This repository provides a reference implementation of Graph Convolutional Networks (GCN) as described in the 2017 ICLR paper by Kipf and Welling. It implements graph convolution operations using TensorFlow to learn representations of nodes in graph-structured data. The implementation includes both standard GCN and a Chebyshev polynomial version, supporting citation network datasets like Cora, Citeseer, and Pubmed for benchmarking.