Skip to main content

Shiwei Liu and Visual Informatics Group Receive Best Paper Award at LoG 2022

Shiwei Liu

Texas ECE and Institute for Foundations of Machine Learning (IFMl) postdoc Shiwei Liu along with Prof. Atlas Wang and the Visual Informatics Group received the Best Paper Award at the Learning on Graphs (LoG) 2022 Conference for their paper “You Can Have Better Graph Neural Networks by Not Training Weights at All: Finding Untrained GNNs Tickets.”

LoG is a new annual research conference that covers areas broadly related to machine learning on graphs and geometry, with a special focus on review quality.

In this paper, the group carries out the first-of-its-kind exploration of discovering matching untrained GNNs (with no weight update).

IFML is a National Science Foundation (NSF) AI Institute.

Shiwei Liu joined Texas ECE in Fall 2022 as a postdoctoral fellow in the VITA group and the Institute for Foundations of Machine Learning (IFML), under the supervision of Dr. Atlas Wang. Shiwei received his Bachelor of Science at North University of China in 2015. He obtained his PhD degree at the Eindhoven University of Technology (TU/e), the Netherlands, under the supervision of Prof. Mykola Pechenizkiy and Dr. Decebal Constantin Mocanu. His research interests include (1) designing efficient (sparse) neural networks and training recipes; and (2) studying the behavior of deep learning from a scientific perspective, by conducting empirical experiments in practice.

All authors:

Tianjin Huang, PhD student at Eindhoven University of Technology (First author)
Tianlong Chen, PhD student at UT Austin
Meng Fang, Assistant Professor at University of Liverpool
Vlado Menkovski, Assistant Professor at Eindhoven University of Technology
Jiaxu Zhao, PhD student at Eindhoven University of Technology
Lu Yin, PhD student at Eindhoven University of Technology
Yulong Pei, Assistant Professor at Eindhoven University of Technology Decebal
Constantin Mocanu, Assistant Professor at University of Twente, Eindhoven University of Technology
Zhangyang Wang, Assistant Professor at UT Austin
Mykola Pechenizkiy, Full Professor at Eindhoven University of Technology
Shiwei Liu, Postdoc at at UT Austin (Lead Author)