探索LLM代理在科学可视化中的交互范式
Exploring LLM Agent Designs and Interaction Modalities for Scientific Visualization
AI总结：
本文研究不同类型的LLM代理在科学可视化任务中的表现，通过评估八种代表性代理在15个基准任务中的可视化质量、效率、鲁棒性和计算成本，发现通用编程代理任务成功率最高但计算成本高，领域专用代理更高效稳定但灵活性差，计算机使用代理在单步任务表现良好但多步流程受限。
摘要：
本文探讨了不同类型的大型语言模型（LLM）代理在科学可视化（SciVis）任务中的表现，其中用户通过自然语言指令生成可视化工作流。我们通过评估八个代表性代理在15个基准任务中的表现，比较了三种主要交互范式：领域专用代理（具有结构化工具使用）、计算机使用代理和通用编程代理。我们进一步分析了交互模式，包括代码脚本和模型上下文协议（MCP）或API调用（用于结构化工具使用），以及命令行界面（CLI）和图形用户界面（GUI）（用于更通用的交互），并研究了选定代理中持久记忆的影响。结果揭示了不同范式和模式之间的明显权衡。通用编程代理在任务成功率最高但计算成本高，而领域专用代理更高效且稳定但灵活性较差。计算机使用代理在单步任务中表现良好，但在多步流程中表现不佳，表明长视图规划是其主要限制。在CLI和GUI设置中，持久记忆在重复试验中提高了性能，尽管其益处取决于底层交互模式和反馈质量。这些发现表明，没有单一方法足够，未来SciVis系统应结合结构化工具使用、交互能力以及自适应记忆机制，以平衡性能、鲁棒性和灵活性。
英文摘要：
This paper examines how large language model (LLM) agents perform on scientific visualization (SciVis) tasks that require generating visualization workflows from natural-language instructions. We compare three representative agent designs: domain-specific agents with structured tool use, computer-use agents, and general-purpose coding agents, across 15 benchmark tasks, evaluating visualization quality, efficiency, robustness, computational cost, and the impact of persistent memory. We further study interaction modalities, including code scripts, model context protocol (MCP) or API calls, command-line interfaces (CLI), and graphical user interfaces (GUI). Our goal is to characterize the tradeoffs among representative SciVis agent configurations used in practice. The results reveal clear tradeoffs across agent designs and interaction modalities. General-purpose coding agents achieve the highest task success rates but incur greater computational cost, whereas domain-specific agents are more efficient and stable but less flexible. Computer-use agents perform well on individual operations but struggle with multi-step workflows. Across both CLI- and GUI-based settings, persistent memory improves performance over repeated trials, but its effectiveness depends on the interaction mode and the quality of feedback. These findings suggest that future SciVis systems should combine structured tool use, interactive capabilities, and adaptive memory mechanisms to balance performance, robustness, and flexibility.
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。
