通过非典型表现重校准增强医疗大语言模型可信度
Enhancing Healthcare LLM Trust with Atypical Presentations Recalibration
AI总结：
本文针对黑盒大语言模型在医疗场景中的过度自信问题，提出利用非典型表现调整置信度估计的重校准方法，在三个医学问答数据集上将校准误差降低约60%，显著优于现有方法。
摘要：
黑盒大语言模型（LLMs）正越来越多地部署于各种环境中，因此这些模型必须能够有效传达其置信度与不确定性，尤其是在高风险场景下。然而，这些模型往往表现出过度自信，从而带来潜在风险和误判。现有的用于引导和校准大语言模型置信度的技术主要集中于通用推理数据集，仅取得了有限的改进。准确的校准对于知情决策和防止不良后果至关重要，但由于这些模型所执行任务的复杂性和多变性，校准仍然具有挑战性。在本工作中，我们研究了黑盒大语言模型在医疗环境中的误校准行为。我们提出了一种新方法，即非典型表现重校准（Atypical Presentations Recalibration），该方法利用非典型表现来调整模型的置信度估计。我们的方法显著改善了校准效果，在三个医学问答数据集上将校准误差降低了约60%，并优于现有方法，如普通语言化置信度（vanilla verbalized confidence）、思维链语言化置信度（CoT verbalized confidence）等。此外，我们还对非典型性在重校准框架中的作用进行了深入分析。
英文摘要：
Black-box large language models (LLMs) are increasingly deployed in various environments, making it essential for these models to effectively convey their confidence and uncertainty, especially in high-stakes settings. However, these models often exhibit overconfidence, leading to potential risks and misjudgments. Existing techniques for eliciting and calibrating LLM confidence have primarily focused on general reasoning datasets, yielding only modest improvements. Accurate calibration is crucial for informed decision-making and preventing adverse outcomes but remains challenging due to the complexity and variability of tasks these models perform. In this work, we investigate the miscalibration behavior of black-box LLMs within the healthcare setting. We propose a novel method, \textit{Atypical Presentations Recalibration}, which leverages atypical presentations to adjust the model's confidence estimates. Our approach significantly improves calibration, reducing calibration errors by approximately 60\% on three medical question answering datasets and outperforming existing methods such as vanilla verbalized confidence, CoT verbalized confidence and others. Additionally, we provide an in-depth analysis of the role of atypicality within the recalibration framework.
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。
