构建多语言可视化代码代理，支持多轮纠错并提升执行成功率。
VisCoder2: Building Multi-Language Visualization Coding Agents
- 基于679K条多语言可执行代码数据，训练支持多轮修正的编码代理。
- 在32B模型规模下实现82.4%的代码执行通过率，尤其擅长符号或编译依赖语言。
- 提供可执行任务与自调试评估协议，适合研究多语言自动化可视化工具。
大型语言模型（LLM）已使代码代理具备生成、执行和修正可视化代码的能力。然而，现有模型因语言覆盖有限、执行不可靠及缺乏迭代修正机制，在实际流程中表现不佳。进展受限于狭窄的数据集与评测基准，多聚焦单轮生成与单语言任务。为此，我们提出三项互补资源：VisCode-Multi-679K是一个大规模监督数据集，包含679,000条经验证的可执行可视化样本，涵盖12种编程语言的多轮修正对话；VisPlotBench是一个系统性评测基准，包含可执行任务、渲染输出及初始生成与多轮自调试的评估协议；最后，我们推出VisCoder2，一系列基于该数据集训练的多语言可视化模型。实验表明，VisCoder2显著优于主流开源基线，接近专有模型GPT-4.1性能，且通过迭代自调试进一步提升，在32B规模下整体执行通过率达82.4%，尤其在符号或编译依赖型语言中表现突出。
原文摘要 · Abstract (English)
Large language models (LLMs) have recently enabled coding agents capable of generating, executing, and revising visualization code. However, existing models often fail in practical workflows due to limited language coverage, unreliable execution, and lack of iterative correction mechanisms. Progress has been constrained by narrow datasets and benchmarks that emphasize single-round generation and single-language tasks. To address these challenges, we introduce three complementary resources for advancing visualization coding agents. VisCode-Multi-679K is a large-scale, supervised dataset containing 679K validated and executable visualization samples with multi-turn correction dialogues across 12 programming languages. VisPlotBench is a benchmark for systematic evaluation, featuring executable tasks, rendered outputs, and protocols for both initial generation and multi-round self-debug. Finally, we present VisCoder2, a family of multi-language visualization models trained on VisCode-Multi-679K. Experiments show that VisCoder2 significantly outperforms strong open-source baselines and approaches the performance of proprietary models like GPT-4.1, with further gains from iterative self-debug, reaching 82.4% overall execution pass rate at the 32B scale, particularly in symbolic or compiler-dependent languages.
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。