Integrated Directional Gradients: Feature Interaction Attribution for Neural NLP Models
Abstract
Sandipan Sikdar, Parantapa Bhattacharya, Kieran Heese. Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2021.
Authors 3
-
Sandipan Sikdar Aachen
RWTH Aachen University · University of Virginia
Affiliation as printed
RWTH Aachen University
University of Virginia
-
Parantapa Bhattacharya Aachen
RWTH Aachen University · University of Virginia
Affiliation as printed
University of Virginia
RWTH Aachen University
-
Kieran Heese Aachen
RWTH Aachen University · University of Virginia
Affiliation as printed
University of Virginia
RWTH Aachen University
Cited by 34 stored of 34
Cited by patents worldwide 1 (Lens.org)
-
DIRECTIONAL DRIVERS OF DEEP LEARNING MODELS BASED ON MODEL GRADIENTSUS20240273353A1 2024-08-15 Pending