Fetching the paper…

Error bounds for approximations with deep ReLU neural networks in $W^{s,p}$ norms · Around