让机器人持续学习安全规则，边做边改还防攻击。
NEXUS: Continual Learning of Symbolic Constraints for Safe and Robust Embodied Planning
- 用符号约束动态更新安全规则，实现闭环进化
- 在安全基准上成功率更高，还能拒绝危险指令
- 适合需要长期安全运行的智能体系统
尽管大语言模型推动了具身智能的发展，但其固有的概率不确定性与物理世界所需的严格确定性和可验证安全性之间存在根本差距。为弥合这一鸿沟，本文提出NEXUS框架，支持具身智能体的持续学习。不同于以往将符号元素视为静态接口的做法，NEXUS利用它们实现符号化语义锚定与知识演进。该框架显式分离物理可行性与安全规范：通过闭环执行反馈提升智能体能力，同时将概率风险评估转化为确定性硬约束，建立预动作防御机制。在SafeAgentBench上的实验表明，NEXUS不仅任务成功率显著提升，还能有效拒绝不安全指令，对对抗攻击具备强鲁棒性，并通过知识积累逐步提高规划效率。
原文摘要 · Abstract (English)
While Large Language Models (LLMs) have catalyzed progress in embodied intelligence, a fundamental gap between their inherent probabilistic uncertainty and the strict determinism and verifiable safety required in the physical world. To mitigate this gap, this paper introduces NEXUS, a modular framework designed for continual learning in embodied agents. Different from prior works that treat symbolic artifacts merely as static interfaces, NEXUS leverages them for symbolic grounding and knowledge evolution. The framework explicitly decouples physical feasibility from safety specifications: capability of agents is improved through closed-loop execution feedback, while probabilistic risk assessments are grounded into deterministic hard constraints to establish a rigorous pre-action defense. Experiments on SafeAgentBench demonstrate that NEXUS achieves superior task success rates while effectively refusing unsafe instructions, exhibiting robust defense against adversarial attacks, and progressively improving planning efficiency through knowledge accumulation.
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。