Concept Embedding Models: Beyond the Accuracy-Explainability Trade-Off

TL;DR

This paper introduces Concept Embedding Models that improve trustworthiness and interpretability in AI systems by learning high-dimensional, meaningful concept representations, balancing accuracy, explanations, and interventions even with limited supervision.

Contribution

The paper proposes a new family of concept bottleneck models that learn interpretable high-dimensional concept embeddings, overcoming limitations of existing models in accuracy and robustness.

Findings

Achieve better or comparable accuracy to standard neural models without concepts

Capture meaningful semantic concepts beyond ground truth labels

Enable effective test-time concept interventions that improve accuracy

Abstract

Deploying AI-powered systems requires trustworthy models supporting effective human interactions, going beyond raw prediction accuracy. Concept bottleneck models promote trustworthiness by conditioning classification tasks on an intermediate level of human-like concepts. This enables human interventions which can correct mispredicted concepts to improve the model's performance. However, existing concept bottleneck models are unable to find optimal compromises between high task accuracy, robust concept-based explanations, and effective interventions on concepts -- particularly in real-world conditions where complete and accurate concept supervisions are scarce. To address this, we propose Concept Embedding Models, a novel family of concept bottleneck models which goes beyond the current accuracy-vs-interpretability trade-off by learning interpretable high-dimensional concept representations. Our experiments demonstrate that Concept Embedding Models (1) attain better or competitive task accuracy w.r.t. standard neural models without concepts, (2) provide concept representations capturing meaningful semantics including and beyond their ground truth labels, (3) support test-time concept interventions whose effect in test accuracy surpasses that in standard concept bottleneck models, and (4) scale to real-world conditions where complete concept supervisions are scarce.