README.md

---
configs:
  - config_name: all
    description: All puzzle types combined.
    data_files:
      - split: train
        path: puzzle_all_train.jsonl
      - split: test
        path: puzzle_all_test.jsonl
  - config_name: dots
    description: Puzzles primarily featuring the dot mechanic.
    data_files:
      - split: train
        path: puzzle_dots_train.jsonl
      - split: test
        path: puzzle_dots_test.jsonl
  - config_name: dots_stars_polys
    description: Puzzles combining dots, stars, and polyominoes.
    data_files:
      - split: train
        path: puzzle_dots_stars_polys_train.jsonl
      - split: test
        path: puzzle_dots_stars_polys_test.jsonl
  - config_name: gaps
    description: Puzzles primarily featuring the line gap mechanic.
    data_files:
      - split: train
        path: puzzle_gaps_train.jsonl
      - split: test
        path: puzzle_gaps_test.jsonl
  - config_name: gaps_dots_triangles
    description: Puzzles combining gaps, dots, and triangles.
    data_files:
      - split: train
        path: puzzle_gaps_dots_triangles_train.jsonl
      - split: test
        path: puzzle_gaps_dots_triangles_test.jsonl
  - config_name: polys
    description: Puzzles primarily featuring polyomino shapes.
    data_files:
      - split: train
        path: puzzle_polys_train.jsonl
      - split: test
        path: puzzle_polys_test.jsonl
  - config_name: polys_ylops
    description: Puzzles featuring polyomino shapes (reversed color logic - 'ylop').
    data_files:
      - split: train
        path: puzzle_polys_ylops_train.jsonl
      - split: test
        path: puzzle_polys_ylops_test.jsonl
  - config_name: stars
    description: Puzzles primarily featuring the star mechanic.
    data_files:
      - split: train
        path: puzzle_stars_train.jsonl
      - split: test
        path: puzzle_stars_test.jsonl
  - config_name: stones
    description: Puzzles primarily featuring the stone (separation) mechanic.
    data_files:
      - split: train
        path: puzzle_stones_train.jsonl
      - split: test
        path: puzzle_stones_test.jsonl
  - config_name: stones_stars
    description: Puzzles combining stones and stars.
    data_files:
      - split: train
        path: puzzle_stones_stars_train.jsonl
      - split: test
        path: puzzle_stones_stars_test.jsonl
  - config_name: triangles
    description: Puzzles primarily featuring the triangle mechanic.
    data_files:
      - split: train
        path: puzzle_triangles_train.jsonl
      - split: test
        path: puzzle_triangles_test.jsonl
size_categories:
- 1K<n<10K
---
<p align="center">
  <img src="logo.png" alt="SPaRC Logo" width="128"/>
</p>

# SPaRC Dataset

[Website](https://sparc.gipplab.org/) • [Solver](https://github.com/lkaesberg/SPaRCSolver) • [Generator](https://github.com/lkaesberg/SPaRCPuzzleGenerator)


A grid-based puzzle dataset for benchmarking LLMs spatial reasoning capabilities.

## Data Schema

Each record (JSON) includes:

* `id` (string): unique puzzle identifier
* `difficulty_level` (int) & `difficulty_score` (float)
* `grid_size`: `{ "height": H, "width": W }`
* `polyshapes`: JSON string mapping shape IDs to binary grids
* `puzzle_array`: 2D array with cell codes (e.g., `S`, `E`, `+`, `P-O-112`)
* `solution_count` (int) and `solutions` list (with `index`, `path`, `pathLength`)
* `text_visualization`: YAML-style summary

## Sample Entry

```json
{
  "id": "10e68dc3a6fbfcdf",
  "difficulty_level": 3,
  "difficulty_score": 2.3261,
  "grid_size": {"height":3,"width":4},
  "polyshapes":"{\"112\":[[0,1,...]]}",
  "puzzle_array":[
    ["+","+","+","E"],
    ["+","P-O-112","+","o-O"],
    ["S","N","+","G"]
  ],
  "solution_count": 10,
  "solutions": [ { "index":0, "pathLength":34, ... } ],
  "text_visualization": "..."
}
```

## Citation Information
If you use the dataset in any way, please cite the following paper. Preprint: https://arxiv.org/abs/2505.16686

```bib
@article{kaesberg2025sparc,
  title={SPaRC: A Spatial Pathfinding Reasoning Challenge},
  author={Kaesberg, Lars Benedikt and Wahle, Jan Philip and Ruas, Terry and Gipp, Bela},
  journal={arXiv preprint arXiv:2505.16686},
  year={2025}
}
```