1 paper
Mo Zhou, Weihang Xu, Simon S. Du +1
Recent work has identified incremental learning in shallow networks trained on single-index and multi-index models. However, existing analyses often rely on simplifying settings, s…