@exc2075

Training two-layer ReLU networks with gradient descent is inconsistent

, and . Journal of Machine Learning Research, 23 (181): 1--82 (2022)

Links and resources

Tags

community