发表机构
Indian Institute of Technology, Roorkee（鲁尔基印度理工学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究针对隐私受限场景下的长文本生成难题，提出基于本地语言模型的多智能体框架RH-RAG，通过三阶段协同生成与双层检索索引提升生成质量，且兼顾数据隐私。
AI 中文摘要
在严格的隐私和安全约束下运营的组织中，从大量内部报告生成长篇内容仍然具有挑战性，专有基于云的大语言模型（LLM）API通常不可行。虽然本地部署的开放权重模型提供了一种隐私保护的替代方案，但现有针对较小模型的检索增强生成（RAG）方法往往缺乏有效的全局规划，且在长输出过程中会积累事实不一致性。为解决这些局限，我们提出RH-RAG，这是一个使用本地语言模型实现安全可信长文本生成的多智能体框架。RH-RAG将生成分解为三个协同阶段：规划智能体从高级语义摘要构建全局文档大纲；写作智能体使用受限连贯性记忆逐步生成连贯的章节内容；校验智能体通过基于自然语言推理的事实验证和基于证明的修正循环缓解幻觉。该框架还采用双层检索索引，支持在消费级硬件上进行高效规划和细粒度上下文生成。在文学、金融和法律领域的评估表明，与标准和分层RAG基线相比，RH-RAG持续提升事实依据、语义连贯性和文档级对齐度，同时在不损害数据隐私的情况下达到与专有云系统相当的可靠性。
英文摘要
Generating long-form content from extensive internal reports remains challenging for organizations operating under strict privacy and security constraints, where proprietary cloud-based LLM APIs are often not viable. While locally deployed open-weight models offer a privacy-preserving alternative, existing retrieval-augmented generation (RAG) approaches on smaller models frequently lack effective global planning and accumulate factual inconsistencies over long outputs. To address these limitations, we present RH-RAG, a multi-agent framework for secure and trustworthy long form generation using local language models. RH-RAG decomposes generation into three coordinated stages: a Planner Agent that constructs a global document outline from high-level semantic summaries, a Writer Agent that incrementally generates coherent section-wise content using bounded coherence memory, and a Checker Agent that mitigates hallucinations through natural language inference-based factual verification and an attestation-driven revision loop. The framework further employs a dual-level retrieval index that supports efficient planning and fine-grained contextual generation on consumer-grade hardware. Evaluations across literary, financial, and legal domains demonstrate that RH-RAG consistently improves factual grounding, semantic coherence, and document-level alignment compared to standard and hierarchical RAG baselines, while achieving reliability competitive with proprietary cloud-based systems without compromising data privacy.
Commentsaccepted in KDD 2026 SeT-LLM Workshop
