Unverified50% confidenceBenchmarkExact time
残差连接支撑了152层超深网络的训练，并在2015年ImageNet竞赛中以低于4%的错误率首次超越人类水平（约5%）
1
Sources
50%
Confidence
Long-term
Relevance
7/14/2026
First Seen
Sources
AI大模型入门：核心概念、发展历程与国产模型格局
Related Claims
Unverified2012年Hinton团队的AlexNet以top-5错误率从26%降至15%的成绩，ResNet（2015，微软研究院）通过残差连接解决深层网络梯度消失问题，EfficientNet（2019，Google Brain）通过复合缩放系数优化网络70% similarUnverifiedIn Google Brain's 2020 ViT paper, it was noted that ViT trained from scratch on ImageNet-1K (1.28 million images) underperforms CNNs of equivalent scale, but begins to surpass CNNs when pre-training data is scaled to ImageNet-21K (14 million images) or JFT-300M (300 million images)68% similarUnverifiedThe DeepSuite benchmark has a false positive rate of 44.9% and a false negative rate of 1.2%.62% similarUnverifiedIn 2012, Hinton's team won the ImageNet competition using GPU + ReLU + Dropout, reducing the image classification error rate from 26% to 15%.61% similarUnverified移除歧义后所有模型准确率提升26.8至40.2个百分点，Gemini甚至达到81%，表明歧义是深度搜索代理最大性能瓶颈之一61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/507994API
curl https://kongchang.com/api/v1/knowledge/claims/507994MCP
get_claim(id=507994)