Data to 14 Oct 2026

Queens are Powerful too: Mitigating Gender Bias in Dialogue Generation

arXiv 1911.03842 ↗First posted 10 Nov 2019cs.CLcited by 152

Not on the map. Why?

Abstract (arXiv)

Models often easily learn biases present in the training data, and their predictions directly reflect this bias. We analyze gender bias in dialogue data, and examine how this bias is actually amplified in subsequent generative chit-chat dialogue models. We measure gender bias in six existing dialogue datasets, and focus on the most biased one, the multi-player text-based fantasy adventure dataset LIGHT, as a testbed for our bias mitigation techniques. The LIGHT dataset is highly imbalanced with respect to gender, containing predominantly male characters, likely because it is entirely collected by crowdworkers and reflects common biases that exist in fantasy or medieval settings. We consider three techniques to mitigate gender bias: counterfactual data augmentation, targeted data collection, and bias controlled training. We show that our proposed techniques mitigate gender bias in LIGHT by balancing the genderedness of generated dialogue utterances and are particularly effective in combination. We quantify performance using various evaluation methods---such as quantity of gendered words, a dialogue safety classifier, and human studies---all of which show that our models generate less gendered, but equally engaging chit-chat responses.

Key facts. Cited by 152 papers in the map (data through 14 Oct 2026). First posted on arXiv 10 Nov 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 33; 2022: 42; 2023: 27; 2024: 29; 2025: 10; 2026: 11.

152 papers.

Sort:

By area:

2022Training language models to follow instructions with human feedbackcited by 15,971
2022OPT: Open Pre-trained Transformer Language Modelscited by 3,512
2021TruthfulQA: Measuring How Models Mimic Human Falsehoodscited by 2,975
2022LaMDA: Language Models for Dialog Applicationscited by 1,220
2022Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learnedcited by 729
2023Bias and Fairness in Large Language Models: A Surveycited by 639
2023Toxicity in ChatGPT: Analyzing Persona-assigned Language Modelscited by 360
2021Towards Understanding and Mitigating Social Biases in Language Modelscited by 296
2021Societal Biases in Language Generation: Progress and Challengescited by 202
2023"Kelly is a Warm Person, Joseph is a Role Model": Gender Biases in LLM-Generated Reference Letterscited by 199
2022BlenderBot 3: a deployed conversational agent that continually learns to responsibly engagecited by 198
2022SODA: Million-scale Dialogue Distillation with Social Commonsense Contextualizationcited by 168

38 more

2022"I'm sorry to hear that": Finding New Biases in Language Models with a Holistic Descriptor Datasetcited by 162
2021RedditBias: A Real-World Resource for Bias Evaluation and Debiasing of Conversational Language Modelscited by 149
2022A Survey of Controllable Text Generation using Transformer-based Pre-trained Language Modelscited by 129
2022Language Models that Seek for Knowledge: Modular Search & Generation for Dialogue and Prompt Completioncited by 102
2021An Empirical Survey of the Effectiveness of Debiasing Techniques for Pre-trained Language Modelscited by 97
2021Trustworthy AI: A Computational Perspectivecited by 93
2021Mitigating Language-Dependent Ethnic Bias in BERTcited by 87
2021On Measures of Biases and Harms in NLPcited by 81
2022Language Generation Models Can Cause Harm: So What Can We Do About It? An Actionable Surveycited by 80
2021Anticipating Safety Issues in E2E Conversational AI: Framework and Toolingcited by 78
2021NL-Augmenter: A Framework for Task-Sensitive Natural Language Augmentationcited by 73
2023Word Embeddings Are Steers for Language Modelscited by 70
2022A Systematic Study of Bias Amplificationcited by 65
2021Few-Shot Bot: Prompt-Based Learning for Dialogue Systemscited by 64
2022Theories of "Gender" in NLP Bias Researchcited by 62
2022Data Feedback Loops: Model-driven Amplification of Dataset Biasescited by 60
2024Self-Debiasing Large Language Models: Zero-Shot Recognition and Reduction of Stereotypescited by 59
2022AlexaTM 20B: Few-Shot Learning Using a Large-Scale Multilingual Seq2Seq Modelcited by 57
2021Just Say No: Analyzing the Stance of Neural Dialogue Generation in Offensive Contextscited by 56
2023"I'm fully who I am": Towards Centering Transgender and Non-Binary Voices to Measure Biases in Open Language Generationcited by 56
2023Multi-Party Chat: Conversational Agents in Group Settings with Humans and Modelscited by 54
2021Dynaboard: An Evaluation-As-A-Service Platform for Holistic Next-Generation Benchmarkingcited by 53
2023BiasAsker: Measuring the Bias in Conversational AI Systemcited by 53
2022Perturbation Augmentation for Fairer NLPcited by 52
2024Towards Human-centered Proactive Conversational Agentscited by 49
2022A Causal Lens for Controllable Text Generationcited by 45
2023Using In-Context Learning to Improve Dialogue Safetycited by 42
2022Mitigating Gender Bias in Distilled Language Models via Counterfactual Role Reversalcited by 40
2024Prompt Stealing Attacks Against Large Language Modelscited by 40
2023Are Personalized Stochastic Parrots More Dangerous? Evaluating Persona Biases in Dialogue Systemscited by 37
2024MBBQ: A Dataset for Cross-Lingual Comparison of Stereotypes in Generative LLMscited by 35
2022AugLy: Data Augmentations for Robustnesscited by 34
2021Revealing Persona Biases in Dialogue Systemscited by 33
2022Why So Toxic? Measuring and Triggering Toxic Behavior in Open-Domain Chatbotscited by 33
2021Uncovering Implicit Gender Bias in Narratives through Commonsense Inferencecited by 32
2022Aligning to Social Norms and Values in Interactive Narrativescited by 32
2021MMChat: Multi-Modal Chat Dataset on Social Mediacited by 30
2022Choose Your Lenses: Flaws in Gender Bias Evaluationcited by 29

50 most cited of 152 All 152 (JSON)