mlsanityData to 7 Oct 2026

Walking in Others' Shoes: How Perspective-Taking Guides Large Language Models in Reducing Toxicity and Bias

arXiv 2407.15366 ↗First posted 22 Jul 2024cs.CL cs.AI cs.CYcited by 17

Not on the map. Why?

Abstract (arXiv)

The common toxicity and societal bias in contents generated by large language models (LLMs) necessitate strategies to reduce harm. Present solutions often demand white-box access to the model or substantial training, which is impractical for cutting-edge commercial LLMs. Moreover, prevailing prompting methods depend on external tool feedback and fail to simultaneously lessen toxicity and bias. Motivated by social psychology principles, we propose a novel strategy named textbfperspective-taking prompting (textscPeT) that inspires LLMs to integrate diverse human perspectives and self-regulate their responses. This self-correction mechanism can significantly diminish toxicity (up to 89%) and bias (up to 73%) in LLMs' responses. Rigorous evaluations and ablation studies are conducted on two commercial LLMs (ChatGPT and GLM) and three open-source LLMs, revealing textscPeT's superiority in producing less harmful responses, outperforming five strong baselines.

Key facts. Cited by 17 papers in the map (data through 7 Oct 2026). Builds on 67 of its 110 references that are in the map. First posted on arXiv 22 Jul 2024.

Builds on

67 of 110 references.

References · 67

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2022Training language models to follow instructions with human feedbackcited by 15,712
  3. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 12,865
  4. 2017A Unified Approach to Interpreting Model Predictionscited by 6,488
  5. 2019BERTScore: Evaluating Text Generation with BERTcited by 5,462
  6. 2022PaLM: Scaling Language Modeling with Pathwayscited by 5,304
  7. 2022Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedbackcited by 3,634
  8. 2023Self-Refine: Iterative Refinement with Self-Feedbackcited by 3,317
  9. 2022Constitutional AI: Harmlessness from AI Feedbackcited by 2,950
  10. 2019The Curious Case of Neural Text Degenerationcited by 2,881
  11. 2019A Survey on Bias and Fairness in Machine Learningcited by 1,911
  12. 2020Don't Stop Pretraining: Adapt Language Models to Domains and Taskscited by 1,548
38 more
  1. 2020RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Modelscited by 1,235
  2. 2015A Diversity-Promoting Objective Function for Neural Conversation Modelscited by 1,068
  3. 2021GLM: General Language Model Pretraining with Autoregressive Blank Infillingcited by 1,026
  4. 2023Large Language Models Cannot Self-Correct Reasoning Yetcited by 965
  5. 2021Ethical and social risks of harm from Language Modelscited by 957
  6. 2023CRITIC: Large Language Models Can Self-Correct with Tool-Interactive Critiquingcited by 751
  7. 2020How Much Knowledge Can You Pack Into the Parameters of a Language Model?cited by 713
  8. 2022Large Language Models Can Self-Improvecited by 656
  9. 2019The Woman Worked as a Babysitter: On Biases in Language Generationcited by 540
  10. 2020Artificial Intelligence, Values and Alignmentcited by 443
  11. 2021DExperts: Decoding-Time Controlled Text Generation with Experts and Anti-Expertscited by 434
  12. 2021BOLD: Dataset and Metrics for Measuring Biases in Open-Ended Language Generationcited by 410
  13. 2023Trustworthy LLMs: a Survey and Guideline for Evaluating Large Language Models' Alignmentcited by 394
  14. 2020GeDi: Generative Discriminator Guided Sequence Generationcited by 377
  15. 2021Self-Diagnosis and Self-Debiasing: A Proposal for Reducing Corpus-Based Bias in NLPcited by 336
  16. 2023Effective Long-Context Scaling of Foundation Modelscited by 301
  17. 2021Towards Understanding and Mitigating Social Biases in Language Modelscited by 294
  18. 2022Generating Sequences by Learning to Self-Correctcited by 271
  19. 2022Prompting GPT-3 To Be Reliablecited by 258
  20. 2023Large Language Model Alignment: A Surveycited by 241
  21. 2022Quark: Controllable Text Generation with Reinforced Unlearningcited by 236
  22. 2023Automatically Correcting Large Language Models: Surveying the landscape of diverse self-correction strategiescited by 236
  23. 2020Learning from the Worst: Dynamically Generated Datasets to Improve Online Hate Detectioncited by 217
  24. 2019Identifying and Reducing Gender Bias in Word-Level Language Modelscited by 214
  25. 2018Gender Bias in Neural Natural Language Processingcited by 210
  26. 2021Societal Biases in Language Generation: Progress and Challengescited by 201
  27. 2019Counterfactual Data Augmentation for Mitigating Gender Stereotypes in Languages with Rich Morphologycited by 200
  28. 2023The Capacity for Moral Self-Correction in Large Language Modelscited by 171
  29. 2020Deep Learning for Text Style Transfer: A Surveycited by 161
  30. 2022"I'm sorry to hear that": Finding New Biases in Language Models with a Holistic Descriptor Datasetcited by 160
  31. 2023SelfCheck: Using LLMs to Zero-Shot Check Their Own Step-by-Step Reasoningcited by 157
  32. 2019Reducing Sentiment Bias in Language Models via Counterfactual Evaluationcited by 148
  33. 2023Large Language Models Understand and Can be Enhanced by Emotional Stimulicited by 143
  34. 2019Does Gender Matter? Towards Fairness in Dialogue Systemscited by 89
  35. 2020Reducing Gender Bias in Neural Machine Translation as a Domain Adaptation Problemcited by 86
  36. 2022Controllable Natural Language Generation with Contrastive Prefixescited by 75
  37. 2022Exploring the Limits of Domain-Adaptive Training for Detoxifying Large-Scale Language Modelscited by 67
  38. 2023Unlocking Context Constraints of LLMs: Enhancing Context Efficiency of LLMs with Self-Information-Based Content Filteringcited by 62

