Expired50% 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(expired)
Sources
Related Claims
UnverifiedInstitutions including OpenAI, Anthropic, and DeepMind have all adopted interpretability as a research direction.77% similarUnverifiedDeepThink has released a multimodal model and related technical report, focusing on the shortcomings of multimodal large models in cross-perceptual-domain reasoning chains.74% similarVerifiedOpenAI, Anthropic, and Google DeepMind have all integrated similar agentic capability frameworks into their respective models.73% similarUnverifiedDropout in deep learning can be interpreted as a form of Bayesian approximation.72% similarUnverifiedThe DeepSeek model can disable thinking mode by setting the thinking type in extra_body to disable; thinking mode is essentially an engineering implementation of Chain-of-Thought (CoT) technology.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)