大型语言模型在分类决策中对齐有效性的系统表征
Systematic Characterization of the Effectiveness of Alignment in Large Language Models for Categorical Decisions
AI总结：
本文提出对齐合规指数（ACI），并以医疗分诊为用例系统评估GPT4o、Claude 3.5 Sonnet和Gemini Advanced在分类决策中与人类偏好的对齐有效性，揭示不同模型和对齐方法间存在显著差异。
摘要：
随着大型语言模型（LLMs）被部署到医疗保健等高风险领域，理解它们的决策与人类偏好和价值观的对齐程度变得至关重要，尤其是当我们认识到这些偏好并不存在单一的金标准时。本文应用一种系统化方法，以医疗分诊作为领域特定的用例，评估LLMs在分类决策中的偏好对齐情况。本文还衡量了对齐流程在多大程度上能有效改变特定模型的对齐状态。该方法的关键是一个新颖且简单的度量指标——对齐合规指数（ACI），它量化了一个LLM能够被对齐到给定偏好函数或金标准的有效程度。由于ACI衡量的是对齐的效果而非对齐的过程，因此它适用于本研究中所使用的上下文学习（in-context learning）之外的对齐方法。\n使用一个模拟患者对的数据集，对三个前沿LLMs（GPT4o、Claude 3.5 Sonnet和Gemini Advanced）进行了评估，考察它们做出与临床专家偏好一致的分诊决策的能力。通过多种提示策略评估了模型在对齐尝试前后的表现。结果揭示了不同模型和对齐方法之间对齐有效性的显著差异。值得注意的是，根据ACI衡量，对齐前表现良好的模型在对齐后有时会出现性能下降，并且目标偏好函数的微小变化会导致模型排名的巨大变动。研究还通过有针对性的提问，探索了人类所理解的、支撑LLMs决策的隐含伦理原则。\n本研究推动了在近期使用一套实用的方法和ACI，以理解在诸如分诊等分类决策中，各种人类与LLM决策价值观之间的对应关系。
英文摘要：
As large language models (LLMs) are deployed in high-stakes domains like healthcare, understanding how well their decision-making aligns with human preferences and values becomes crucial, especially when we recognize that there is no single gold standard for these preferences. This paper applies a systematic methodology for evaluating preference alignment in LLMs on categorical decision-making with medical triage as a domain-specific use case. It also measures how effectively an alignment procedure will change the alignment of a specific model. Key to this methodology is a novel simple measure, the Alignment Compliance Index (ACI), that quantifies how effectively a LLM can be aligned to a given preference function or gold standard. Since the ACI measures the effect rather than the process of alignment, it is applicable to alignment methods beyond the in-context learning used in this study. Using a dataset of simulated patient pairs, three frontier LLMs (GPT4o, Claude 3.5 Sonnet, and Gemini Advanced) were assessed on their ability to make triage decisions consistent with an expert clinician's preferences. The models' performance before and after alignment attempts was evaluated using various prompting strategies. The results reveal significant variability in alignment effectiveness across models and alignment approaches. Notably, models that performed well, as measured by ACI, pre-alignment sometimes degraded post-alignment, and small changes in the target preference function led to large shifts in model rankings. The implicit ethical principles, as understood by humans, underlying the LLMs' decisions were also explored through targeted questioning. This study motivates the use of a practical set of methods and the ACI, in the near term, to understand the correspondence between the variety of human and LLM decision-making values in categorical decision-making such as triage.
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。
