Unverified50% confidenceFactExact time
拜登时代的AI行政令曾以10²⁶次浮点运算（FLOP）作为训练算力监管红线
1
Sources
50%
Confidence
Long-term
Relevance
7/12/2026
First Seen
Sources
Related Claims
UnverifiedAnthropic采用宪法AI（Constitutional AI）训练方法训练Opus系列模型64% similarUnverifiedCurrent mainstream AI alignment methods include RLHF, Constitutional AI, and DPO, which internalize safety constraints into model weights during the model training phase.63% similarVerifiedConstitutional AI is a training method proposed by Anthropic in 2022, whose core idea is to provide an AI system with a set of explicit written principles and introduce an AI feedback (RLAIF) mechanism63% similarUnverifiedClaude Code的逐步审批设计与Anthropic在AI安全研究中强调的可控性原则一致，是Constitutional AI训练方法的延伸体现63% similarUnverifiedThe Biden administration's AI executive order required that training of models exceeding a specific compute threshold be reported to the government.62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/491859API
curl https://kongchang.com/api/v1/knowledge/claims/491859MCP
get_claim(id=491859)