README.md

metadata

dataset_info:
  features:
    - name: data_source
      dtype: string
    - name: prompt
      list:
        - name: content
          dtype: string
        - name: role
          dtype: string
    - name: images
      list:
        - name: bytes
          dtype: binary
    - name: ability
      dtype: string
    - name: env_name
      dtype: string
    - name: reward_model
      struct:
        - name: ground_truth
          dtype: string
        - name: style
          dtype: string
    - name: extra_info
      struct:
        - name: answer
          dtype: string
        - name: index
          dtype: string
        - name: question
          dtype: string
        - name: split
          dtype: string
    - name: randomized_to_original
      dtype: string
  splits:
    - name: train
      num_bytes: 1971977439
      num_examples: 15000
    - name: validation
      num_bytes: 74206529
      num_examples: 500
  download_size: 1931231003
  dataset_size: 2046183968
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
task_categories:
  - image-text-to-text
license: apache-2.0
language:
  - en
tags:
  - visual-reasoning
  - tool-use
  - multimodal
  - mllm

AdaReasoner Dataset

Project Page | Paper | GitHub

AdaReasoner is a family of multimodal models that learn tool use as a general reasoning skill rather than as tool-specific behavior. This dataset provides the training and evaluation data used to enable long-horizon, multi-step tool interactions. It was curated to help models infer tool utility from task context and intermediate outcomes, enabling the coordination of multiple tools and generalization to unseen ones.

🧩 Data Format

The data is stored in Parquet format. According to the official documentation, each item in the dataset typically follows this structure:

prompt = [
    {
        "content": system_prompt,
        "role": "system"
    },
    {
        "content": f"{question_text}",
        "role": "user"
    }
]
item = {
    "data_source": "jigsaw_coco",
    "prompt": prompt,
    "images": [{"bytes": question_image_bytes}] + choice_images,
    "ability": "visual_reasoning",
    "env_name": "jigsaw",
    "reward_model": {
        "ground_truth": correct_letter.lower(),
        "style": "model"
    },
    "extra_info": { # Used for reward calculation
        "extra_info1": "...",
    }
}

📚 Citation

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

@article{song2026adareasoner,
  title={AdaReasoner: Dynamic Tool Orchestration for Iterative Visual Reasoning},
  author={Song, Mingyang and Sun, Haoyu and Gu, Jiawei and Li, Linjie and Xu, Luxin and Krishna, Ranjay and Cheng, Yu},
  journal={arXiv preprint arXiv:2601.18631},
  year={2026}
}

📄 License

This dataset is licensed under the Apache 2.0 License.

🤝 Acknowledgments

This model is part of the AdaReasoner project. For more information, visit our GitHub repository.

📧 Contact

For questions and feedback, please open an issue in our GitHub repository.