SIVA: A Conversational and Declarative Scientific Visualization Tool
Authors
Michael Ballantyne (Washington State University), Aayush Kaji Shrestha (Washington State University), Pascal Grosset (Los Alamos National Laboratory), Thomas Gilray (Washington State University), Roxana Bujack (Los Alamos National Laboratory)
Presentation
- Session
- Me, Myself, and AI
- Time
- Wednesday, Nov 11, 13:54 – 14:03 (US/Eastern) · session 13:00 – 14:30
- Location
- Hall America center
Links
Sign in to access the preprint PDF.
Sign in- Download Supplemental Material
Keywords
Scientific visualization, domain-specific languages, large language models, AI agents, MCP
Abstract
The growing adoption of LLMs has led to increased interest in LLM-assisted visualization. While these systems can generate visualizations from natural language, concerns remain regarding their reliability, interpretability, and user control. In this paper we present our initial work on SIVA, a conversational and declarative scientific visualization system that supports iterative human–AI collaboration. In SIVA, visualizations are expressed as specifications in a domain-specific language (DSL) that serves as a shared, inspectable artifact. This approach ensures that the visualization shown is always in sync with a human-readable representation for scientists to audit and manually revise. We demonstrate SIVA on a wildfire simulation dataset.
For Practitioners
The SIVA system is currently targeted at simulation scientists with VTK-compatible structured or image grid data who would like AI assistance in exploring and visualizing their data. Visualization practitioners building their own AI-assisted tooling may also be interested in applying the architectural pattern in their own systems: an MCP-based AI workflow that exposes the generated visualization as code in a declarative, deeply-embedded DSL to allow direct user auditing and modifications.