File size: 2,885 Bytes
3a2d55e 3cb1822 3a2d55e 3cb1822 f4ab3a3 3cb1822 6baf440 3cb1822 f4ab3a3 6baf440 3cb1822 e98f189 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 | ---
pretty_name: SciVisAgentBench
task_categories:
- text-to-3d
- other
---
# SciVisAgentBench Tasks
This repository is a secondary repo of [SciVisAgentBench](https://github.com/KuangshiAi/SciVisAgentBench), contains scientific data analysis and visualization datasets and tasks for benchmarking scientific visualization agents.
## Download Volume Datasets
`download_and_organize.py` downloads and organizes datasets under 512MB. Make sure you run the following script before you evaluate your agents through [SciVisAgentBench](https://github.com/KuangshiAi/SciVisAgentBench):
```shell
python download_and_organize.py
```
## Data Organization
All the volume datasets from http://klacansky.com/open-scivis-datasets/ have been organized into a consistent structure.
### Directory Structure
The datasets and tasks for ParaView-MCP and ChatVis are organized into the `main`, the `sci_volume_data`, and the `chatvis_bench` folders, while `napari_mcp_evals` holds tasks and datasets for bioimage-agent. The `chatvis_bench` folder contains 20 test cases from the official [ChatVis](https://github.com/tpeterka/ChatVis) benchmark.
Each dataset in the `main`, the `sci_volume_data`, and the `chatvis_bench` folders follows this structure:
```
dataset_name/
├── data/
│ ├── dataset_file.raw # The actual data file
│ └── dataset_name.txt # Metadata about the dataset
├── GS/ # Ground truth folder (ParaView state + pvpython code)
├── task_description.txt # ParaView visualization task
└── visualization_goals.txt # Evaluation criteria for the visualization
```
### Available Volume Datasets
- **37 datasets under 512MB** are suggested to be downloaded
- **18 datasets over 512MB** are listed but not downloaded
See `datasets_list.md` for a complete list with specifications. And `datasets_info.json` is the complete JSON file with all dataset metadata.
### Task Descriptions
Each dataset has:
1. **Task descriptions** - Based on dataset type (medical, simulation, molecular, etc.)
2. **Visualization goals** - Evaluation criteria tailored to the dataset characteristics
3. **Ground Truth** - Ground truth pvpython code, ParaView state and screenshots
## Acknowledgement
SciVisAgentBench was mainly created by Kuangshi Ai (kai@nd.edu), Shusen Liu (liu42@llnl.gov), and Haichao Miao (miao1@llnl.gov). Some of the test cases are provided by Kaiyuan Tang (ktang2@nd.edu). We sincerely thank the open-source community for their invaluable contributions. This project is made possible thanks to the following outstanding projects:
- [ParaView-MCP](https://github.com/LLNL/paraview_mcp)
- [Bioimage-agent](https://github.com/LLNL/bioimage-agent)
- [ChatVis](https://github.com/tpeterka/ChatVis)
## License
© 2025 University of Notre Dame.
Released under the [License](./LICENSE). |