mlsanityData to 13 Oct 2026

A benchmark for toxic comment classification on Civil Comments dataset

arXiv 2301.11125 ↗First posted 26 Jan 2023cs.CL eess.AScited by 11

Not on the map. Why?

Abstract (arXiv)

Toxic comment detection on social media has proven to be essential for content moderation. This paper compares a wide set of different models on a highly skewed multi-label hate speech dataset. We consider inference time and several metrics to measure performance and bias in our comparison. We show that all BERTs have similar performance regardless of the size, optimizations or language used to pre-train the models. RNNs are much faster at inference than any of the BERT. BiLSTM remains a good compromise between performance and inference time. RoBERTa with Focal Loss offers the best performance on biases and AUROC. However, DistilBERT combines both good AUROC and a low inference time. All models are affected by the bias of associating identities. BERT, RNN, and XLNet are less sensitive than the CNN and Compact Convolutional Transformers.

Key facts. Cited by 11 papers in the map (data through 13 Oct 2026). Builds on 14 of its 20 references that are in the map. First posted on arXiv 26 Jan 2023.

Builds on

14 of 20 references.

References · 14

  1. 2017Attention Is All You Needcited by 60,106
  2. 2015Deep Residual Learning for Image Recognitioncited by 51,950
  3. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,304
  4. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,043
  5. 2015Distilling the Knowledge in a Neural Networkcited by 10,180
  6. 2019DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lightercited by 4,281
  7. 2019Unsupervised Cross-lingual Representation Learning at Scalecited by 3,916
  8. 2015Neural Machine Translation of Rare Words with Subword Unitscited by 2,910
  9. 2019XLNet: Generalized Autoregressive Pretraining for Language Understandingcited by 2,810
  10. 2019ALBERT: A Lite BERT for Self-supervised Learning of Language Representationscited by 2,779
  11. 2019Nuanced Metrics for Measuring Unintended Bias with Real Data for Text Classificationcited by 400
  12. 2020BERTweet: A pre-trained language model for English Tweetscited by 346
2 more
  1. 2021Escaping the Big Data Paradigm with Compact Transformerscited by 263
  2. 2020HateBERT: Retraining BERT for Abusive Language Detection in Englishcited by 206

Built on by

Cited by year (2026 to 13 Oct)

Citing papers per year: 2023: 2; 2024: 4; 2025: 3; 2026: 2.

11 papers.

  1. 2023Harnessing the Power of LLMs in Practice: A Survey on ChatGPT and Beyondcited by 411
  2. 2024When is Multicalibration Post-Processing Necessary?cited by 16
  3. 2024Language model developers should report train-test overlapcited by 16
  4. 2024Complexity Matters: Dynamics of Feature Learning in the Presence of Spurious Correlationscited by 12
  5. 2024BadFair: Backdoored Fairness Attacks with Group-conditioned Triggerscited by 4
  6. 2025Toxicity in Online Platforms and AI Systems: A Survey of Needs, Challenges, Mitigations, and Future Directionscited by 3
  7. 2026I Can't Believe It's Not Robust: Catastrophic Collapse of Safety Classifiers under Embedding Driftcited by 3
  8. 2025Probe-based Fine-tuning for Reducing Toxicitycited by 2
  9. 2023Measuring Misogyny in Natural Language Generation: Preliminary Results from a Case Study on two Reddit Communitiescited by 1
  10. 2025ASCenD-BDS: Adaptable, Stochastic and Context-aware framework for Detection of Bias, Discrimination and Stereotypingcited by 1
  11. 2026Rank-Constrained Adaptation for Reliable Real-World Performancecited by 0