README.md

metadata

license: mit
task_categories:
  - image-to-text
  - text-to-image
language:
  - en
pretty_name: simons ARC (abstraction & reasoning corpus) solve augment version 3
size_categories:
  - 10K<n<100K
configs:
  - config_name: default
    data_files:
      - split: train
        path: data.jsonl

Version 1

Augmentation of the ARC-AGI tasks.

example count: 1-3.

test count: 1.

Version 2

Only skew up/down/left/right

Version 3

Instead of making tasks out of input/output images.

I'm now focusing on preserving the original puzzle, with some transformations applied.