Internal Covariate Shift? How Does Batch Normalization Help Optimization? https://arxiv.org/abs/1805.11604 A B Good for a, update But not for a b 00●●●0 Batch normalization make a and a'have similar statistics. Experimental results do not support the above idea. Internal Covariate Shift? How Does Batch Normalization Help Optimization? https://arxiv.org/abs/1805.11604 11 𝒙 𝐴 𝒂 𝐵 𝒃 …… 𝒙 𝐴′ 𝒂′ 𝐵′ 𝒃′ …… update Good for 𝒂, But not for 𝒂’ Batch normalization make 𝒂 and 𝒂’ have similar statistics. Experimental results do not support the above idea