---
license: cc-by-4.0
pretty_name: COSPLAN
configs:
- config_name: checkboard
  data_files:
  - split: test
    path: checkboard/metadata.jsonl
- config_name: blocksworld
  data_files:
  - split: test
    path: blocksworld/metadata.jsonl
- config_name: imagenet
  data_files:
  - split: test
    path: imagenet/metadata.jsonl
- config_name: robovqa
  data_files:
  - split: test
    path: robovqa/metadata.jsonl
---

# COSPLAN

A composite visual-planning benchmark with four multiple-choice subsets. Each
subset is a separate config in HuggingFace `imagefolder` format (images +
`metadata.jsonl`). Every row has an `image`, a `prompt`, and a `correct_option`
(A/B/C/D).

| Config | Items | Task |
|--------|-------|------|
| `checkboard`   | 4896 | Grid path planning around obstacles |
| `blocksworld`  | 5000 | Block-stacking move-sequence selection |
| `imagenet`     | 1000 | Image-patch rearrangement validation |
| `robovqa`      |  300 | Robot action-sequence selection |

## Usage

```python
from datasets import load_dataset
ds = load_dataset("shrg7/COSPLAN", "blocksworld", split="test")
print(ds[0]["prompt"], ds[0]["correct_option"])
```