How good is your explanation? Algorithmic stability measures to assess the qualityof explanations for deep neural networks - IRT Saint Exupéry - Institut de Recherche Technologique Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2021

How good is your explanation? Algorithmic stability measures to assess the qualityof explanations for deep neural networks

Résumé

A plethora of methods have been proposed to explain howdeep neural networks reach a decision but comparativelylittle effort has been made to ensure that the explanationsproduced by these methods are objectively relevant. Whiledesirable properties for a good explanation are easy to come,objective measures have been harder to derive. Here, we pro-pose two new measures to evaluate explanations borrowedfrom the field of algorithmic stability: relative consistencyReCo and mean generalizability MeGe. We conduct severalexperiments on multiple image datasets and network archi-tectures to demonstrate the benefits of the proposed measuresover representative methods. We show that popular fidelitymeasures are not sufficient to guarantee good explanations.Finally, we show empirically that 1-Lipschitz networks pro-vide general and consistent explanations, regardless of theexplanation method used, making them a relevant directionfor explainability.
Fichier principal
Vignette du fichier
main.pdf (5.14 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-02930949 , version 1 (04-09-2020)
hal-02930949 , version 2 (01-07-2021)
hal-02930949 , version 3 (08-11-2021)

Identifiants

Citer

Thomas Fel, David Vigouroux, Rémi Cadène, Thomas Serre. How good is your explanation? Algorithmic stability measures to assess the qualityof explanations for deep neural networks. 2021. ⟨hal-02930949v2⟩
411 Consultations
324 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More