CourtGuard：一种用于大语言模型安全的模型无关框架，用于零样本策略适应
CourtGuard: A Model-Agnostic Framework for Zero-Shot Policy Adaptation in LLM Safety
AI总结：
CourtGuard通过证据辩论框架实现大语言模型的安全零样本适应，超越传统方法在多个安全基准测试中的表现，并具备跨领域泛化和自动数据审计能力。
摘要：
当前用于大型语言模型（LLM）的安全机制严重依赖静态、微调的分类器，这些分类器存在适应刚性的问题，无法在不进行昂贵再训练的情况下强制执行新的治理规则。为了解决这一问题，我们引入了CourtGuard，一种检索增强的多代理框架，将安全评估重新想象为证据辩论。通过组织基于外部政策文件的对抗性辩论，CourtGuard在7个安全基准测试中实现了最先进的性能，优于专门的政策遵循基线，而无需微调。除了标准指标外，我们还突显了两个关键能力：（1）零样本适应性，其中我们的框架通过交换参考政策成功泛化到一个非领域维基百科篡改任务（达到90%的准确率）；（2）自动数据整理和审计，我们利用CourtGuard来整理和审计九个新的复杂对抗性攻击数据集。我们的结果表明，将安全逻辑与模型权重解耦，为满足当前和未来的AI治理监管要求提供了一条稳健、可解释且可适应的路径。
英文摘要：
Current safety mechanisms for Large Language Models (LLMs) rely heavily on static, fine-tuned classifiers that suffer from adaptation rigidity, the inability to enforce new governance rules without expensive retraining. To address this, we introduce CourtGuard, a retrieval-augmented multi-agent framework that reimagines safety evaluation as Evidentiary Debate. By orchestrating an adversarial debate grounded in external policy documents, CourtGuard achieves state-of-the-art performance across 7 safety benchmarks, outperforming dedicated policy-following baselines without fine-tuning. Beyond standard metrics, we highlight two critical capabilities: (1) Zero-Shot Adaptability, where our framework successfully generalized to an out-of-domain Wikipedia Vandalism task (achieving 90\% accuracy) by swapping the reference policy; and (2) Automated Data Curation and Auditing, where we leveraged CourtGuard to curate and audit nine novel datasets of sophisticated adversarial attacks. Our results demonstrate that decoupling safety logic from model weights offers a robust, interpretable, and adaptable path for meeting current and future regulatory requirements in AI governance.
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。
