Unverified50% confidenceFactExact time
Between 2023 and 2024, Anthropic identified millions of interpretable features in the Claude model, including the Golden Gate Bridge feature experiment
1
Sources
50%
Confidence
Long-term
Relevance
9/3/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedAnthropic published mechanistic interpretability research on the internal feature interpretation of Claude in 2024, discovering millions of interpretable feature units.79% similarUnverifiedIn 2024, Anthropic successfully decomposed millions of interpretable features in the Claude model using sparse autoencoders (SAE).77% similarUnverifiedAnthropic在2023年发表的研究表明，通过SAE可以从Claude模型中提取出对应'金门大桥'、'代码错误'等具体概念的单元特征74% similarUnverifiedAnthropic introduced tool_use capability for the Claude model in 2024.71% similarUnverifiedAnthropic于2023年发表论文《Towards Monosemanticity》，首次在百万参数级别模型中识别出数万个单语义特征69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/850080API
curl https://kongchang.com/api/v1/knowledge/claims/850080MCP
get_claim(id=850080)