Unverified50% confidenceFactExact time
Dictionary learning refers to decomposing high-dimensional internal activation vectors of a model into a linear combination of a set of sparse, semantically interpretable atomic features.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
Unverified集成学习的理论基础来自偏差-方差权衡框架，单一复杂模型存在高方差而简单模型存在高偏差64% similarUnverified机制可解释性入门通常需要同时具备线性代数、深度学习基础以及实验编程能力63% similarUnverified分子天然具有SE(3)/E(3)等变性，即物理性质不随整体旋转和平移改变；几何深度学习将物理对称先验内嵌于网络结构，使模型参数量更小、泛化更强62% similarVerifiedAnthropic uses Sparse Autoencoders (SAE) to decompose the polysemantic activations inside the model into purer features62% similarVerifiedAnthropic's sparse autoencoder research has been able to isolate millions of interpretable features from model activations62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1003172API
curl https://kongchang.com/api/v1/knowledge/claims/1003172MCP
get_claim(id=1003172)