ProMoral-Bench：评估大语言模型中道德推理与安全性的提示策略
ProMoral-Bench: Evaluating Prompting Strategies for Moral Reasoning and Safety in LLMs
AI总结：
ProMoral-Bench通过统一道德安全评分评估多种提示策略，发现紧凑示例引导框架在道德和安全方面表现更优。
摘要：
提示设计显著影响大型语言模型（LLMs）的道德能力和安全对齐，但现有实证比较在数据集和模型之间仍显碎片化。我们引入ProMoral-Bench，一个统一的基准，评估11种提示范式跨四个LLM家族的表现。利用ETHICS、Scruples、WildJailbreak以及我们新提出的鲁棒性测试ETHICS-Contrast，我们通过所提出的统一道德安全分数（UMSS）来衡量性能，该指标平衡了准确性和安全性。我们的结果表明，紧凑的示例引导框架优于复杂的多阶段推理，提供了更高的UMSS分数和更强的鲁棒性，同时以更低的token成本。虽然多轮推理在扰动下显得脆弱，但少量示例在增强道德稳定性和对抗性抵抗方面表现一致。ProMoral-Bench建立了标准化的框架，用于原则性、成本效益的提示工程。
英文摘要：
Prompt design significantly impacts the moral competence and safety alignment of large language models (LLMs), yet empirical comparisons remain fragmented across datasets and models.We introduce ProMoral-Bench, a unified benchmark evaluating 11 prompting paradigms across four LLM families. Using ETHICS, Scruples, WildJailbreak, and our new robustness test, ETHICS-Contrast, we measure performance via our proposed Unified Moral Safety Score (UMSS), a metric balancing accuracy and safety. Our results show that compact, exemplar-guided scaffolds outperform complex multi-stage reasoning, providing higher UMSS scores and greater robustness at a lower token cost. While multi-turn reasoning proves fragile under perturbations, few-shot exemplars consistently enhance moral stability and jailbreak resistance. ProMoral-Bench establishes a standardized framework for principled, cost-effective prompt engineering.
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。
