Kaan Sancak — accepted AI-conference papers

2025

A Scalable and Effective Alternative to Graph Transformers

Kaan Sancak, Zhigang Hua, Jin Fang, Yan Xie, Andrey Malevich, Bo Long +2

AAAI 2025technical

Graph Neural Networks (GNNs) have shown impressive performance in graph representation learning, but they face challenges in capturing long-range dependencies due to their limited expressive power. To address this, Graph Transformers (GTs) were introduced, utilizing self-attention mechanism to effec…

Cited by 1 Source ↗PDF ↗code ↗Scholar ↗

2024

VCR-Graphormer: A Mini-batch Graph Transformer via Virtual Connections

Dongqi Fu, Zhigang Hua, Yan Xie, Jin Fang, Si Zhang, Kaan Sancak +4

ICLR 2024poster

Graph transformer has been proven as an effective graph learning method for its adoption of attention mechanism that is capable of capturing expressive representations from complex topological and feature information of graphs. Graph transformer conventionally performs dense attention (or global att…

Cited by 31 Source ↗PDF ↗code ↗Scholar ↗