Are BERT-based fact-checking models robust against adversarial attack?

Bachelor thesis (2023)

Authors

E.E. Afriat Electrical Engineering, Mathematics and Computer Science

Contributors

Avishek Anand Leibniz Universität (graduation committee member)

Lijun Lyu L3S (mentor)

L. Corti Web Information Systems - (mentor)

Faculty

Electrical Engineering, Mathematics and Computer Science, Electrical Engineering, Mathematics and Computer Science

Adversarial attacks NLP BERT Fact Checking

To reference this document use:

http://resolver.tudelft.nl/uuid:9b8e7e90-2487-4a52-8631-bd10e1537246

More Info

expand_more

Published Date

03-02-2023

Language

English

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Faculty

Electrical Engineering, Mathematics and Computer Science

Abstract

We seek to examine the vulnerability of BERT-based fact-checking. We implement a gradient based, adversarial attack strategy, based on Hotflip swapping individual tokens from the input. We use this on a pre-trained ExPred model for fact-checking. We find that gradient based adversarial attacks are ineffective against ExPred. Uncertainties about the similitude of the examples generated by our adversarial attack implementation cast doubts on the results.

Files

Main.pdf

(.pdf | 0.308 Mb)