Reasoning models struggle to control their chains of thought, and that’s good

【RSS 原文 · 未经 AI 筛选或翻译】OpenAI introduces CoT-Control and finds reasoning models struggle to control their chains of thought, reinforcing monitorability as an AI safety safeguard.