SVGenius：面向SVG理解、编辑与生成的大语言模型基准评测
SVGenius: Benchmarking LLMs in SVG Understanding, Editing and Generation
AI总结：
SVGenius构建了首个系统性SVG处理基准，通过2,377个查询评测22个主流模型在理解、编辑和生成上的能力，揭示推理增强训练优于单纯扩大规模，且风格迁移仍是最大挑战。
摘要：
大语言模型（LLMs）和多模态大语言模型在SVG处理方面已展现出有前景的能力，然而现有基准评测存在真实世界覆盖范围有限、缺乏复杂度分层以及评估范式碎片化等问题。我们提出SVGenius，一个包含2,377个查询的综合基准，涵盖理解、编辑和生成三个递进维度。SVGenius基于来自24个应用领域的真实世界数据构建，并具有系统化的复杂度分层，通过8个任务类别和18项指标对模型进行评估。我们评测了22个主流模型，涵盖不同规模、架构、训练范式和可访问性水平。分析结果表明，虽然闭源专有模型显著优于开源模型，但所有模型都随着复杂度增加而出现系统性的性能下降，这表明当前方法存在根本性局限；然而，推理增强训练比单纯扩大规模更能有效克服这些局限，不过风格迁移仍然是所有模型类型中最具挑战性的能力。SVGenius建立了首个面向SVG处理的系统性评估框架，为开发能力更强的矢量图形模型和推进自动化图形设计应用提供了关键见解。附录和补充材料（包括所有数据和代码）可在https://zju-real.github.io/SVGenius获取。
英文摘要：
Large Language Models (LLMs) and Multimodal LLMs have shown promising capabilities for SVG processing, yet existing benchmarks suffer from limited real-world coverage, lack of complexity stratification, and fragmented evaluation paradigms. We introduce SVGenius, a comprehensive benchmark comprising 2,377 queries across three progressive dimensions: understanding, editing, and generation. Built on real-world data from 24 application domains with systematic complexity stratification, SVGenius evaluates models through 8 task categories and 18 metrics. We assess 22 mainstream models spanning different scales, architectures, training paradigms, and accessibility levels. Our analysis reveals that while proprietary models significantly outperform open-source counterparts, all models exhibit systematic performance degradation with increasing complexity, indicating fundamental limitations in current approaches; however, reasoning-enhanced training proves more effective than pure scaling for overcoming these limitations, though style transfer remains the most challenging capability across all model types. SVGenius establishes the first systematic evaluation framework for SVG processing, providing crucial insights for developing more capable vector graphics models and advancing automated graphic design applications. Appendix and supplementary materials (including all data and code) are available at https://zju-real.github.io/SVGenius.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
