Scientific visualization tools enable exploration of complex scientific datasets but present a steep learning curve, creating hurdles for scientists by demanding technical expertise beyond their domain. AI agents powered by Large Language Models (LLMs) have the potential to address this issue, and yet their effectiveness and reliability in doing so are currently unclear. In this project we will develop a benchmark suite, which focuses on generating simple scientific visualization actions in ParaView via Python code. We will evaluate the efficacy of popular LLMs in performing the tasks in the suite under various conditions, checking execution and pipeline validity across extensive trials. We will evaluate the agentic workflow with science experts at Argonne. The results will indicate how LLMs can assist domain scientists in using scientific visualization software.
Stand Out From the Crowd
Upload your resume and get instant feedback on how well it matches this job.
Job Type
Full-time
Career Level
Intern
Education Level
No Education Listed
Number of Employees
1,001-5,000 employees