AISTATS 2024poster5 citations

Deep Classifier Mimicry without Data Access

Steven Braun, Martin Mundt, Kristian Kersting

Abstract

Access to pre-trained models has recently emerged as a standard across numerous machine learning domains. Unfortunately, access to the original data the models were trained on may not equally be granted. This makes it tremendously challenging to fine-tune, compress models, adapt continually, or to do any other type of data-driven update. We posit that original data access may however not be required. Specifically, we propose Contrastive Abductive Knowledge Extraction (CAKE), a model-agnostic knowledge distillation procedure that mimics deep classifiers without access to the original data. To this end, CAKE generates pairs of noisy synthetic samples and diffuses them contrastively toward a model’s decision boundary. We empirically corroborate CAKE’s effectiveness using several benchmark datasets and various architectural choices, paving the way for broad application.

BibTeX
@InProceedings{pmlr-v238-braun24b,
  title = 	 {Deep Classifier Mimicry without Data Access},
  author =       {Braun, Steven and Mundt, Martin and Kersting, Kristian},
  booktitle = 	 {Proceedings of The 27th International Conference on Artificial Intelligence and Statistics},
  pages = 	 {4762--4770},
  year = 	 {2024},
  editor = 	 {Dasgupta, Sanjoy and Mandt, Stephan and Li, Yingzhen},
  volume = 	 {238},
  series = 	 {Proceedings of Machine Learning Research},
  month = 	 {02--04 May},
  publisher =    {PMLR},
  pdf = 	 {https://proceedings.mlr.press/v238/braun24b/braun24b.pdf},
  url = 	 {https://proceedings.mlr.press/v238/braun24b.html},
  abstract = 	 {Access to pre-trained models has recently emerged as a standard across numerous machine learning domains. Unfortunately, access to the original data the models were trained on may not equally be granted. This makes it tremendously challenging to fine-tune, compress models, adapt continually, or to do any other type of data-driven update. We posit that original data access may however not be required. Specifically, we propose Contrastive Abductive Knowledge Extraction (CAKE), a model-agnostic knowledge distillation procedure that mimics deep classifiers without access to the original data. To this end, CAKE generates pairs of noisy synthetic samples and diffuses them contrastively toward a model’s decision boundary. We empirically corroborate CAKE’s effectiveness using several benchmark datasets and various architectural choices, paving the way for broad application.}
}
Deep Classifier Mimicry without Data Access · AISTATS 2024