Scientific visualization agent and benchmark

Areas: Data and visualization, Uses AI

CASS member: RAPIDS

Description

ChatVis is both an agent for automatically generating scientific visualizations from natural language prompts and a benchmark for evaluating synthetically-generated visualizations, whether generated by ChatVis or other tools. ChatVis_agent is an LLM agent that aids the LLM to generate Python code for ParaView scientific visualization tasks. ChatVis_benchmark is a benchmark suite of canonical visualization tasks, ParaView regression tests, and scientific use cases that includes comprehensive evaluation metrics.

Target audience

ParaView users, domain scientists, visualization researchers.

Additional resources