Interpretable On-The-Fly Repair of Deep Neural Classifiers

Deep neural networks (DNNs) are vital in safety-critical systems but remain imperfect, leading to misclassification post-deployment. Prior works either make the model abstain from predicting in uncertain cases and so reduce its overall accuracy, or suffer from being uninterpretable. To overcome the limitations of prior work, we propose an interpretable approach to repair misclassifications after model deployment, instead of discarding them, by reducing the multi-classification problem into a simple binary classification. Our proposed technique specifically targets the predictions that the model is uncertain about them, extracts the training data that is positively and negatively incorporated into those uncertain decisions, and uses them to repair the cases where uncertainty leads to misclassification. We evaluate our approach on MNIST. The preliminary results show that our technique can repair 10.7% of the misclassifications on average, improving the performance of the models, and motivating the applicability of on-The-fly repair for more complex classifiers and different modalities.