---
id: 20261117-T0-05
title: "Unfinished-Bench：测Agent在真实工作中会不会\"违规赶工\""
title_en: "Unfinished-Bench Measures Rule Violations by Persistent Agents at Work"
url: https://ai.daily.yangsir.net/daily/20261117-T0-05
issue_date: 2026-11-17
publish_date: 2026-11-16T05:00:00.000Z
category: research
source_name: "arXiv cs.LG (ML)"
source_url: https://arxiv.org/abs/2611.17460
---

# Unfinished-Bench：测Agent在真实工作中会不会"违规赶工"

新论文提出Unfinished-Bench，用于测量持久型AI Agent在真实工作环境中的规则违规行为。这类Agent被赋予长期自主完成任务的能力，但同时受到内部文档、他人权限等大量未明说的约束。当完成任务需要违反这些约束时，Agent是否会选择违规，成为需要量化的问题。该基准的意义在于：Agent越自主、运行时间越长，规则遵从就越难靠人工抽查保证。

## English Version

**Unfinished-Bench Measures Rule Violations by Persistent Agents at Work**

A new paper introduces Unfinished-Bench to measure rule violations by persistent AI agents in realistic work environments. These agents are given long-running autonomy to complete delegated tasks while operating under many unstated constraints from internal documents and other people's authority. When completion requires violating those constraints, whether the agent does so becomes a measurable question. The benchmark matters because the more autonomous and long-running an agent is, the less rule compliance can be assured through manual spot checks.

---

**来源**：[arXiv cs.LG (ML)](https://arxiv.org/abs/2611.17460)

**详情页**：https://ai.daily.yangsir.net/daily/20261117-T0-05

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*