- 谁
- OpenAI
- 引文
In a new analysis from OpenAI's safety team , researchers systematically studied which specific training steps trigger this erosion.
发生了什么
OpenAI 安全团队分析了从 GPT-5 到 GPT-6 Astra 的模型，并找出与思维链透明度降低相关的四个因素：基础模型能力、架构、强化学习，以及从教师模型迁移。分析发现，有些架构在不显露的多步推理方面提升更大，而在 Astra 类模型中，增加强化学习会提高可控性。分析还指出一种权衡：更长、更加可见的推理过程更容易监控，但会增加部署时的计算成本。
推荐理由
该分析将思维链透明度降低追溯到具体的训练选择，并说明能力与计算成本之间的权衡，为读者理解模型设计如何影响监督提供了一个框架。
接下来读什么
查看这件事的完整进展OpenAI Researchers Link Training Choices to Reduced Reasoning Transparency这件事到目前为止
OpenAI’s safety team analyzed models from GPT-5 through GPT-6 Astra and linked four training factors to reduced chain-of-thought transparency: base-model capability, architecture, reinforcement learning, and transfer from teacher models. The analysis found that some architectures gain more from silent multi-step reasoning, while additional reinforcement learning improves controllability in Astra-class models. OpenAI’s researchers described a tradeoff: longer, more visible reasoning can be easier to monitor, but it increases compute costs in deployment.