing visualizations to repair flawed charts or adapt them to desired styles. Existing benchmarks primarily evaluate generation from scratch, leaving visualization code editing from multimodal feedback largely unexplored. We introduce VisEditBench, a benchmark of 1,395 human-annotated visualization code-editing tasks grounded in realistic visualization workflows and failure cases. VisEditBench covers two practical settings: feedback-guided repair, where models revise visualization code using buggy or marked charts together with textual feedback, and reference-guided restyling, where models modify code to match a target chart image. Evaluating 20 state-of-the-art VLMs reveals that visualization code editing remains challenging: Claude-4.6-Sonnet achieves the best overall pass rate of 74.46%, while most open-source models remain below 50%. Performance is particularly weak on visually grounded style adaptation, where Claude-4.6-Sonnet achieves only 55.71%. To establish a strong baseline, we further propose VisEditAgent, a render-grounded editing framework that iteratively generates, executes, validates, and refines candidate edits. Built on GPT-4o, VisEditAgent improves overall pass rate from 55.75% to 67.99%, demonstrating the importance of render-grounded feedback for faithful visualization editing. We will release VisEditBench at this https URL.