AI Paper Atlas

Research map / Adversarial robustness and security / Attacks on text and graphs

TextShield: Beyond Successfully Detecting Adversarial Sentences in text classification

Lingfeng Shen, Ze Zhang, Haiyun Jiang, Ying Chen

ICLR 2023 · Adversarial robustness and security › Attacks on text and graphs · approach: Defenses against universal perturbations · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use TextShield instead of SEMBERT defense · IMDB adversarial examples (IGA, PWWS, GA, TextFooler)accuracy: 87.5 -> 88.4; 88.4 -> 88.6; 87.3 -> 90.5; 85.5 -> 86.5Improved
Use TextShield instead of SEMBERT defense · AGNews adversarial examples (IGA, PWWS, GA, TextFooler)accuracy: 88.5 -> 90.2; 88.5 -> 90.0; 88.5 -> 90.0; 88.5 -> 89.5Improved
Use TextShield instead of SEMBERT defense · Yahoo adversarial examples (IGA, PWWS, GA, TextFooler)accuracy: 66.8 -> 70.5; 66.8 -> 69.0; 66.4 -> 70.0; 62.0 -> 64.5Improved
Remove the VG sub-detector (ablation)TextShield · AGNews, TextCNN, TL/PWWS/GA/IGAaccuracy: -8.9; -7.9; -6.4; -7.0Hurt
Add the VG sub-detector (ablation)TextShield · AGNews, TextCNN, TL/PWWS/GA/IGAaccuracy: +8.9; +7.9; +6.4; +7.0Improved
Remove all sub-detectors (ablation)TextShield · AGNews, TextCNN, TL/PWWS/GA/IGAaccuracy: -34.8; -29.6; -40.3; -36.9Hurt
Add all four saliency sub-detectors (ablation)TextShield · AGNews, TextCNN, TL/PWWS/GA/IGAaccuracy: +34.8; +29.6; +40.3; +36.9Improved
Set corrector threshold β to 0.4 instead of a small or large value (ablation)TextShield corrector · AGNews, three victim classifiersaccuracyImproved
Increase the number of adversarial training sentences per class K (ablation)TextShield detector · TextShield defense configurationsaccuracyImproved
Increase K beyond 2000 adversarial sentences per class (ablation)TextShield detector · TextShield defense configurationsaccuracyNo clear effect

Later papers that compared with this method

1 later accepted paper reports TextShield as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy