Unverified50% confidenceFactExact time
Sigmoid将输出压缩到0~1之间，Tanh是早期常用激活函数，但随网络层数加深会导致梯度消失问题
1
Sources
50%
Confidence
Long-term
Relevance
7/23/2026
First Seen
Sources
Related Claims
VerifiedThe vanishing gradient problem originates from saturating activation functions such as Sigmoid or Tanh, whose derivatives approach zero for large or small input values, causing gradients to decay layer by layer.80% similarVerifiedThe Sigmoid function has a vanishing gradient problem: when the absolute value of the input is large, its derivative approaches 0, causing the gradient to decay layer by layer during backpropagation.76% similarUnverifiedsigmoid激活函数的导数最大值仅为0.25，多层连乘后导致梯度迅速消失70% similarUnverified在高维空间中通过令导数为零求解析解几乎不可能，梯度下降是可扩展到高维的策略57% similarUnverifiedVanishing gradient refers to the exponential decay of gradient values during backpropagation in deep neural networks, causing parameters near the input layer to receive almost no updates.56% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/598511API
curl https://kongchang.com/api/v1/knowledge/claims/598511MCP
get_claim(id=598511)