AGENTSAFE：面向危险指令的具身智能体安全基准
AGENTSAFE: Benchmarking the Safety of Embodied Agents on Hazardous Instructions
对做 VLM 具身智能体的视觉研究者，SAFE 提供了可复用的对抗仿真沙盒与通用适配器（把高层 VLM 输出映射到低层具身控制），以及按感知-规划-执行分阶段诊断的评测协议，可用于定位「识别到危险却无法转化为安全规划」这一具体失效环节。
原文依据 we uncover systematic failures in translating hazard recognition into safe planning and execution
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论视觉语言模型（VLM）的集成正在推动新一代具身智能体在以人为本的环境中运行。然而，随着部署范围扩大，这些系统面临日益增长的安全风险，尤其是在执行危险指令时。当前的安全评测基准仍然有限：它们只覆盖狭窄的危险范围，且主要关注最终结果，忽视了智能体完整的感知—规划—执行过程，从而掩盖了关键的失效模式。因此，我们提出 SAFE，一个用于系统评估具身 VLM 智能体在危险指令下安全性的基准。SAFE 包含三个组成部分：SAFE-THOR，一个可扩展的对抗仿真沙盒，配有通用适配器，可将高层 VLM 输出映射到低层具身控制，支持多种智能体工作流的集成；SAFE-VERSE，一个受阿西莫夫机器人三定律启发、具备风险意识的任务套件，包含 45 个对抗场景、1,350 个危险任务和 9,900 条指令，覆盖对人类、环境和智能体自身的风险；以及 SAFE-DIAGNOSE，一个多层次、细粒度的评测协议，从感知、规划和执行三个层面衡量智能体表现。将 SAFE 应用于九个最优 VLM 和两种具身智能体工作流后，我们发现了在将危险识别转化为安全规划与执行方面的系统性失败。我们的发现揭示了当前安全对齐的根本局限，并证明了开展全面、多阶段评测以发展更安全的具身智能的必要性。
英文原文摘要 1599 字符
The integration of vision-language models (VLMs) is driving a new generation of embodied agents capable of operating in human-centered environments. However, as deployment expands, these systems face growing safety risks, particularly when executing hazardous instructions. Current safety evaluation benchmarks remain limited: they cover only narrow scopes of hazards and focus primarily on final outcomes, neglecting the agent's full perception-planning-execution process and thereby obscuring critical failure modes. Therefore, we present SAFE, a benchmark for systematically assessing the safety of embodied VLM agents on hazardous instructions. SAFE comprises three components: SAFE-THOR, an extensible adversarial simulation sandbox with a universal adapter that maps high-level VLM outputs to low-level embodied controls, supporting diverse agent workflow integration; SAFE-VERSE, a risk-aware task suite inspired by Asimov's Three Laws of Robotics, comprising 45 adversarial scenarios, 1,350 hazardous tasks, and 9,900 instructions that span risks to humans, environments, and agents; and SAFE-DIAGNOSE, a multi-level and fine-grained evaluation protocol measuring agent performance across perception, planning, and execution. Applying SAFE to nine state-of-the-art VLMs and two embodied agent workflows, we uncover systematic failures in translating hazard recognition into safe planning and execution. Our findings reveal fundamental limitations in current safety alignment and demonstrate the necessity of a comprehensive, multi-stage evaluation for developing safer embodied intelligence.