Unverified50% confidenceFactExact time
AlexNet was the first to systematically use ReLU instead of Sigmoid/Tanh in a large-scale deep convolutional network, making training deep networks feasible and increasing training speed several times over.
1
Sources
50%
Confidence
Long-term
Relevance
10/13/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAlexNet adopted an 8-layer deep convolutional neural network architecture, leveraged GPU-accelerated training, and reduced the Top-5 error rate from the previous year's 25.8% to 15.3%.74% similarUnverifiedResNet（2015，He 等人）通过引入跳跃连接构造恒等映射的学习旁路，将可训练网络深度推进至 152 层乃至千层，并在 ImageNet 上以显著优势夺冠73% similarVerifiedResNet (Residual Network) was proposed by Microsoft Research's He Kaiming team in 2015 and solved the vanishing gradient problem in deep network training by introducing residual connections.72% similarVerifiedAlexNet在ImageNet竞赛中夺冠正式开启了深度学习时代72% similarUnverified残差连接与层归一化解决了深层网络训练不稳定的问题，使模型可堆叠到数十层69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1013757API
curl https://kongchang.com/api/v1/knowledge/claims/1013757MCP
get_claim(id=1013757)