@inproceedings{zhang-etal-2026-searchgym,
    title = "{S}earch{G}ym: Bootstrapping Real-World Search Agents via Cost-Effective and High-Fidelity Environment Simulation",
    author = "Zhang, Xichen  and
      He, Ziyi  and
      Zhu, Yinghao  and
      Wu, Sitong  and
      Yu, Shaozuo  and
      Chu, Meng  and
      Zhang, Wenhu  and
      Tan, Haoru  and
      Jia, Jiaya",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.acl-long.848/",
    pages = "18633--18665",
    ISBN = "979-8-89176-390-6",
    abstract = "Search agents have emerged as a pivotal paradigm for solving open-ended, knowledge-intensive reasoning tasks. However, training these agents via Reinforcement Learning (RL) faces a critical dilemma: interacting with live commercial Web APIs is prohibitively expensive, while relying on static data snapshots often introduces noise due to data misalignment. This misalignment generates corrupted reward signals that destabilize training by penalizing correct reasoning or rewarding hallucination. To address this, we propose SearchGym, a simulation environment designed to bootstrap robust search agents. SearchGym employs a rigorous generative pipeline to construct a verifiable knowledge graph and an aligned document corpus, ensuring that every reasoning task is factually grounded and strictly solvable. Building on this controllable environment, we introduce SearchGym-RL, a curriculum learning methodology that progressively optimizes agent policies through purified feedback, evolving from basic interactions to complex, long-horizon planning. Extensive experiments across the Llama and Qwen families demonstrate strong Sim-to-Real generalization. Notably, our Qwen2.5-7B-Base model trained within SearchGym surpasses the web-enhanced ASearcher baseline across nine diverse benchmarks by an average relative margin of 10.6{\%}. Our results validate that high-fidelity simulation serves as a scalable and highly cost-effective methodology for developing capable search agents."
}