@inproceedings{zhou-etal-2026-toolgrad,
    title = "{T}ool{G}rad: Efficient Tool-use Dataset Generation with Textual ``Gradients''",
    author = "Zhou, Zhongyi  and
      Uehara, Kohei  and
      Zhang, Haoyu  and
      Zhou, Jingtao  and
      Gu, Lin  and
      Du, Ruofei  and
      Xu, Zheng  and
      Harada, Tatsuya",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.findings-acl.950/",
    pages = "19040--19056",
    ISBN = "979-8-89176-395-1",
    abstract = "Prior work synthesizes tool-use LLM datasets by first generating a user query, followed by complex tool-use annotations like DFS. This inherently leads to inevitable annotation failures and low efficiency in data generation. We introduce ToolGrad, an agentic framework that inverts this paradigm. ToolGrad first constructs valid tool-use chains through an iterative process guided by textual ``gradients'', and then synthesizes corresponding user queries. This ``answer-first'' approach led to ToolGrad-500, a dataset generated with more complex tool use, lower cost, and almost 100{\%} pass rate. Experiments show that models trained on ToolGrad-500 outperform those trained on expensive baseline datasets and proprietary LLMs."
}