Concept Evolution in Deep Learning Training: A Unified Interpretation Framework and Discoveries
This addresses a critical gap in DNN interpretation research for AI practitioners by providing a unified framework applicable to various architectures, though it is incremental as it builds on existing interpretation methods.
The paper tackles the problem of interpreting deep neural networks during training by introducing ConceptEvo, a framework that reveals the inception and evolution of learned concepts, and demonstrates its effectiveness through human evaluation and quantitative experiments, showing it identifies comprehensible and crucial concept evolutions.
We present ConceptEvo, a unified interpretation framework for deep neural networks (DNNs) that reveals the inception and evolution of learned concepts during training. Our work addresses a critical gap in DNN interpretation research, as existing methods primarily focus on post-training interpretation. ConceptEvo introduces two novel technical contributions: (1) an algorithm that generates a unified semantic space, enabling side-by-side comparison of different models during training, and (2) an algorithm that discovers and quantifies important concept evolutions for class predictions. Through a large-scale human evaluation and quantitative experiments, we demonstrate that ConceptEvo successfully identifies concept evolutions across different models, which are not only comprehensible to humans but also crucial for class predictions. ConceptEvo is applicable to both modern DNN architectures, such as ConvNeXt, and classic DNNs, such as VGGs and InceptionV3.