Human-Guided Harm Recovery for Computer Use Agents 事件

Human-Guided Harm Recovery for Computer Use Agents arXiv:2604.18847v2 Announce Type: replace-cross Abstract: As LM agents gain the ability to execute actions on real computer systems, we need ways to not only prevent harmful actions at scale but also effectively remediate harm when prevention fails. We formalize a solution to this neglected challenge in post-execution safeguards as harm recovery: the problem of optimally steering an agent from a harmful state back to a safe one in alignment wit