@inproceedings{song-etal-2025-smart,
    title = "Smart-Searcher: Incentivizing the Dynamic Knowledge Acquisition of {LLM}s via Reinforcement Learning",
    author = "Song, Huatong  and
      Jiang, Jinhao  and
      Tian, Wenqing  and
      Chen, Zhipeng  and
      Wu, Yuhuan  and
      Zhao, Jiahao  and
      Min, Yingqian  and
      Zhao, Xin  and
      Fang, Lei  and
      Wen, Ji-Rong",
    editor = "Christodoulopoulos, Christos  and
      Chakraborty, Tanmoy  and
      Rose, Carolyn  and
      Peng, Violet",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2025",
    month = nov,
    year = "2025",
    address = "Suzhou, China",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/author-page-yu-wang-polytechnic/2025.findings-emnlp.731/",
    doi = "10.18653/v1/2025.findings-emnlp.731",
    pages = "13572--13586",
    ISBN = "979-8-89176-335-7",
    abstract = "Large Language Models (LLMs) are powerful but prone to hallucinations due to static knowledge. Retrieval-Augmented Generation (RAG) helps by injecting external information, but current methods often are costly, generalize poorly, or ignore the model{'}s internal knowledge.In this paper, we introduce Smart-Searcher, a novel framework designed to train LLMs to adaptively leverage both internal and external knowledge sources. Smart-Searcher employs a two-stage training strategy: an initial SFT Cold-start phase for preliminary format learning, followed by RL for Dynamic Knowledge Acquisition. The RL stage uses outcome-supervision to encourage exploration, incorporates a reward mechanism for internal knowledge utilization, and integrates a memorization mechanism to continuously assimilate retrieved information, thereby enriching the model{'}s internal knowledge. By leveraging internal knowledge and external search engine, the model continuously improves its capabilities, enabling efficient retrieval-augmented reasoning.Our experiments demonstrate that Smart-Searcher outperforms previous RAG and reasoning methods and achieves efficient retrieval.The code is available at \url{https://github.com/RUCAIBox/R1-Searcher-plus}."
}