Aidan Gomez:这次异常行为暴露了思维链监控的局限
中文全文 · AI 翻译
更令人不安的一点是,模型的推理内容完全没有透露它接下来要做什么。这意味着,思维链(CoT)监控在这里会毫无用处。
对照原文
One of the more disturbing things was that the model’s reasoning didn’t indicate anything about what it was about to do. Meaning that CoT monitoring would have been useless here.