Reproducing: Ma, Bin, et al. “How Explanations Leak the Decision Logic: Stealing Graph Neural Networks via Explanation Alignment.” ArXiv (Cornell University), 3 June 2025, https://doi.org/10.48550/arxiv.2506.03087.
This repository contains my reproduction work for the Model Extraction Attack and Defense project in which I am working on through Florida State University's Undergraduate Research Program (UROP). It includes code, Colab notebooks, and results from weekly experiments (Week 3 — Week 5). The goal is to reproduce the experiments and compare results with the paper.