mlsanityData to 7 Oct 2026

Societal Biases in Language Generation: Progress and Challenges

arXiv 2105.04054 ↗First posted 10 May 2021cs.CLcited by 201

Not on the map. Why?

Abstract (arXiv)

Technology for language generation has advanced rapidly, spurred by advancements in pre-training large models on massive amounts of data and the need for intelligent agents to communicate in a natural manner. While techniques can effectively generate fluent text, they can also produce undesirable societal biases that can have a disproportionately negative impact on marginalized populations. Language generation presents unique challenges for biases in terms of direct user interaction and the structure of decoding techniques. To better understand these challenges, we present a survey on societal biases in language generation, focusing on how data and techniques contribute to biases and progress towards reducing biases. Motivated by a lack of studies on biases from decoding techniques, we also conduct experiments to quantify the effects of these techniques. By further discussing general trends and open challenges, we call to attention promising directions for research and the importance of fairness and inclusivity considerations for language generation applications.

Key facts. Cited by 201 papers in the map (data through 7 Oct 2026). Builds on 73 of its 106 references that are in the map. First posted on arXiv 10 May 2021.

Builds on

73 of 106 references.

References · 73

  1. 2017Attention Is All You Needcited by 59,932
  2. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  3. 2020Language Models are Few-Shot Learnerscited by 32,199
  4. 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,491
  5. 2019The Curious Case of Neural Text Degenerationcited by 2,881
  6. 2019XLNet: Generalized Autoregressive Pretraining for Language Understandingcited by 2,801
  7. 2020Extracting Training Data from Large Language Modelscited by 2,144
  8. 2016Equality of Opportunity in Supervised Learningcited by 2,000
  9. 2019Transformer-XL: Attentive Language Models Beyond a Fixed-Length Contextcited by 1,830
  10. 2018Datasheets for Datasetscited by 1,371
  11. 2018Hierarchical Neural Story Generationcited by 1,182
  12. 2020Language (Technology) is Power: A Critical Survey of "Bias" in NLPcited by 951
38 more
  1. 2019DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generationcited by 842
  2. 2019Plug and Play Language Models: A Simple Approach to Controlled Text Generationcited by 793
  3. 2019Universal Adversarial Triggers for Attacking and Analyzing NLPcited by 709
  4. 2019The Woman Worked as a Babysitter: On Biases in Language Generationcited by 540
  5. 2019Release Strategies and the Social Impacts of Language Modelscited by 505
  6. 2021DExperts: Decoding-Time Controlled Text Generation with Experts and Anti-Expertscited by 434
  7. 2021BOLD: Dataset and Metrics for Measuring Biases in Open-Ended Language Generationcited by 410
  8. 2020GeDi: Generative Discriminator Guided Sequence Generationcited by 377
  9. 2021Persistent Anti-Muslim Bias in Large Language Modelscited by 376
  10. 2018Fairness Without Demographics in Repeated Loss Minimizationcited by 363
  11. 2021Self-Diagnosis and Self-Debiasing: A Proposal for Reducing Corpus-Based Bias in NLPcited by 336
  12. 2019Mitigating Gender Bias in Natural Language Processing: Literature Reviewcited by 287
  13. 2020Data and its (dis)contents: A survey of dataset development and use in machine learning researchcited by 259
  14. 2019Evaluating Gender Bias in Machine Translationcited by 247
  15. 2018Examining Gender and Race Bias in Two Hundred Sentiment Analysis Systemscited by 234
  16. 2019Identifying and Reducing Gender Bias in Word-Level Language Modelscited by 214
  17. 2018Gender Bias in Neural Natural Language Processingcited by 210
  18. 2019Counterfactual Data Augmentation for Mitigating Gender Stereotypes in Languages with Rich Morphologycited by 200
  19. 2019BERT has a Mouth, and It Must Speak: BERT as a Markov Random Field Language Modelcited by 184
  20. 2018Learning Gender-Neutral Word Embeddingscited by 169
  21. 2021Understanding the Capabilities, Limitations, and Societal Impact of Large Language Modelscited by 159
  22. 2018Reducing Gender Bias in Abusive Language Detectioncited by 152
  23. 2019Queens are Powerful too: Mitigating Gender Bias in Dialogue Generationcited by 152
  24. 2019Reducing Sentiment Bias in Language Models via Counterfactual Evaluationcited by 148
  25. 2019Predictive Biases in Natural Language Processing Models: A Conceptual Framework and Overviewcited by 137
  26. 2018Assessing Gender Bias in Machine Translation -- A Case Study with Google Translatecited by 121
  27. 2021Gender Bias in Machine Translationcited by 116
  28. 2020Towards Controllable Biases in Language Generationcited by 112
  29. 2019Automatically Neutralizing Subjective Bias in Textcited by 105
  30. 2021Re-imagining Algorithmic Fairness in India and Beyondcited by 102
  31. 2019Getting Gender Right in Neural Machine Translationcited by 98
  32. 2019Does Gender Matter? Towards Fairness in Dialogue Systemscited by 89
  33. 2017Ethical Challenges in Data-Driven Dialogue Systemscited by 86
  34. 2020Reducing Gender Bias in Neural Machine Translation as a Domain Adaptation Problemcited by 86
  35. 2020Multi-Dimensional Gender Bias Classificationcited by 80
  36. 2019Equalizing Gender Biases in Neural Machine Translation with Word Embeddings Techniquescited by 65
  37. 2019On Measuring Gender Bias in Translation of Gender-neutral Pronounscited by 59
  38. 2020PowerTransformer: Unsupervised Controllable Revision for Biased Language Correctioncited by 59

50 most cited of 73 All 73 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 7; 2022: 39; 2023: 52; 2024: 42; 2025: 43; 2026: 18.

