PASTABench: Proactive Assessment of Sequential Trajectories for Agent Safety

2026-09-23

AI Summary

为解决自主代理操作安全问题，提出PASTABench基准及Optimal Intervention Window方法，评估显示现有模型在及时干预方面表现不佳。

Abstract

As Large Language Models (LLMs) evolve into autonomous agents that alter real-world states, ensuring operational safety across multi-step workflows has become a critical challenge. While recent work has moved beyond single-turn evaluation toward multi-turn paradigms, key limitations persist: step-level methods treat actions in isolation, missing how risks accumulate, while trajectory-level evaluations operate post-hoc, offering no opportunity for timely intervention. To address these limitations, we formalize Decoupled Proactive Safety Monitoring along three dimensions: whether to intervene, when to intervene, and what the risk is. We introduce PASTABench, a benchmark of 1,139 multi-turn trajectories spanning 5 risk categories and 13 subcategories. We further propose the Optimal Intervention Window (OIW), anchored by annotated Earliest-Signal and Trigger turns, to quantify intervention timeliness. Evaluation of 16 LLMs reveals that proactive intervention remains largely unsolved, with the best model achieving only 40.74% optimal-timing interventions. Fine-grained diagnosis further uncovers pervasive lexical overfitting: competitive safety scores of smaller models mask keyword hypersensitivity rather than genuine risk comprehension, as their proactive capability largely collapses once hazard vocabulary is neutralized.

Problem

• Large Language Models
• Autonomous Agents
• Operational Safety
• Multi-step Workflows
• Proactive Intervention

Innovation

• Decoupled Proactive Safety Monitoring
• PASTABench
• Optimal Intervention Window
• lexical overfitting