% Key papers for DP-SGD refinements (scime atlas, https://atlas.scime.ai). Authors are not included: add them from the paper pages.

@inproceedings{neurips2018_cpsgd_57788,
  title = {{cpSGD: Communication-efficient and differentially-private distributed SGD}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2018},
  url = {https://nips.cc/virtual/2018/poster/11727},
}

@inproceedings{neurips2020_understanding_61038,
  title = {{Understanding Gradient Clipping in Private SGD: A Geometric Perspective}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2020},
  url = {https://nips.cc/virtual/2020/poster/18162},
}

@inproceedings{aaai2023_differentially_3466,
  title = {{Differentially Private Learning with Per-Sample Adaptive Clipping}},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2023},
  url = {https://ojs.aaai.org/index.php/AAAI/article/view/26242},
}

@inproceedings{neurips2023_automatic_68880,
  title = {{Automatic Clipping: Differentially Private Deep Learning Made Easier and Stronger}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2023},
  url = {https://nips.cc/virtual/2023/poster/70978},
}

@inproceedings{icml2021_practical_39798,
  title = {{Practical and Private (Deep) Learning Without Sampling or Shuffling}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2021},
  url = {https://icml.cc/virtual/2021/poster/9697},
}

@inproceedings{neurips2021_enabling_62913,
  title = {{Enabling Fast Differentially Private SGD via Just-in-Time Compilation and Vectorization}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2021},
  url = {https://nips.cc/virtual/2021/poster/28839},
}

@inproceedings{iclr2021_bypassing_30035,
  title = {{Bypassing the Ambient Dimension: Private SGD with Gradient Subspace Identification}},
  booktitle = {International Conference on Learning Representations (ICLR)},
  year = {2021},
  url = {https://iclr.cc/virtual/2021/poster/3266},
}

@inproceedings{aaai2022_the_2836,
  title = {{The Role of Adaptive Optimizers for Honest Private Hyperparameter Selection}},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2022},
  url = {https://aaai.org/papers/07806-the-role-of-adaptive-optimizers-for-honest-private-hyperparameter-selection/},
}

@inproceedings{icml2021_large_39506,
  title = {{Large Scale Private Learning via Low-rank Reparametrization}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2021},
  url = {https://icml.cc/virtual/2021/poster/8901},
}

@inproceedings{aaai2024_dp_5572,
  title = {{DP-AdamBC: Your DP-Adam Is Actually DP-SGD (Unless You Apply Bias Correction)}},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2024},
  url = {https://ojs.aaai.org/index.php/AAAI/article/view/29451},
}

@inproceedings{neurips2021_fast_62421,
  title = {{Fast and Memory Efficient Differentially Private-SGD via JL Projections}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2021},
  url = {https://nips.cc/virtual/2021/poster/28171},
}

@inproceedings{neurips2022_scalable_64973,
  title = {{Scalable and Efficient Training of Large Convolutional Neural Networks with Differential Privacy}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2022},
  url = {https://nips.cc/virtual/2022/poster/54548},
}

@inproceedings{icml2022_differentially_40450,
  title = {{Differentially Private Coordinate Descent for Composite Empirical Risk Minimization}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2022},
  url = {https://icml.cc/virtual/2022/poster/18089},
}

@inproceedings{iclr2022_hyperparameter_30780,
  title = {{Hyperparameter Tuning with Renyi Differential Privacy}},
  booktitle = {International Conference on Learning Representations (ICLR)},
  year = {2022},
  url = {https://iclr.cc/virtual/2022/poster/6746},
}

@inproceedings{aaai2024_online_6852,
  title = {{Online Sensitivity Optimization in Differentially Private Learning}},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2024},
  url = {https://ojs.aaai.org/index.php/AAAI/article/view/29099},
}

@inproceedings{icml2023_differentially_41512,
  title = {{Differentially Private Optimization on Large Model at Small Cost}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2023},
  url = {https://icml.cc/virtual/2023/poster/24636},
}

@inproceedings{iclr2023_exploring_33183,
  title = {{Exploring the Limits of Differentially Private Deep Learning with Group-wise Clipping}},
  booktitle = {International Conference on Learning Representations (ICLR)},
  year = {2023},
  url = {https://iclr.cc/virtual/2023/poster/11123},
}

@inproceedings{neurips2024_scalable_70614,
  title = {{Scalable DP-SGD: Shuffling vs. Poisson Subsampling}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2024},
  url = {https://neurips.cc/virtual/2024/poster/96505},
}

