为照护者AI对话设计风险评估框架，有效降低98%的回应风险。
RubRIX: Rubric-Driven Risk Mitigation in Caregiver-AI Interactions
- 基于照护伦理构建五维风险评分体系，量化回应中的忽视、偏见等隐患。
- 在2万条真实照护咨询上测试，单轮优化使风险成分下降45%-98%。
- 适合关注医疗AI安全、人机交互伦理的研究者与开发者使用。
寻求AI支持的照护者提出复杂需求——信息获取、情感认同和压力信号——需要对回应的安全性与恰当性进行细致评估。现有AI评估框架多聚焦通用风险（如毒性、幻觉、政策违规等），难以捕捉大语言模型在照护场景中的细微风险。本文提出RubRIX（基于评价标准的风险指数），一个基于理论、经临床专家验证的框架，用于评估大模型在照护对话中的风险。该框架以照护伦理要素为基础，将五种实证衍生的风险维度具体化：忽视、偏见与污名、信息不准、无批判肯定、认知傲慢。我们在Reddit和ALZConnected超过2万条照护者提问上评估了六款先进大模型。通过评价标准引导的优化，各模型在一轮迭代后风险成分平均降低45%-98%。本研究贡献了一种面向高负担场景的领域敏感、用户中心的评估方法，强调在照护支持中开展情境化互动风险评估的重要性。我们公开了基准数据集，以推动未来对AI支持中情境风险评估的研究。
原文摘要 · Abstract (English)
Caregivers seeking AI-mediated support express complex needs -- information-seeking, emotional validation, and distress cues -- that warrant careful evaluation of response safety and appropriateness. Existing AI evaluation frameworks, primarily focused on general risks (toxicity, hallucinations, policy violations, etc), may not adequately capture the nuanced risks of LLM-responses in caregiving-contexts. We introduce RubRIX (Rubric-based Risk Index), a theory-driven, clinician-validated framework for evaluating risks in LLM caregiving responses. Grounded in the Elements of an Ethic of Care, RubRIX operationalizes five empirically-derived risk dimensions: Inattention, Bias & Stigma, Information Inaccuracy, Uncritical Affirmation, and Epistemic Arrogance. We evaluate six state-of-the-art LLMs on over 20,000 caregiver queries from Reddit and ALZConnected. Rubric-guided refinement consistently reduced risk-components by 45-98% after one iteration across models. This work contributes a methodological approach for developing domain-sensitive, user-centered evaluation frameworks for high-burden contexts. Our findings highlight the importance of domain-sensitive, interactional risk evaluation for the responsible deployment of LLMs in caregiving support contexts. We release benchmark datasets to enable future research on contextual risk evaluation in AI-mediated support.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。