OpenAI News·· 2026-03-05AI 评分42
OpenAI 推出 CoT-Control：推理模型难以控制思维链，监控成为 AI 安全保障
Reasoning models struggle to control their chains of thought, and that’s good
AI 导读
OpenAI 推出 CoT-Control，发现推理模型难以控制思维链，强调监控在 AI 安全中的重要性。
原文
OpenAI introduces CoT-Control and finds reasoning models struggle to control their chains of thought, reinforcing monitorability as an AI safety safeguard.
来源：OpenAI News · openai.com