metadata
license: other
license_name: openxvqa-research-only
license_link: https://arxiv.org/abs/2310.08864
task_categories:
- visual-question-answering
- multiple-choice
language:
- en
size_categories:
- 1K<n<10K
pretty_name: OpenXVQA
configs:
- config_name: default
data_files:
- split: test
path: data/test-*.parquet
OpenXVQA
VQA benchmark derived from Open X-Embodiment robot manipulation data. Tests visual understanding of diverse robotic scenes and actions.
Format
Each row contains:
{
"id": "<unique id>",
"question": "<question text>",
"choices": "['option1', 'option2', ...]",
"correct_answer": <int index>,
"image": <PNG bytes>
}
choices is a Python-literal string of a list of answer options.
correct_answer is the integer index into choices of the correct option.
image is embedded as binary in the parquet (HF Image() decodes on load).
Citation
@article{padalkar2023open,
title={Open X-Embodiment: Robotic Learning Datasets and RT-X Models},
author={Padalkar, Abhishek and ...},
journal={arXiv:2310.08864},
year={2023}
}
License
Annotations are derived from Open X-Embodiment, which is openly published. The MCQ formatting here is a research-only artifact.