新研究：AI模型书面推理步骤对应独特的内部模式
原标题：AI models' written reasoning steps correspond to distinct internal patterns, a new study finds
AI 摘要
韩国KAIST与Naver AI Lab的一项新研究发现，语言模型文本输出中的推理步骤在模型内部数值表示中也能被区分开来，且信号在中间层最强。研究团队定义了八种推理操作，用Qwen2.5-7B、Qwen3-8B和Gemma4-31B求解数学题，并用GPT-5标注各步骤。结果表明，内部状态携带的推理步骤类型信息超越了表面用词，且推理步骤依赖前文语境而非孤立形成。该发现对AI安全中的思维链监督有潜在意义，但实验仅限于数学任务和少数模型。
正文节选
AI models' written reasoning steps correspond to distinct internal patterns, a new study finds Can the distinct reasoning steps a language model shows in its text output also be found in its internal states? A new study put it to the test. When a reasoning model solves a task step by step, it does different things along the way: reading data, breaking down the problem, retrieving a formula, running a calculation. Researchers at South Korea's KAIST and Naver AI Lab wanted to know whether those re