README.md

---
license: mit
task_categories:
  - visual-question-answering
  - question-answering
language:
  - en
tags:
  - multimodal
  - agents
  - tool-use
  - benchmark
  - visualtoolbench
size_categories:
  - 1K<n<10K
---

# VisualToolBench (XSkill-formatted)

This is **VisualToolBench** (originally from `ScaleAI/VisualToolBench`) repackaged
into the JSON layout consumed by [XSkill](https://github.com/XSkill-Agent/XSkill)'s
evaluation pipeline (`eval/infer_api.py`). All images are extracted from the
original parquet snapshots and stored as PNG files referenced by relative paths
in the JSON splits.

## Contents

A single archive `VisualToolBench.zip` (~3.7 GB) containing:

```
VisualToolBench/
├── val_50.json        #   50 sanity-check samples
├── val_single.json    # ~603 single-turn samples
├── val_multi.json     # ~601 multi-turn samples
├── val_full.json      # 1204 full samples (single + multi)
└── images/
    └── <doc_id>/img_*.png    # one folder per sample
```

Each sample in the JSON files follows XSkill's expected schema:

```json
{
  "doc_id":      "<unique-id>",
  "problem":     "<image>\n<question text>",
  "images":      ["VisualToolBench/images/<doc_id>/img_0.png", ...],
  "solution":    "<gold answer>",
  "data_source": "<original prompt category>"
}
```

The `<image>` placeholder marks where each image is to be injected during
prompt assembly; the order matches the order of paths in `images`.

## Usage

### Download

```bash
huggingface-cli download wan288972153/VisualToolBench-XSkill \
    VisualToolBench.zip \
    --repo-type dataset \
    --local-dir .
unzip VisualToolBench.zip -d ./
# → ./VisualToolBench/  (contains the JSON splits + images/)
```

### Plug into XSkill

Either drop the `VisualToolBench/` folder under `<XSkill>/benchmark/`, or
point the ablation script to wherever you put it:

```bash
VTB_DATA_DIR=/path/to/VisualToolBench bash scripts_local/run_ablation.sh
```

## Source

- Original benchmark: [ScaleAI/VisualToolBench](https://huggingface.co/datasets/ScaleAI/VisualToolBench)
- Conversion script: see `scripts_local/convert_visualtoolbench.py` in the XSkill repo

## Citation

If you use this data, please cite the original VisualToolBench authors and the
XSkill paper:

```bibtex
@misc{jiang2026xskillcontinuallearningexperience,
  title  = {XSkill: Continual Learning from Experience and Skills in Multimodal Agents},
  author = {Guanyu Jiang and Zhaochen Su and Xiaoye Qu and Yi R. Fung},
  year   = {2026},
  eprint = {2603.12056},
  archivePrefix = {arXiv},
  primaryClass  = {cs.AI}
}
```