Deep Classifier Mimicry without Data Access
This addresses a practical challenge for machine learning practitioners who need to update models when data access is restricted, offering a novel solution that is incremental in its approach.
The paper tackles the problem of fine-tuning or compressing pre-trained models without access to the original training data by proposing CAKE, a model-agnostic knowledge distillation method that uses noisy synthetic samples to mimic classifiers, achieving effectiveness across benchmark datasets and architectures.
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.