AutoProtoNet: Interpretability for Prototypical Networks

TL;DR

AutoProtoNet enhances interpretability in prototypical networks by enabling visualization and debugging of class representations in the embedding space, leading to improved accuracy and better understanding of the model.

Contribution

It introduces AutoProtoNet, a method that incorporates interpretability into prototypical networks through input reconstruction and prototype refinement techniques.

Findings

Prototype refinement improves validation accuracy.

Embedding space visualization aids understanding of class representations.

Debugging with AutoProtoNet enhances model performance on in-the-wild images.

Abstract

In meta-learning approaches, it is difficult for a practitioner to make sense of what kind of representations the model employs. Without this ability, it can be difficult to both understand what the model knows as well as to make meaningful corrections. To address these challenges, we introduce AutoProtoNet, which builds interpretability into Prototypical Networks by training an embedding space suitable for reconstructing inputs, while remaining convenient for few-shot learning. We demonstrate how points in this embedding space can be visualized and used to understand class representations. We also devise a prototype refinement method, which allows a human to debug inadequate classification parameters. We use this debugging technique on a custom classification task and find that it leads to accuracy improvements on a validation set consisting of in-the-wild images. We advocate for interpretability in meta-learning approaches and show that there are interactive ways for a human to enhance meta-learning algorithms.