Unverified50% confidenceFactExact time
In 2023, researchers used Sparse Autoencoders to extract millions of interpretable features from the Claude model.
1
Sources
50%
Confidence
Long-term
Relevance
9/6/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAnthropic's mechanistic interpretability research paper "Towards Monosemanticity" identified millions of interpretable features in the Claude model using sparse autoencoders71% similarUnverifiedUsing SAEs, researchers have identified millions of interpretable features in models such as Claude, including representations associated with abstract concepts such as deception or power-seeking.71% similarUnverified2024年安全研究人员展示了通过对抗性后缀（Adversarial Suffix）——一串看似无意义的字符序列——来操纵模型输出的技术65% similarUnverifiedThe Claude model is known for its long context window exceeding 100,000 tokens and its complex reasoning capabilities.63% similarUnverifiedAI在解决困扰学界60年的埃尔多斯1196号本原集猜想时，引入了原本用于素数研究的冯·曼戈尔特函数58% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/865811API
curl https://kongchang.com/api/v1/knowledge/claims/865811MCP
get_claim(id=865811)