@misc{indiciaecd3ce2270ab1,
  title = {Massive Activations in Graph Neural Networks: Decoding Attention for Domain-Dependent Interpretability},
  author = {Lorenzo Bini and Marco Sorbi and Stephane Marchand-Maillet},
  year = {2025},
  url = {https://arxiv.org/abs/2409.03463},
  note = {Source identifier: 2409.03463}
}