Causally Reliable Concept Bottleneck Models

Pietro Barbiero (IBM Research) · Giovanni De Felice (Università della Svizzera Italiana) · Arianna Casanova Flores (Universität Liechtenstein) · Francesco De Santis (Polytechnic Institute of Turin) · Silvia Santini (Universita della Svizzera Italiana) · Johannes Schneider (University of Liechtenstein) · Alberto Termine (SUPSI - University of Applied Sciences Southern Switzerland)

bottleneck of conceptscausal mechanismscausal reasoningcausally reliablecausally reliable concept bottleneck modelsconcept-based models deep learning architectures explainability fairness constraintshuman-interpretable variablesinterpretable models observational data out-of-distribution generalizationresponsiveness to interventionsscientific literatureunstructured background knowledge

Concept-based models are an emerging paradigm in deep learning that constrains the inference process to operate through human-interpretable variables, facilitating explainability and human interaction. However, these architectures, on par with popular opaque neural models, fail to account for the true causal mechanisms underlying the target phenomena represented in the data. This hampers their ability to support causal reasoning tasks, limits out-of-distribution generalization, and hinders the implementation of fairness constraints. To overcome these issues, we propose Causally reliable Concept Bottleneck Models (C$^2$BMs), a class of concept-based architectures that enforce reasoning through a bottleneck of concepts structured according to a model of the real-world causal mechanisms. We also introduce a pipeline to automatically learn this structure from observational data and unstructured background knowledge (e.g., scientific literature). Experimental evidence suggests that C$^2$BMs are more interpretable, causally reliable, and improve responsiveness to interventions w.r.t. standard opaque and concept-based models, while maintaining their accuracy.