跳到正文
arXiv cs.CL· William Rudman, Abhishek Divekar, Kanishk Jain, Sebastian Joseph, Stella S. R. Offner, Matthew Lease, Kyle Mahowald, Greg Durrett, Junyi Jessy Li·· 10 小时前AI 评分40

VESTA:用统计工具智能体做可视化数据探索

VESTA: Visual Exploration with Statistical Tool Agents

AI 导读

VESTA 是一个为 VLM 配备动态增长工具包的框架,通过数据变换、假设驱动的可视化和统计检验来引导模型改进。团队同时推出 DAWN 基准,覆盖分布拟合与时间序列建模,并包含初始质量函数、引力波啁啾信号等真实天文任务。动态工具创建优于既有智能体流程,在全部时间序列任务和复杂分布拟合任务上超过 Claude Code、Codex 与 OpenCode,但后者在简单分布上拟合更好。

正文 · 原文

View PDF HTML (experimental)

Abstract:Fitting quantitative models to data is a central step in scientific workflows, yet it remains one of the least automated. Recent agent-based systems leverage language and vision-language models (VLMs) to propose statistical models, but these systems struggle on challenging modeling tasks. To address these limitations, we introduce VESTA (Visual Exploration with Statistical Tool Agents), a framework that equips VLMs with a dynamically growing toolkit to guide model refinement through data transformations, hypothesis-driven visualizations, and statistical tests. Unlike prior systems that rely on iterative critique alone, VESTA explores data before and during refinement by selecting or creating reusable diagnostic tools. We evaluate VESTA using three toolkit configurations: no tools, static expert-written tools, and dynamic model-written tools. To support this evaluation, we introduce DAWN (Dataset for Automated Workflows and Numerical Modeling), a benchmark targeting distribution fitting and time series modeling with varying difficulty tiers that culminates in real-world astronomy tasks including modeling initial mass functions and gravitational-wave chirp signals. VESTA's dynamic tool creation outperforms prior agentic pipelines for statistical analysis and visual tool creation, with the largest gains on complex and domain-specific tasks. Against strong agentic coding harnesses (Claude Code, Codex, and OpenCode), VESTA is better on all time series tasks and on complex distribution fitting tasks, though the harnesses achieve better fits on simpler distributions. Further, dynamically generated tools are substantially more sophisticated than those produced by existing visual tool-creation systems, covering more diagnostic categories per function and strongly preferring visual outputs that the VESTA VLM critic can reason over directly.
Subjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Computation (stat.CO)
MSC classes: 68T07, 68T50, 62F15, 62M10, 85A35
ACM classes: I.2.1; I.2.6; I.2.11; G.3; I.2.10; J.2
Cite as: arXiv:2606.00384 [cs.AI]
  (or arXiv:2606.00384v3 [cs.AI] for this version)
  https://doi.org/10.48550/arXiv.2606.00384

arXiv-issued DOI via DataCite

Submission history

From: William Rudman Jr [view email]
[v1] Fri, 29 May 2026 21:57:37 UTC (3,090 KB)
[v2] Sun, 7 Jun 2026 19:48:24 UTC (5,354 KB)
[v3] Thu, 19 Nov 2026 16:02:04 UTC (4,151 KB)

来源:arXiv cs.CL · arxiv.org