Why do Larger Models Generalize Better? A Theoretical Perspective via the XOR Problem

By Alon Brutzkus, Amir Globerson. 2018

79 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Jianping Gou
2. Dacheng Tao
3. Baosheng Yu
4. S. Maybank
5. Xingjiao Wu
6. Tianlong Ma
7. Luwei Xiao
8. Liangbo He

Citations in the corpus

• Knowledge Distillation: A Survey (2020, background): “Large deep models tend to achieve good performance in practice, because the over parameterization improves the generalization performance when new data is considered (Brutzkus and Globerson, 2019; Allen-Zhu et al., 2019; Arora et al., 2018; Zhang et al., 2018; Tu et al., 2020).”
• A Survey of Human-in-the-loop for Machine Learning (2021, background): “The great success of deep learning is due to larger-scale models, which include hundreds of millions of parameters [2].”