@misc{indiciae9d265fa3a0f5,
  title = {Gradient-based grand canonical optimization enabled by graph neural networks with fractional atomic existence},
  author = {Mads-Peter Verner Christiansen and Bjørk Hammer},
  year = {2025},
  doi = {10.1088/2632-2153/ae1bf6},
  url = {https://arxiv.org/abs/2507.19438},
  note = {Source identifier: 2507.19438}
}