审计具有文化特定道德梯度的LLM驱动社交机器人
Auditing LLM-Governed Social Robots with Culture-Specific Moral Gradients
AI总结：
针对LLM驱动社交机器人在跨文化场景中优先分配资源时的道德偏差，提出基于梯度的多语言审计框架，通过对称控制场景测试模型对文化偏好梯度的区分能力，发现提示工程无法可靠纠正的持续性不对称问题。
摘要：
LLM驱动的社交机器人越来越多地决定谁先获得现实世界中的帮助。由于不同文化中基于年龄、地位和群体规模的优先顺序规范不同，未能多元校准可能导致不平等的扩大。然而，LLM道德审计仍以英语为中心，很少测试具身化情境，在LLM-机器人部署日益加剧的背景下，多元校准成为一个紧迫的诊断缺口。我们引入一个基于梯度的审计框架，用于对LLM道德权衡行为进行多语言评估，以文化偏好梯度为基准。基于九项跨领域社交机器人综述（>8000篇论文），我们推导出护理、教育和服务领域的对称控制场景，将“道德机器实验”中的“拯救谁”转化为“先帮助谁”的困境，并保留身份权衡（多数vs少数；年轻vs年老；高地位vs低地位）。我们在四种提示模式下对四个LLM进行跨四个国家-语言对的审计（57,600个决策），并以国家特定的MME偏好梯度为基准。通过序数一致性检验模型是否能区分文化背景；治理类型学映射梯度区分、方向趋势和 deliberative 方面的脆弱性。我们发现存在持续的、文化不对称的梯度跟踪失败，仅靠提示无法可靠纠正：西方语言决策的质量校准几乎是中文和日文的两倍；多数优先权衡中的高确定性往往消除跨文化梯度；对年龄和地位规范的局部敏感性可能边缘化少数群体。提示效果不均，只有对比性示例能带来一致改进，而仅推理提示可能恶化跟踪。我们的结果促使将多语言、多元审计作为LLM-机器人部署前的门控，并表明模型因素比单独提示更稳健。
英文摘要：
LLM-governed social robots increasingly decide who receives real-world assistance first. As prioritization norms vary across cultures by age, status, and group size, failure to calibrate pluralistically can scale into unequal access. Yet LLM moral audits remain English-centered, rarely test embodied contexts, leaving pluralistic calibration as an urgent diagnostic gap amid intensifying LLM-robot deployment. We introduce a gradient-based audit framework for multilingual evaluation of LLM moral trade-off behavior against cultural preference gradients. Grounded in nine cross-domain social robotics reviews (>8,000 papers), we derive symmetry-controlled scenarios across care, education, and services, translating the Moral Machine Experiment's "whom to spare" into "whom to assist first" dilemmas with preserved identity trade-offs (many vs. few; young vs. old; higher vs. lower status). We audit four LLMs across four country-language pairs in four prompting regimes (57,600 decisions), benchmarked against country-specific MME preference gradients. Ordinal concordance tests whether models differentiate cultural contexts; a governance typology maps vulnerabilities in gradient differentiation, directional tendency, and deliberation. We find persistent, culturally asymmetric gradient tracking failures that prompting alone cannot reliably correct: quality calibration is nearly twice as strong for Western-language decisions as for Chinese and Japanese; high determinism in majority-first trade-offs often erases cross-cultural gradients; partial sensitivity to age- and status-based norms risks sidelining minorities. Prompting effects are uneven; only contrastive exemplars yield consistent gains, while reasoning-only prompts can worsen tracking. Our results motivate multilingual, pluralistic audits as an LLM-robot pre-deployment gate and suggest model factors are a more robust lever than prompting alone.
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。
