Exploring the Plausibility of Hate and Counter Speech Detectors with Explainable AI
Not on the map. Why?
Abstract (arXiv)
In this paper we investigate the explainability of transformer models and their plausibility for hate speech and counter speech detection. We compare representatives of four different explainability approaches, i.e., gradient-based, perturbation-based, attention-based, and prototype-based approaches, and analyze them quantitatively with an ablation study and qualitatively in a user study. Results show that perturbation-based explainability performs best, followed by gradient-based and attention-based explainability. Prototypebased experiments did not yield useful results. Overall, we observe that explainability strongly supports the users in better understanding the model predictions.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 23 of its 50 references that are in the map. First posted on arXiv 25 Jul 2024.
Builds on
23 of 50 references.
References · 23
- 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
- 2019BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehensioncited by 5,874
- 2016"Why Should I Trust You?": Explaining the Predictions of Any Classifiercited by 5,040
- 2017Axiomatic Attribution for Deep Networkscited by 3,350
- 2018Sanity Checks for Saliency Mapscited by 1,149
- 2020Towards Faithfully Interpretable NLP Systems: How should we define and evaluate faithfulness?cited by 590
- 2019ERASER: A Benchmark to Evaluate Rationalized NLP Modelscited by 523
- 2018Metrics for Explainable AI: Challenges and Prospectscited by 257
- 2022The Disagreement Problem in Explainable Machine Learning: A Practitioner's Perspectivecited by 156
- 2019A Benchmark Dataset for Learning to Intervene in Online Hate Speechcited by 128
- 2018Thou shalt not hate: Countering Online Hate Speechcited by 64
- 2022GlobEnc: Quantifying Global Token Attribution by Incorporating the Whole Encoder Layer in Transformerscited by 44
11 more
- 2020Racism is a Virus: Anti-Asian Hate and Counterspeech in Social Media during the COVID-19 Crisiscited by 40
- 2022Metrics for saliency map evaluation of deep learning explanation methodscited by 40
- 2022Hate Speech and Counter Speech Detection: Conversational Context Does Mattercited by 32
- 2020Countering hate on social media: Large scale classification of hate and counter speechcited by 22
- 2018Analyzing the hate and counter speech accounts on Twittercited by 21
- 2021On Explaining Your Explanations of BERT: An Empirical Study with Sequence Classificationcited by 21
- 2022ProtoTEx: Explaining Model Decisions with Prototype Tensorscited by 21
- 2019On the Explanation of Machine Learning Predictions in Clinical Gait Analysiscited by 15
- 2022Robust and Explainable Identification of Logical Fallacies in Natural Language Argumentscited by 12
- 2023Hate Cannot Drive out Hate: Forecasting Conversation Incivility following Replies to Hate Speechcited by 10
- 2022ELF22: A Context-based Counter Trolling Dataset to Combat Internet Trollscited by 7
Built on by
Citing papers per year: 2025: 1.
1 paper.