Unverified50% confidenceEventExact time
研究者向OpenAI和Anthropic的模型提供名为deep_think的工具时，模型通常隐藏的思维链出现了意外泄露
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
8/13/2026
First Seen
Valid until: 8/27/2026
Sources
Related Claims
UnverifiedOpenAI、Anthropic、DeepMind等机构均已将可解释性作为研究方向77% similarUnverifiedDeepThink发布了多模态模型及相关技术报告，重点讨论了多模态大模型在跨感知域推理链路上的不足74% similarVerifiedOpenAI、Anthropic和Google DeepMind均已在各自模型中集成了类似的智能体能力框架73% similarUnverified深度学习中Dropout可以被解释为一种贝叶斯近似72% similarUnverifiedDeepSeek模型可通过设置extra_body中的thinking类型为disable来关闭思考模式，思考模式本质是Chain-of-Thought（CoT）技术的工程实现72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/741014API
curl https://kongchang.com/api/v1/knowledge/claims/741014MCP
get_claim(id=741014)