The authors of The merged-staircase property: a necessary and nearly sufficient condition for SGD learning of sparse functions on two-layer neural networks have not publicly listed the code yet.
Request code directly from the authors:
Get an expert to implement this paper:
(OR if you have code to share with the community, please add it here 😊🙏 )