让大模型主动操控检索过程，突破传统黑箱查询限制。
Interact-RAG: Reason and Interact with the Corpus, Beyond Black-Box Retrieval
- 设计可交互的语料操作引擎，实现对检索的精细控制。
- 在6个基准上超越现有方法，显著提升复杂任务表现。
- 适合需要深度信息探索的智能体研发与系统构建者。
检索增强生成（RAG）通过引入外部信息显著提升了大语言模型的能力。然而，当前基于智能体的RAG方法受限于将检索过程视为黑箱查询操作，导致智能体仅能发出查询指令，难以应对复杂的寻址任务。为此，我们提出Interact-RAG，一种新范式，使大模型智能体从被动查询者转变为检索过程的主动操控者。通过引入语料交互引擎，赋予智能体一系列动作原语，实现对信息检索的细粒度控制。为进一步赋能整个RAG流程，我们首先构建了增强推理的工作流，支持零样本执行及交互轨迹合成；随后利用合成数据通过监督微调（SFT）训练全自主端到端智能体，并结合强化学习（RL）进行优化。在六个基准上的广泛实验表明，Interact-RAG显著优于其他先进方法，验证了推理-交互策略的有效性。
原文摘要 · Abstract (English)
Retrieval-Augmented Generation (RAG) has significantly enhanced LLMs by incorporating external information. However, prevailing agentic RAG approaches are constrained by a critical limitation: they treat the retrieval process as a black-box querying operation. This confines agents' actions to query issuing, hindering its ability to tackle complex information-seeking tasks. To address this, we introduce Interact-RAG, a new paradigm that elevates the LLM agent from a passive query issuer into an active manipulator of the retrieval process. We dismantle the black-box with a Corpus Interaction Engine, equipping the agent with a set of action primitives for fine-grained control over information retrieval. To further empower the agent on the entire RAG pipeline, we first develop a reasoning-enhanced workflow, which enables both zero-shot execution and the synthesis of interaction trajectories. We then leverage this synthetic data to train a fully autonomous end-to-end agent via Supervised Fine-Tuning (SFT), followed by refinement with Reinforcement Learning (RL). Extensive experiments across six benchmarks demonstrate that Interact-RAG significantly outperforms other advanced methods, validating the efficacy of our reasoning-interaction strategy.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。