Xuanyuan, H.; Barbiero, P.; Georgiev, D.; Magister, L. C.; Liò, P. Global Concept-Based Interpretability for Graph Neural Networks via Neuron Analysis. AAAI 2023, 37, 10675-10683.