mlsanityData to 7 Oct 2026

ToKen: Task Decomposition and Knowledge Infusion for Few-Shot Hate Speech Detection

arXiv 2205.12495 ↗First posted 25 May 2022cs.CLcited by 15

On the map: LLM reasoning → Reasoning benchmarks and evaluation

Abstract (arXiv)

Hate speech detection is complex; it relies on commonsense reasoning, knowledge of stereotypes, and an understanding of social nuance that differs from one culture to the next. It is also difficult to collect a large-scale hate speech annotated dataset. In this work, we frame this problem as a few-shot learning task, and show significant gains with decomposing the task into its "constituent" parts. In addition, we see that infusing knowledge from reasoning datasets (e.g. Atomic2020) improves the performance even further. Moreover, we observe that the trained models generalize to out-of-distribution datasets, showing the superiority of task decomposition and knowledge infusion compared to previously used methods. Concretely, our method outperforms the baseline by 17.83% absolute gain in the 16-shot case.

Key facts. Cited by 15 papers in the map (data through 7 Oct 2026). Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning. Builds on 18 of its 31 references that are in the map. First posted on arXiv 25 May 2022.

Builds on

18 of 31 references.

References · 18

  1. 2019BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehensioncited by 5,874
  2. 2019HuggingFace's Transformers: State-of-the-art Natural Language Processingcited by 2,201
  3. 2020StereoSet: Measuring stereotypical bias in pretrained language modelscited by 911
  4. 2017Automated Hate Speech Detection and the Problem of Offensive Languagecited by 724
  5. 2020HateXplain: A Benchmark Dataset for Explainable Hate Speech Detectioncited by 403
  6. 2019Social Bias Frames: Reasoning about Social and Power Implications of Languagecited by 377
  7. 2019Predicting the Type and Target of Offensive Posts in Social Mediacited by 268
  8. 2018Large Scale Crowdsourcing and Characterization of Twitter Abusive Behaviorcited by 265
  9. 2020COMET-ATOMIC 2020: On Symbolic and Neural Commonsense Knowledge Graphscited by 254
  10. 2021Latent Hatred: A Benchmark for Understanding Implicit Hate Speechcited by 216
  11. 2019Racial Bias in Hate Speech and Abusive Language Detection Datasetscited by 208
  12. 2018Hate Speech Dataset from a White Supremacy Forumcited by 196
6 more
  1. 2022Jury Learning: Integrating Dissenting Voices into Machine Learning Modelscited by 115
  2. 2020Cross-lingual Zero- and Few-shot Hate Speech Detection Utilising Frozen Transformer Language Models and AXELcited by 23
  3. 2020Preserving Integrity in Online Social Networkscited by 14
  4. 2021An Information Retrieval Approach to Building Datasets for Hate Speech Detectioncited by 9
  5. 2020Annotating for Hate Speech: The MaNeCo Corpus and Some Input from Critical Discourse Analysiscited by 6
  6. 2022Meta AI at Arabic Hate Speech 2022: MultiTask Learning with Self-Correction for Hate Speech Classificationcited by 2

Built on by

Cited by year

Citing papers per year: 2022: 1; 2023: 8; 2024: 2; 2025: 4.

15 papers.

Sort:

By area:

  1. 2023Question Decomposition Improves the Faithfulness of Model-Generated Reasoningcited by 107
  2. 2023Facilitating Fine-grained Detection of Chinese Toxic Language: Hierarchical Taxonomy, Resources, and Benchmarkscited by 45
  3. 2023HARE: Explainable Hate Speech Detection with Step-by-Step Reasoningcited by 30
  4. 2023Moderating New Waves of Online Hate with Chain-of-Thought Reasoning in Large Language Modelscited by 17
  5. 2024Tox-BART: Leveraging Toxicity Attributes for Explanation Generation of Implicit Hate Speechcited by 11
  6. 2023Revisiting Hate Speech Benchmarks: From Data Curation to System Deploymentcited by 7
  7. 2023Towards Legally Enforceable Hate Speech Detection for Public Forumscited by 4
  8. 2022Hypothesis Engineering for Zero-Shot Hate Speech Detectioncited by 3
  9. 2023Hate Speech Detection via Dual Contrastive Learningcited by 3
  10. 2025Conditioning Large Language Models on Legal Systems? Detecting Punishable Hate Speechcited by 2
  11. 2025Fine-Grained Chinese Hate Speech Understanding: Span-Level Resources, Coded Term Lexicon, and Enhanced Detection Frameworkscited by 2
  12. 2024Alternative Speech: Complementary Method to Counter-Narrative for Better Discoursecited by 1
3 more
  1. 2023A Group-Specific Approach to NLP for Hate Speech Detectioncited by 0
  2. 2025U-GIFT: Uncertainty-Guided Firewall for Toxic Speech in Few-Shot Scenariocited by 0
  3. 2025Selective Demonstration Retrieval for Improved Implicit Hate Speech Detectioncited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025HatePrototypes: Interpretable and Transferable Representations for Implicit and Explicit Hate Speech Detection
  2. 2026X-MuTeST: A Multilingual Benchmark for Explainable Hate Speech Detection and A Novel LLM-consulted Explanation Framework
  3. 2025Reasoning-Aware Multimodal Fusion for Hateful Video Detection
  4. 2024Towards Efficient and Explainable Hate Speech Detection via Model Distillation
  5. 2023HateRephrase: Zero- and Few-Shot Reduction of Hate Intensity in Online Posts using Large Language Models
  6. 2025AmpleHate: Amplifying the Attention for Versatile Implicit Hate Detection
  7. 2025Cross-Modal Transfer from Memes to Videos: Addressing Data Scarcity in Hateful Video Detection
  8. 2025Rethinking Hate Speech Detection on Social Media: Can LLMs Replace Traditional Models?