Distill n' Explain: explaining graph neural networks using simple surrogates

Tamara Pereira, Erik Nascimento, Lucas E. Resck, Diego Mesquita, Amauri Souza

Tutkimustuotos: Artikkeli kirjassa/konferenssijulkaisussaConference article in proceedingsScientificvertaisarvioitu

7 Lataukset (Pure)

Abstrakti

Explaining node predictions in graph neural networks (GNNs) often boils down to finding graph substructures that preserve predictions. Finding these structures usually implies back-propagating through the GNN, bonding the complexity (e.g., number of layers) of the GNN to the cost of explaining it. This naturally begs the question: Can we break this bond by explaining a simpler surrogate GNN? To answer the question, we propose Distill n' Explain (DnX). First, DnX learns a surrogate GNN via knowledge distillation. Then, DnX extracts node or edge-level explanations by solving a simple convex program. We also propose FastDnX, a faster version of DnX that leverages the linear decomposition of our surrogate model. Experiments show that DnX and FastDnX often outperform state-of-the-art GNN explainers while being orders of magnitude faster. Additionally, we support our empirical findings with theoretical results linking the quality of the surrogate model (i.e., distillation error) to the faithfulness of explanations.

AlkuperäiskieliEnglanti
OtsikkoProceedings of The 26th International Conference on Artificial Intelligence and Statistics (AISTATS) 2023
ToimittajatFrancisco Ruiz, Jennifer Dy, Jan-Willem van de Meent
KustantajaJMLR
Sivut6199-6214
Sivumäärä16
TilaJulkaistu - 2023
OKM-julkaisutyyppiA4 Artikkeli konferenssijulkaisussa
TapahtumaInternational Conference on Artificial Intelligence and Statistics - Valencia, Espanja
Kesto: 25 huhtik. 202327 huhtik. 2023
Konferenssinumero: 26
http://aistats.org/aistats2023/

Julkaisusarja

NimiProceedings of Machine Learning Research
KustantajaJMLR
Vuosikerta206
ISSN (painettu)2640-3498

Conference

ConferenceInternational Conference on Artificial Intelligence and Statistics
LyhennettäAISTATS
Maa/AlueEspanja
KaupunkiValencia
Ajanjakso25/04/202327/04/2023
www-osoite

Sormenjälki

Sukella tutkimusaiheisiin 'Distill n' Explain: explaining graph neural networks using simple surrogates'. Ne muodostavat yhdessä ainutlaatuisen sormenjäljen.

Siteeraa tätä