Mixture of Concept Bottleneck Experts
Polytechnic Institute of Turin · Intesa Sanpaolo Innovation Center · Università della Svizzera Italiana · Universität Liechtenstein · University of Pisa · Department of Information Engineering and Mathematical Sciences, University of Siena · University of Oxford · IBM Research · Politecnico di Torino
PDF 由论文原始站点提供，PaperCompass 不保存论文文件。
摘要
Concept Bottleneck Models (CBMs) promote interpretability by grounding predictions in human-understandable concepts. However, existing CBMs typically fix their task predictor to a single linear or Boolean expression, limiting both predictive accuracy and adaptability to diverse user needs. We propose Mixture of Concept Bottleneck Experts (M-CBEs), a framework that generalizes existing CBMs along two dimensions: the number of experts and the functional form of each expert, exposing an underexplored region of the design space. We investigate this region by instantiating two novel models: Linear M-CBE, which learns a finite set of linear expressions, and Symbolic M-CBE, which leverages symbolic regression to discover expert functions from data under user-specified operator vocabularies. Empirical evaluation demonstrates that varying the mixture size and functional form provides a robust framework for navigating the accuracy-interpretability trade-off, adapting to different user and task needs.