While generative retrieval (GR) demonstrates competitive performance on standard retrieval benchmarks, existing approaches directly map queries to document identifiers (docids) without intermediate deliberation, limiting their effectiveness for complex queries that require multi-step reasoning. As a preliminary study on integrating chain-of-thought (CoT) into generative retrieval, we introduce ThinkGR, a unified framework that interleaves CoT with docid generation, enabling iterative thinking and retrieval within a single generative process. To bridge the gap between free-form thought generation and structured retrieval targets, we design (1) a hybrid decoding strategy that dynamically switches between unconstrained thought generation and constrained docid decoding, and (2) a two-phase training approach that first aligns thought-retrieval patterns through supervised fine-tuning, then optimizes thought quality via retrieval-grounded reinforcement learning. Experiments on four multi-hop retrieval benchmarks demonstrate that ThinkGR achieves state-of-the-art performance with an average improvement of +6.86\%. Our work opens new avenues for enhancing generative retrieval with explicit deliberation capabilities, with promising implications for retrieval tasks requiring complex reasoning.

翻译：尽管生成式检索（GR）在标准检索基准上展现了竞争性能，但现有方法直接将查询映射到文档标识符（docid），缺乏中间推演过程，限制了其对需要多步推理的复杂查询的有效性。作为将链式思维（CoT）融入生成式检索的初步研究，我们提出了ThinkGR——一个将链式思维与文档标识符生成交织的统一框架，能够在单一生成过程中实现迭代思考与检索。为弥合自由形式思维生成与结构化检索目标之间的差距，我们设计了：（1）混合解码策略，在无约束思维生成与受约束文档标识符解码之间动态切换；（2）两阶段训练方法，先通过监督微调对齐思考-检索模式，再通过基于检索的强化学习优化思维质量。在四个多跳检索基准上的实验表明，ThinkGR实现了最先进的性能，平均提升+6.86%。我们的工作为通过显式推演能力增强生成式检索开辟了新路径，对需要复杂推理的检索任务具有重要启示。