@misc{indiciae84618eb90e57,
  title = {Stealing Training Data from Large Language Models in Decentralized Training through Activation Inversion Attack},
  author = {Chenxi Dai and Lin Lu and Pan Zhou},
  year = {2025},
  url = {https://arxiv.org/abs/2502.16086},
  note = {Source identifier: 2502.16086}
}