SciVisAgentBench: A Benchmark for Evaluating Scientific Data Analysis and Visualization Agents

TL;DR

SciVisAgentBench is a new comprehensive benchmark designed to evaluate scientific data analysis and visualization agents, enabling systematic comparison and progress tracking in realistic multi-step SciVis tasks.

Contribution

It introduces a structured taxonomy, a multimodal evaluation pipeline, and initial baselines, addressing the lack of principled benchmarks for SciVis agents.

Findings

Expert and LLM judges show good agreement in evaluations.

Baseline agents reveal significant capability gaps.

The benchmark supports systematic comparison and failure diagnosis.

Abstract

Recent advances in large language models (LLMs) have enabled agentic systems that translate natural language intent into executable scientific visualization (SciVis) tasks. Despite rapid progress, the community lacks a principled and reproducible benchmark for evaluating these emerging SciVis agents in realistic, multi-step analysis settings. We present SciVisAgentBench, a comprehensive and extensible benchmark for evaluating scientific data analysis and visualization agents. Our benchmark is grounded in a structured taxonomy spanning four dimensions: application domain, data type, complexity level, and visualization operation. It currently comprises 108 expert-crafted cases covering diverse SciVis scenarios. To enable reliable assessment, we introduce a multimodal outcome-centric evaluation pipeline that combines LLM-based judging with deterministic evaluators, including image-based metrics, code checkers, rule-based verifiers, and case-specific evaluators. We also conduct a validity study with 12 SciVis experts to examine the agreement between human and LLM judges. Using this framework, we evaluate representative SciVis agents and general-purpose coding agents to establish initial baselines and reveal capability gaps. SciVisAgentBench is designed as a living benchmark to support systematic comparison, diagnose failure modes, and drive progress in agentic SciVis. The benchmark is available at https://scivisagentbench.github.io/.