@inproceedings{lee-etal-2025-token,
    title = "From Token to Action: State Machine Reasoning to Mitigate Overthinking in Information Retrieval",
    author = "Lee, Dohyeon  and
      Jeong, Yeonseok  and
      Hwang, Seung-won",
    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.371/",
    doi = "10.18653/v1/2025.findings-emnlp.371",
    pages = "7048--7064",
    ISBN = "979-8-89176-335-7",
    abstract = "Chain-of-Thought (CoT) prompting enables complex reasoning in large language models (LLMs), including applications in information retrieval (IR). However, it often leads to overthinking, where models produce excessively long and semantically redundant traces with little or no benefit. We identify two key challenges in IR: redundant trajectories that revisit similar states and misguided reasoning that diverges from user intent. To address these, we propose State Machine Reasoning (SMR), a transition-based reasoning framework composed of discrete actions (REFINE, RERANK, STOP) that support early stopping and fine-grained control. Experiments on the BEIR and BRIGHT benchmarks show that improves retrieval performance (nDCG@10) by 3.4{\%} while reducing token usage by 74.4{\%}. It generalizes across LLMs and retrievers without requiring task-specific tuning, offering a practical alternative to conventional CoT reasoning."
}