Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Lymperopoulos, Dimitris, Lymperaiou, Maria, Filandrianos, Giorgos, Stamou, Giorgos
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2408.01969
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913968786767872
author Lymperopoulos, Dimitris
Lymperaiou, Maria
Filandrianos, Giorgos
Stamou, Giorgos
author_facet Lymperopoulos, Dimitris
Lymperaiou, Maria
Filandrianos, Giorgos
Stamou, Giorgos
contents As NLP models become increasingly integral to decision-making processes, the need for explainability and interpretability has become paramount. In this work, we propose a framework that achieves the aforementioned by generating semantically edited inputs, known as counterfactual interventions, which change the model prediction, thus providing a form of counterfactual explanations for the model. We test our framework on two NLP tasks - binary sentiment classification and topic classification - and show that the generated edits are contrastive, fluent and minimal, while the whole process remains significantly faster that other state-of-the-art counterfactual editors.
format Preprint
id arxiv_https___arxiv_org_abs_2408_01969
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Optimal and efficient text counterfactuals using Graph Neural Networks
Lymperopoulos, Dimitris
Lymperaiou, Maria
Filandrianos, Giorgos
Stamou, Giorgos
Computation and Language
As NLP models become increasingly integral to decision-making processes, the need for explainability and interpretability has become paramount. In this work, we propose a framework that achieves the aforementioned by generating semantically edited inputs, known as counterfactual interventions, which change the model prediction, thus providing a form of counterfactual explanations for the model. We test our framework on two NLP tasks - binary sentiment classification and topic classification - and show that the generated edits are contrastive, fluent and minimal, while the whole process remains significantly faster that other state-of-the-art counterfactual editors.
title Optimal and efficient text counterfactuals using Graph Neural Networks
topic Computation and Language
url https://arxiv.org/abs/2408.01969