50 most cited of 67 All 67 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 5; 2025: 9; 2026: 3.

17 papers.

Sort:

By area:

  1. 2025A Comprehensive Survey in LLM(-Agent) Full Stack Safety: Data, Training and Deploymentcited by 110
  2. 2024Understanding the Dark Side of LLMs' Intrinsic Self-Correctioncited by 42
  3. 2024Position: Theory of Mind Benchmarks are Broken for Large Language Modelscited by 30
  4. 2025Nuclear Deployed: Analyzing Catastrophic Risks in Decision-making of Autonomous LLM Agentscited by 23
  5. 2024Course-Correction: Safety Alignment Using Synthetic Preferencescited by 13
  6. 2024DebateQA: Evaluating Question Answering on Debatable Knowledgecited by 12
  7. 2025Exploring Multimodal Challenges in Toxic Chinese Detection: Taxonomy, Benchmark, and Findingscited by 10
  8. 2025Pragmatic Inference Chain (PIC) Improving LLMs' Reasoning of Authentic Implicit Toxic Languagecited by 8
  9. 2025AI Awarenesscited by 6
  10. 2024DiPT: Enhancing LLM reasoning through diversified perspective-takingcited by 4
  11. 2025Chinese Toxic Language Mitigation via Sentiment Polarity Consistent Rewritescited by 2
  12. 2025Something Just Like TRuST : Toxicity Recognition of Span and Targetcited by 2
5 more
  1. 2025REFER: Mitigating Bias in Opinion Summarisation via Frequency Framed Promptingcited by 2
  2. 2025The Social Cost of Intelligence: Emergence, Propagation, and Amplification of Stereotypical Bias in Multi-Agent Systemscited by 2
  3. 2026Robust Persona-Aware Toxicity Detection with Prompt Optimization and Learned Ensemblingcited by 0
  4. 2026DeFrame: Debiasing Large Language Models Against Framing Effectscited by 0
  5. 2026Debias It Yourself: Teaching LLMs Cognitive Bias Mitigation Interventionscited by 0