@misc{indiciae4c1688963279,
  title = {Equivariant Differentially Private Deep Learning: Why DP-SGD Needs Sparser Models},
  author = {Florian A. Hölzl and Daniel Rueckert and Georgios Kaissis},
  year = {2023},
  url = {https://arxiv.org/abs/2301.13104},
  note = {Source identifier: 2301.13104}
}