Sebastian Raschka 谈推理模型 CoT 可监控性：预训练越强越难监控
Raschka 把一篇关于 CoT 可监控性的研究讲得很清楚：预训练越强、RL 用得越多，模型思维链越难看懂。做 AI 安全评估或部署推理模型的人应该看看，能帮你判断什么时候不能只信模型的书面推理。
Sebastian Raschka 在 X 上讨论了推理模型开发中优化思维链（CoT）与保持可监控性之间的矛盾：只要模型把思考过程写出来，人们就能检查它是否作弊，但显式优化 CoT 可能失去这一能力。
他引用一篇研究指出两个关键变量：一是 no-CoT 能力，即模型在单次前向传播中不写字就能完成多少推理，更好的预训练（更低 perplexity）会提升这一能力，而架构基本无影响；二是 CoT 可控性，即模型被要求增删改写推理时能否照做，仅用 RL 就能大幅提升可控性，且可控性会通过蒸馏传递给其他模型。这两项能力越高，监控就越难。
研究还确认了已知直觉：模型越强，CoT 写得越少（如同更聪明的学生考试时不需要草稿），而字面思考未必反映模型真实做法。一个附带的好消息是，过程监督（process supervision）似乎不是导致难监控的主要因素。
In reasoning model development, it's always been this interesting tension between training/optimizing chain-of-thoughts (CoTs) vs not touching them to preserve monitorability.
CoTs are cool because we can check what the model is up to (e.g., whether it cheats or not) as long as it spells out its thought process. And we risk losing that if we explicitly optimize CoTs.
Two caveats:
1) I.e., the catch of CoT and monitorability is that this only works if the written thinking reflects what the model is actually doing.
2) Also, when models become more capable, they spell out less in the CoT (like a cleverer math student needs less note-taking when solving problems during an exam).
Nothing new on that front, but an interesting thread below on how to measure "harder to monitor", with some interesting results that confirm the existing intuitions:
1) no-CoT capability (meaning how much reasoning the model can do in a single forward pass without writing anything down)
- Better pretraining (lower perplexity) raises this no-CoT capability - Architecture mostly doesn't matter 2) CoT controllability (meaning how well the model can change its written reasoning when asked to add, remove, or change parts of it)
- RL alone raises controllability a lot - Controllability is passed on through distillation
* Higher no-CoT capability and higher controllability both make monitoring harder.
PS: Perhaps the good news is that it seems that process supervision wasn't the main culprit.
本条情报归属 AITOP「论文」核心栏目，重点聚焦北美前沿实验室与硅谷科技巨头。在跨区域产业观察坐标中，该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线，反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括：CoT、RL、推理模型、AI安全，读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明：本文原始线索采编自「Sebastian Raschka」（@rasbt），经 AITOP 多维评分管线结构化提炼。
