arXiv cs.AI· Lewen Yan, Yanxu Zhu, Jie Zhang, Jing Shao·· 13 小时前AI 评分38
OwlSight：面向思维链推理的全景监控器
OwlSight: A Panoramic Monitor for Chain-of-Thought Reasoning
AI 导读
研究者提出 CoT 监控策略 Factor，通过在提示词中自定义监控要求，让单一监控器依靠不同 factor 覆盖多种任务，并区分细粒度与粗粒度两类 factor。
正文 · 原文
Abstract:Chain-of-Thought (CoT) monitoring offers actionable evidence for oversight AI systems, with the potential to enable early detection and intervention. However, existing approaches are focused on specific tasks, whereas the regulation of AI, particularly reasoning models, requires monitoring a broader range of tasks. We introduce Factor for CoT monitoring, a strategy that supports customizing monitoring requirements in the prompt, which enables a single monitor to cover different tasks by relying on different factors. Furthermore, we introduce two types of factors, fine-grained and coarse-grained (e.g., monitoring whether a specific email address has been leaked versus monitoring privacy leakage), to accommodate diverse user monitoring needs. Based on this, we develop a CoT monitor named OwlSight that relies solely on factor to monitor behaviors related to multiple dimensions, including capability, safety, and trustworthiness. Experiments across 15,971 reasoning trajectories show that OwlSight surpasses general-purpose LLMs by 1.32-30.57 and 11.71-27.81 points at fine and coarse granularity, respectively, and achieves state-of-the-art results on OpenAI's CoT monitorability benchmark. It also generalizes to unseen models' CoTs and remains robust to Factor rephrasings, CoT compression, and obfuscation, offering a promising paradigm for CoT monitoring that can contribute to the oversight of frontier AI.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.23670 [cs.AI] |
| (or arXiv:2610.23670v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.23670 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Lewen Yan [view email]
[v1]
Fri, 16 Oct 2026 16:01:49 UTC (3,008 KB)
来源：arXiv cs.AI · arxiv.org