Concept Embedding Models - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

Concept Embedding Models

Résumé

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 highdimensional 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.
Fichier principal
Vignette du fichier
2209.09056.pdf (3.63 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03854550 , version 1 (15-11-2022)

Identifiants

Citer

Mateo Espinosa Zarlenga, Pietro Barbiero, Gabriele Ciravegna, Giuseppe Marra, Francesco Giannini, et al.. Concept Embedding Models. NeurIPS 2022 - 36th Conference on Neural Information Processing Systems, Nov 2022, New Orleans, United States. ⟨hal-03854550⟩
50 Consultations
17 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More