201 papers.

Sort:

By area:

  1. 2022PaLM: Scaling Language Modeling with Pathwayscited by 5,304
  2. 2023A Survey on Evaluation of Large Language Modelscited by 1,227
  3. 2021Ethical and social risks of harm from Language Modelscited by 957
  4. 2022GLM-130B: An Open Bilingual Pre-trained Modelcited by 837
  5. 2022Galactica: A Large Language Model for Sciencecited by 736
  6. 2023Safe RLHF: Safe Reinforcement Learning from Human Feedbackcited by 630
  7. 2023Bias and Fairness in Large Language Models: A Surveycited by 626
  8. 2022Using Large Language Models to Simulate Multiple Humans and Replicate Human Subject Studiescited by 506
  9. 2024From Generation to Judgment: Opportunities and Challenges of LLM-as-a-judgecited by 480
  10. 2023Towards Measuring the Representation of Subjective Global Opinions in Language Modelscited by 424
  11. 2023Can Large Language Models Transform Computational Social Science?cited by 245
  12. 2022Quark: Controllable Text Generation with Reinforced Unlearningcited by 236
38 more
  1. 2023PromptRobust: Towards Evaluating the Robustness of Large Language Models on Adversarial Promptscited by 218
  2. 2023"Kelly is a Warm Person, Joseph is a Role Model": Gender Biases in LLM-Generated Reference Letterscited by 196
  3. 2022SSD-LM: Semi-autoregressive Simplex-based Diffusion Language Model for Text Generation and Modular Controlcited by 191
  4. 2023Do All Languages Cost the Same? Tokenization in the Era of Commercial Language Modelscited by 186
  5. 2022"I'm sorry to hear that": Finding New Biases in Language Models with a Holistic Descriptor Datasetcited by 160
  6. 2023Having Beer after Prayer? Measuring Cultural Bias in Large Language Modelscited by 153
  7. 2024Investigating Cultural Alignment of Large Language Modelscited by 140
  8. 2021Harms of Gender Exclusivity and Challenges in Non-Binary Representation in Language Technologiescited by 128
  9. 2023Language Model Behavior: A Comprehensive Surveycited by 92
  10. 2024Measuring Political Bias in Large Language Models: What Is Said and How It Is Saidcited by 89
  11. 2021Quantifying Social Biases in NLP: A Generalization and Empirical Comparison of Extrinsic Fairness Metricscited by 83
  12. 2021On Measures of Biases and Harms in NLPcited by 81
  13. 2022Language Generation Models Can Cause Harm: So What Can We Do About It? An Actionable Surveycited by 80
  14. 2022FLUTE: Figurative Language Understanding through Textual Explanationscited by 78
  15. 2024Bias in Large Language Models: Origin, Evaluation, and Mitigationcited by 77
  16. 2025LLM Generated Persona is a Promise with a Catchcited by 77
  17. 2022COLD: A Benchmark for Chinese Offensive Language Detectioncited by 74
  18. 2022Gradient-Based Constrained Sampling from Language Modelscited by 64
  19. 2023TESS: Text-to-Text Self-Conditioned Simplex Diffusioncited by 63
  20. 2024Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systemscited by 63
  21. 2024Self-Debiasing Large Language Models: Zero-Shot Recognition and Reduction of Stereotypescited by 58
  22. 2022AlexaTM 20B: Few-Shot Learning Using a Large-Scale Multilingual Seq2Seq Modelcited by 57
  23. 2023"I'm fully who I am": Towards Centering Transgender and Non-Binary Voices to Measure Biases in Open Language Generationcited by 56
  24. 2022Re-contextualizing Fairness in NLP: The Case of Indiacited by 55
  25. 2022Characteristics of Harmful Text: Towards Rigorous Benchmarking of Language Modelscited by 49
  26. 2024MEGAnno+: A Human-LLM Collaborative Annotation Systemcited by 47
  27. 2023Inspecting the Geographical Representativeness of Images from Text-to-Image Modelscited by 44
  28. 2022Controllable Text Generation with Neurally-Decomposed Oraclecited by 41
  29. 2023Evaluating Evaluation Metrics: A Framework for Analyzing NLG Evaluation Metrics using Measurement Theorycited by 41
  30. 2022Mitigating Gender Bias in Distilled Language Models via Counterfactual Role Reversalcited by 40
  31. 2024AgentsCoDriver: Large Language Model Empowered Collaborative Driving with Lifelong Learningcited by 40
  32. 2023Are Personalized Stochastic Parrots More Dangerous? Evaluating Persona Biases in Dialogue Systemscited by 37
  33. 2024Leveraging Large Language Models for NLG Evaluation: Advances and Challengescited by 37
  34. 2025FairSteer: Inference Time Debiasing for LLMs with Dynamic Activation Steeringcited by 37
  35. 2025Neither Valid nor Reliable? Investigating the Use of LLMs as Judgescited by 37
  36. 2022Towards Identifying Social Bias in Dialog Systems: Frame, Datasets, and Benchmarkscited by 36
  37. 2023Evaluation of African American Language Bias in Natural Language Generationcited by 36
  38. 2022Deconstructing NLG Evaluation: Evaluation Practices, Assumptions, and Their Implicationscited by 32

50 most cited of 201 All 201 (JSON)