metadata

license: mit
size_categories:
  - 1K<n<10K
task_categories:
  - visual-question-answering
pretty_name: VisualPuzzles
dataset_info:
  features:
    - name: id
      dtype: int64
    - name: category
      dtype: string
    - name: image
      dtype: image
    - name: question
      dtype: string
    - name: options
      sequence: string
    - name: answer
      dtype: string
  splits:
    - name: train
      num_bytes: 139582416.624
      num_examples: 1168
  download_size: 137679574
  dataset_size: 139582416.624
configs:
  - config_name: default
    data_files:
      - split: train
        path: data.parquet

VisualPuzzles: Decoupling Multimodal Reasoning Evaluation from Domain Knowledge

Overview

VisualPuzzles is a multimodal benchmark specifically designed to evaluate reasoning abilities in large models while deliberately minimizing reliance on domain-specific knowledge.

Key features:

1168 diverse puzzles
5 reasoning categories: Algorithmic, Analogical, Deductive, Inductive, Spatial
Difficulty labels: Easy, Medium, Hard
Less knowledge-intensive than existing benchmarks (e.g., MMMU)
More reasoning-complex than existing benchmarks (e.g., MMMU)

Key Findings

All models perform worse than humans; most can't surpass even 5th-percentile human performance.
Strong performance on knowledge-heavy benchmarks does not transfer well.
Larger models and structured "thinking modes" don't guarantee better results.
Scaling model size does not ensure stronger reasoning

Usage

To load this dataset via Hugging Face’s datasets library:

from datasets import load_dataset

dataset = load_dataset("neulab/VisualPuzzles")
data = dataset["train"]

sample = data[0]
print("ID:", sample["id"])
print("Category:", sample["category"])
print("Question:", sample["question"])
print("Options:", sample["options"])
print("Answer:", sample["answer"])

Citation

If you use or reference this dataset in your work, please cite:

@article{song2025visualpuzzles,
  title         = {VisualPuzzles: Decoupling Multimodal Reasoning Evaluation from Domain Knowledge},
  author        = {Song, Yueqi and Ou, Tianyue and Kong, Yibo and Li, Zecheng and Neubig, Graham and Yue, Xiang},
  year          = {2025},
  journal       = {arXiv preprint arXiv:2504.10342},
  url           = {https://arxiv.org/abs/2504.10342}
}