根据用户意图动态调整外部信息信任度，提升生成准确性与可靠性。
When Context Misleads: Intent-Guided Decoding for Robust Retrieval-Augmented Generation
- 通过用户意图判断何时信任检索内容，动态调节生成路径。
- 在事实冲突测试中相比直接RAG最高提升65.4个百分点的准确率。
- 适合需要高事实正确性且兼顾上下文忠实性的问答场景。
检索增强生成（RAG）通过引入外部证据提升大模型表现，但检索内容可能有用、无关甚至误导。现有系统对检索信息采用固定信任策略，导致错误信息被过度依赖或关键信息被忽略。为此，我们提出意图引导解码（IGD）框架，根据用户意图在检索内容与参数记忆间动态权衡。IGD通过答案级过滤和词元级修正，引导最终生成轨迹。我们在三个忠实问答基准和三个事实冲突基准上评估了五种大模型，结果显示IGD显著提升事实恢复能力，在事实冲突任务中相比直接RAG最高提升65.4个百分点，同时保持或改善严格遵循上下文的行为。该结果凸显了在RAG中平衡事实准确性和上下文忠实性的重要性。
原文摘要 · Abstract (English)
Retrieval-augmented generation (RAG) improves large language models by grounding generation in external evidence, but it also introduces a source trust problem: retrieved context may be useful, irrelevant, or even misleading. Existing RAG systems often apply a fixed trust policy toward retrieved evidence, which can either over-trust incorrect context or underuse context when the user explicitly asks for context-following behavior. Therefore, we propose Intent-Guided Decoding (IGD), a framework that arbitrates between retrieved context and parametric memory according to user intent. IGD uses answer-level filtering and token-level correction to steer the final decoding trajectory between retrieved context and parametric memory. We evaluate IGD on three faithful QA benchmarks and three factual-conflict benchmarks across five LLMs, IGD substantially improves factual recovery, achieving gains of up to 65.4 percentage points on factual-conflict benchmarks over Direct RAG, while preserving or improving strict context-following behavior, this findings highlight the importance of balancing factuality and faithfulness in RAG.
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。