OpenAI 研究：模型越强，CoT 监控越难
OpenAI 的这项研究发现模型能力越强，思维链越难当监控窗口用，等于给靠 CoT 做安全审计的团队敲了警钟。做对齐、评估或想理解推理模型黑箱机制的人应该看看这条推文串，里面把问题拆成了可量化的两个维度。
OpenAI 研究员 Jeffrey Wang 发布新研究，考察思维链（CoT）可监控性在模型训练生命周期中的演变，包括 no-CoT 能力与 CoT 可控性两个维度。
研究发现：随着模型能力随训练规模提升，通过 CoT 监控模型行为正变得越来越困难。研究将可监控性拆解为能力与可控性两条线，为定位这种失控发生在训练哪个阶段、以何种机制出现提供了实验框架。
作者承认许多问题尚未有答案，但认为这是研究 CoT 可监控性衰减的一个可行起点，对依赖 CoT 监控做 AI 安全对齐的团队是直接的警示信号。
Recommended read. They find that increasingly intelligent models are harder to monitor through their CoT. Lots of things remain unanswered, but this is a nice setup for studying how and where this emerges. Jeffrey Wang @jeffreygwang (1/10) Today, we’re sharing new research from @OpenAI : How does CoT monitorability (no-CoT capability and CoT controllability) evolve over the model training lifecycle? We emerged with unsurprising but sobering findings: as we scale toward more intelligent systems, models are becoming harder to monitor through their CoT. 🧵 🔗 View Quoted Tweet📊 9 ⚡
This report is categorized under AITOP's "Papers" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to US-China foundation model benchmarks and commercialization comparison and AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include OpenAI, 思维链, 推理模型, AI安全.
Source attribution: curated from elvis (@omarsar0).
