3 citations · 4 across the 2 of their papers we have counts for
3 papers
cs.LG2023★ 3 cited
Type-II Saddles and Probabilistic Stability of Stochastic Gradient Descent
Liu Ziyin, Botao Li, Tomer Galanti +1
Characterizing and understanding the dynamics of stochastic gradient descent (SGD) around saddle points remains an open problem. We first show that saddle points in neural networks…
stat.ML2022★ 1 cited
Exact Solutions of a Deep Linear Network
Liu Ziyin, Botao Li, Xiangming Meng
This work finds the analytical expression of the global minima of a deep linear network with weight decay and stochastic neurons, a fundamental model for understanding the landscap…
cs.LG2021
SGD with a Constant Large Learning Rate Can Converge to Local Maxima
Liu Ziyin, Botao Li, James B. Simon +1
Previous works on stochastic gradient descent (SGD) often focus on its success. In this work, we construct worst-case optimization problems illustrating that, when not in the regim…