Visualizing the Embedding Space to Explain the Effect of Knowledge Distillation

Recent research has found that knowledge distillation can be effective in reducing the size of a network and in increasing generalization. A pre-trained, large teacher network, for example, was shown to be able to bootstrap a student model that eventually outperforms the teacher in a limited label environment. Despite these advances, it still is relatively unclear why this method works, that is, what the resulting student model does ‘better’. To address this issue, here, we utilize two non-linear, low-dimensional embedding methods (t-SNE and IVIS) to visualize representation spaces of different layers in a network. We perform a set of extensive experiments with different architecture parameters and distillation methods. The resulting visualizations and metrics clearly show that distillation guides the network to find a more compact representation space for higher accuracy already in earlier layers compared to its non-distilled version.