What Makes a Good Example? Modeling Exemplar Selection with Neural Network Representations
This work addresses the computational principles of teaching exemplar selection, offering insights for dataset distillation in machine learning, though it is incremental in applying existing methods to a new problem.
The study modeled human exemplar selection for teaching using neural network feature representations and subset selection criteria, finding that strategies based on joint representativeness or its combination with diversity best captured human judgments, with transformer-based representations outperforming convolutional networks.
Teaching requires distilling a rich category distribution into a small set of informative exemplars. Although prior work shows that humans consider both representativeness and diversity when teaching, the computational principles underlying these tradeoffs remain unclear. We address this gap by modeling human exemplar selection using neural network feature representations and principled subset selection criteria. Novel visual categories were embedded along a one-dimensional morph continuum using pretrained vision models, and selection strategies varied in their emphasis on prototypicality, joint representativeness, and diversity. Adult participants selected one to three exemplars to teach a learner. Model-human comparisons revealed that strategies based on joint representativeness, or its combination with diversity, best captured human judgments, whereas purely prototypical or diversity-based strategies performed worse. Moreover, transformer-based representations consistently aligned more closely with human behavior than convolutional networks. These results highlight the potential utility of dataset distillation methods in machine learning as computational models for teaching.