Unverified50% confidenceFactExact time
Sigmoid函数的导数最大值仅为0.25，经过数十层连乘后梯度会指数级衰减
1
Sources
50%
Confidence
Long-term
Relevance
8/29/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedIn long-chain tasks, if each step has an error rate of approximately 5%, the overall success rate may drop to below 36% after 20 steps.63% similarVerifiedIf the accuracy per step is 95%, after 20 steps the overall accuracy will drop to approximately 36%.62% similarUnverifiedResearch by Gunasekar et al. in 2017 proved that for linear models, gradient descent is equivalent to minimum-norm interpolation.61% similarUnverifiedRidge regression shrinks regression coefficients toward zero but does not set them to zero by adding an L2 regularization term to the loss function, whereas LASSO (L1 regularization) produces sparse solutions.59% similarUnverified在LLM超低比特量化中，Transformer模型某些特定通道的激活值可能比其他通道大100倍以上，这些离群值会导致极端量化的精度崩溃58% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/817113API
curl https://kongchang.com/api/v1/knowledge/claims/817113MCP
get_claim(id=817113)