How Explanations Leak the Decision Logic: Stealing Graph Neural Networks via Explanation Alignment

Published in The Thirty-ninth Annual Conference on Neural Information Processing Systems (**NeurIPS 2025**), 2025