用可解释的反馈机制，让大模型生成更符合证据的内容。
COMPASS: Context-Modulated PID Attention Steering System for Hallucination Mitigation
- 通过上下文依赖得分实时监控注意力分配，识别偏离证据的生成行为。
- 在多个评测中将幻觉率降低2.8至5.8个百分点，无需重训练。
- 适合关注模型可解释性与事实性提升的研究者和工程师。
大语言模型虽能生成流畅文本，却常因过度依赖参数知识而产生事实错误。本文提出轻量级可解释控制框架COMPASS，通过内置反馈环路在解码过程中动态调节注意力。该系统引入透明指标——上下文依赖得分（CRS），在线监测注意力头是否基于证据生成内容。利用此可解释信号，PID控制器实时调整注意力，确保生成事实一致，无需重新训练或多轮解码。在HotpotQA、XSum、HaluEval、RAGTruth等基准上，幻觉率绝对降低2.8至5.8个百分点，同时揭示了不同注意力头对证据对齐的贡献差异。结果表明，基于反馈的可解释性是理解大模型行为的关键路径。
原文摘要 · Abstract (English)
Large language models (LLMs) often generate fluent but factually incorrect statements despite having access to relevant evidence, a failure mode rooted in how they allocate attention between contextual and parametric knowledge. Understanding and steering this internal behavior is key both for trustworthy deployment and for scientific interpretability of model mechanisms. We introduce COMPASS (Context-Modulated PID Attention Steering System), a lightweight, interpretable control framework that embeds a model-based feedback loop directly within decoding. COMPASS quantifies context reliance via a transparent metric, the Context Reliance Score (CRS), which serves as an online probe of how attention heads ground generation in evidence. Using this interpretable signal, a PID controller dynamically modulates attention heads to maintain factual consistency without retraining or multi-pass decoding. Across benchmarks (HotpotQA, XSum, HaluEval, RAGTruth), COMPASS consistently reduces contextual hallucination rates (2.8 to 5.8 percent absolute) while revealing how distinct attention heads contribute to evidence alignment. These results highlight feedback-driven interpretability as a pathway toward scientific understanding of LLM behavior.
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。