Dataset Card for SciFIBench

Dataset Summary

The SciFIBench (Scientific Figure Interpretation Benchmark) contains 1000 multiple-choice scientific figure interpretation questions covering two tasks. Task 1: Figure -> Caption involves selecting the most appropriate caption given a figure; Task 2: Caption -> Figure involves the opposite -- selecting the most appropriate figure given a caption. This benchmark was curated from the SciCap dataset, using adversarial filtering to obtain hard negatives. Human verification has been performed on each question to ensure high-quality, answerable questions.

Example Usage

from datasets import load_dataset

# load dataset
dataset = load_dataset("jonathan-roberts1/SciFIBench") # optional: set cache_dir="PATH/TO/MY/CACHE/DIR"
# figure2caption_dataset = load_dataset("jonathan-roberts1/SciFIBench", split="Figure2Caption")
# caption2figure_dataset = load_dataset("jonathan-roberts1/SciFIBench", split="Caption2Figure")
"""
DatasetDict({
    Caption2Figure: Dataset({
        features: ['ID', 'Question', 'Options', 'Answer', 'Category', 'Images'],
        num_rows: 500
    })
    Figure2Caption: Dataset({
        features: ['ID', 'Question', 'Options', 'Answer', 'Category', 'Images'],
        num_rows: 500
    })
})
"""

# select task
figure2caption_dataset = dataset['Figure2Caption']
"""
Dataset({
    features: ['ID', 'Question', 'Options', 'Answer', 'Category', 'Images'],
    num_rows: 500
})
"""

# query items
figure2caption_dataset[40] # e.g., the 41st element
"""
{'ID': 40,
 'Question': 'Which caption best matches the image?',
 'Options': ['A)  ber vs snr for fft size=2048 using ls , lmmse , lr-lmmse .',
  'B)  ber vs snr for fft size=1024 using ls , lmmse , lr-lmmse algorithms .',
  'C)  ber vs snr for fft size=512 using ls , lmmse , lr-lmmse algorithms .',
  'D)  ber vs snr for fft size=256 using ls , lmmse , lr-lmmse algorithms with a 16 qam modulation .',
  'E)  ber vs snr for a bpsk modulation .'],
 'Answer': 'D',
 'Category': 'other cs',
 'Images': [<PIL.PngImagePlugin.PngImageFile image mode=RGB size=501x431>]}
"""

Source Data

More information regarding the source data can be found at: https://github.com/tingyaohsu/SciCap

Dataset Curators

This dataset was curated by Jonathan Roberts, Kai Han, Neil Houlsby, and Samuel Albanie

Citation Information

@article{roberts2024scifibench,
  title={SciFIBench: Benchmarking Large Multimodal Models for Scientific Figure Interpretation},
  author={Roberts, Jonathan and Han, Kai and Houlsby, Neil and Albanie, Samuel},
  journal={arXiv preprint arXiv:2405.08807},
  year={2024}
}