SLM 遇见 LLM：在幻觉检测中平衡延迟、可解释性与一致性
SLM Meets LLM: Balancing Latency, Interpretability and Consistency in Hallucination Detection
AI总结：
针对 LLM 在实时幻觉检测中的延迟问题，提出结合 SLM 分类器与 LLM 受限推理器的新框架，通过提示技术对齐两者决策与解释，有效优化了实时可解释检测并提升用户体验。
摘要：
大型语言模型（LLMs）能力强大，但在实时应用中面临延迟挑战，如进行在线幻觉检测。为克服此问题，我们提出一种新框架，利用小型语言模型（SLM）分类器进行初步检测，随后由 LLM 作为受限推理器为检测到的幻觉内容生成详细解释。本研究通过引入有效的提示技术，使 LLM 生成的解释与 SLM 决策对齐，从而优化了实时可解释的幻觉检测。实验结果证明了其有效性，进而提升了整体用户体验。
英文摘要：
Large language models (LLMs) are highly capable but face latency challenges in real-time applications, such as conducting online hallucination detection. To overcome this issue, we propose a novel framework that leverages a small language model (SLM) classifier for initial detection, followed by a LLM as constrained reasoner to generate detailed explanations for detected hallucinated content. This study optimizes the real-time interpretable hallucination detection by introducing effective prompting techniques that align LLM-generated explanations with SLM decisions. Empirical experiment results demonstrate its effectiveness, thereby enhancing the overall user experience.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
