Visual Graph Arena Dataset

Welcome to the Visual Graph Arena dataset page!

The Visual Graph Arena Benchmark is a collection of six datasets designed to evaluate and enhance the visual reasoning capabilities of AI. The datasets are structured around three primary concepts, each divided into two tasks: (for more information visit Tasks)

  1. Graph Isomorphism Tasks:
    • Easy Isomorphism: Determining whether two given graphs are isomorphic, with non-isomorphic graphs chosen randomly.
    • Hard Isomorphism: Determining whether two given graphs are isomorphic, with non-isomorphic graphs being degree-equivalent.
  2. Graph Path Finding Tasks:
    • Hamiltonian Path: Determining whether a given graph contains a Hamiltonian path.
    • Shortest Path: Finding the shortest path length between two marked nodes in a graph.
  3. Graph Cycle Finding Tasks:
    • Hamiltonian Cycle: Determining whether a given graph contains a Hamiltonian cycle.
    • Biggest Chordless Cycle: Finding the size of the largest chordless cycle in a graph.
Research Paper
Paper will be available for
download upon publication. Square Image Download
Paper Here

Download

To download the Visual Graph Arena dataset, click the button below:

visual_graph_arena.zip

Citation

If you use this dataset in your research, please cite the following paper:


Citation will be available upon publication.

Contact

If you have any questions or feedback regarding the Visual Graph Arena dataset, please contact us at zahra{at}csail{dot}mit{dot}edu.

DOI

On the day I was creating this page, I received the bitter news of my grandmother's passing, coming just a month after we lost her son, also my uncle, to cancer. With a deep sadness and sorrow, I write this to their memories. Ziba and Ehsan.
- Peyman