@inproceedings{iclr2023_differentially_33032,
  title = {{Differentially Private Adaptive Optimization with Delayed Preconditioners}},
  booktitle = {International Conference on Learning Representations (ICLR)},
  year = {2023},
  url = {https://iclr.cc/virtual/2023/poster/11050},
}

@inproceedings{icml2022_private_41029,
  title = {{Private Adaptive Optimization with Side information}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2022},
  url = {https://icml.cc/virtual/2022/poster/17541},
}

@inproceedings{neurips2024_nearly_72979,
  title = {{Nearly Tight Black-Box Auditing of Differentially Private Machine Learning}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2024},
  url = {https://neurips.cc/virtual/2024/poster/94437},
}

@inproceedings{aaai2025_differentially_8736,
  title = {{Differentially Private Prototypes for Imbalanced Transfer Learning}},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2025},
  url = {https://ojs.aaai.org/index.php/AAAI/article/view/35395},
}

@inproceedings{icml2024_a_43486,
  title = {{A New Linear Scaling Rule for Private Adaptive Hyperparameter Optimization}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2024},
  url = {https://icml.cc/virtual/2024/poster/34966},
}

@inproceedings{neurips2024_differentially_72305,
  title = {{Differentially Private Stochastic Gradient Descent with Fixed-Size Minibatches: Tighter RDP Guarantees with or without Replacement}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2024},
  url = {https://neurips.cc/virtual/2024/poster/95041},
}

@inproceedings{neurips2025_geoclip_76763,
  title = {{GeoClip: Geometry-Aware Clipping for Differentially Private SGD}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2025},
  url = {https://openreview.net/forum?id=MnT0QgNWFC},
}

@inproceedings{neurips2025_private_78110,
  title = {{Private Training Large-scale Models with Efficient DP-SGD}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2025},
  url = {https://openreview.net/forum?id=b6SWqFEOSF},
}

@inproceedings{iclr2025_the_81338,
  title = {{The Last Iterate Advantage: Empirical Auditing and Principled Heuristic Analysis of Differentially Private SGD}},
  booktitle = {International Conference on Learning Representations (ICLR)},
  year = {2025},
  url = {https://iclr.cc/virtual/2025/poster/30434},
}

@inproceedings{aaai2026_an_107158,
  title = {{An Improved Privacy and Utility Analysis of Differentially Private SGD with Bounded Domain and Smooth Losses}},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2026},
  url = {https://ojs.aaai.org/index.php/AAAI/article/view/39510},
}

@inproceedings{iclr2024_dp_33850,
  title = {{DP-SGD Without Clipping: The Lipschitz Neural Network Way}},
  booktitle = {International Conference on Learning Representations (ICLR)},
  year = {2024},
  url = {https://iclr.cc/virtual/2024/poster/19225},
}

@inproceedings{iclr2024_the_34629,
  title = {{The importance of feature preprocessing for differentially private linear optimization}},
  booktitle = {International Conference on Learning Representations (ICLR)},
  year = {2024},
  url = {https://iclr.cc/virtual/2024/poster/18445},
}

@inproceedings{iclr2024_differentially_35490,
  title = {{Differentially Private SGD Without Clipping Bias: An Error-Feedback Approach}},
  booktitle = {International Conference on Learning Representations (ICLR)},
  year = {2024},
  url = {https://iclr.cc/virtual/2024/poster/17583},
}

@inproceedings{icml2024_proactive_43318,
  title = {{Proactive DP: A Multiple Target Optimization Framework for DP-SGD}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2024},
  url = {https://icml.cc/virtual/2024/poster/35134},
}

@inproceedings{icml2024_how_45747,
  title = {{How Private are DP-SGD Implementations?}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2024},
  url = {https://icml.cc/virtual/2024/poster/32705},
}

@inproceedings{icml2026_understanding_51051,
  title = {{Understanding Private Learning From Feature Perspective}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2026},
  url = {https://openreview.net/forum?id=IP6W8i7XPG},
}

@inproceedings{icml2026_efficient_53323,
  title = {{Efficient DP-SGD for LLMs with Randomized Clipping}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2026},
  url = {https://openreview.net/forum?id=eTO6G25Jbo},
}

@inproceedings{icml2026_fiber_54964,
  title = {{FIBER: A Differentially Private Optimizer with Filter-Aware Innovation Bias Correction}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2026},
  url = {https://openreview.net/forum?id=uMdZHVXdOL},
}

@inproceedings{neurips2024_banded_71609,
  title = {{Banded Square Root Matrix Factorization for Differentially Private Model Training}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2024},
  url = {https://neurips.cc/virtual/2024/poster/95662},
}