@inproceedings{hong-etal-2025-dp,
    title = "{DP}-{FROST}: Differentially Private Fine-tuning of Pre-trained Models with Freezing Model Parameters",
    author = "Hong, Daeyoung  and
      Jung, Woohwan  and
      Shim, Kyuseok",
    editor = "Rambow, Owen  and
      Wanner, Leo  and
      Apidianaki, Marianna  and
      Al-Khalifa, Hend  and
      Eugenio, Barbara Di  and
      Schockaert, Steven",
    booktitle = "Proceedings of the 31st International Conference on Computational Linguistics",
    month = jan,
    year = "2025",
    address = "Abu Dhabi, UAE",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/jlcl-multiple-ingestion/2025.coling-main.465/",
    pages = "6966--6984",
    abstract = "Training models with differential privacy has received a lot of attentions since differential privacy provides theoretical guarantee of privacy preservation. For a task in a specific domain, since a large-scale pre-trained model in the same domain contains general knowledge of the task, using such a model requires less effort in designing and training the model. However, differentially privately fine-tuning such models having a large number of trainable parameters results in large degradation of utility. Thus, we propose methods that effectively fine-tune the large-scale pre-trained models with freezing unimportant parameters for downstream tasks while satisfying differential privacy. To select the parameters to be fine-tuned, we propose several efficient methods based on the gradients of model parameters. We show the effectiveness of the proposed method by performing experiments with real datasets."
}