GNRK: Graph Neural Runge-Kutta method for solving partial differential equations

1 code implementation • 1 Oct 2023 • Hoyun Choi, Sungyeop Lee, B. Kahng, Junghyo Jo

Neural networks have proven to be efficient surrogate models for tackling partial differential equations (PDEs).

Scale-invariant representation of machine learning

1 code implementation • 7 Sep 2021 • Sungyeop Lee, Junghyo Jo

We observe that the frequency of internal codes or labels follows power laws in both supervised and unsupervised learning models.

BIG-bench Machine Learning Clustering

Information flows of diverse autoencoders

1 code implementation • 15 Feb 2021 • Sungyeop Lee, Junghyo Jo

Thus, we conclude that the compression phase is not necessary for generalization in representation learning.

Information Plane Representation Learning