🤝 Safety Research

168 papers · page 2 of 3 · 26–125 · the most cited first · all shelves

Safety Research is a subfield of Social Sciences in the OpenAlex classification of science. The topics most of its papers here are on: «Ethics and Social Impacts of AI», «Academic integrity and plagiarism», «Disability Rights and Representation».

📄 Papers

← prev 1 2 3 next →

26. Persistent Anti-Muslim Bias in Large Language ModelsAbubakar Abid, Maheen Farooqi, James Zou2021 · 640 citations
27. Evaluating Gender Bias in Machine TranslationGabriel Stanovsky, Noah A. Smith, Luke Zettlemoyer2019 · 588 citations
28. Measuring Bias in Contextualized Word RepresentationsKeita Kurita, Nidhi Vyas, Ayush Pareek, et al.2019 · 584 citations
29. Racial Bias in Hate Speech and Abusive Language Detection DatasetsThomas Davidson, Debasmita Bhattacharya, Ingmar Weber2019 · 577 citations
30. Examining Gender and Race Bias in Two Hundred Sentiment Analysis SystemsSvetlana Kiritchenko, Saif Mohammad2018 · 574 citations
31. Release Strategies and the Social Impacts of Language ModelsIrene Solaiman, Miles Brundage, Jack Clark, et al.2019 · 538 citations
32. The Social Impact of Natural Language ProcessingDirk Hovy, Shannon L. Spruit2016 · 516 citations
33. Learning Gender-Neutral Word EmbeddingsJieyu Zhao, Yichao Zhou, Zeyu Li, et al.2018 · 512 citations
34. Gender Bias in Contextualized Word EmbeddingsJieyu Zhao, Tianlu Wang, Mark Yatskar, et al.2019 · 503 citations
35. BBQ: A hand-built bias benchmark for question answeringAlicia Parrish, Angelica Chen, Nikita Nangia, et al.2022 · 487 citations
36. Null It Out: Guarding Protected Attributes by Iterative Nullspace ProjectionShauli Ravfogel, Yanai Elazar, Hila Gonen, et al.2020 · 480 citations
37. Reducing Gender Bias in Abusive Language DetectionJi Ho Park, Jamin Shin, Pascale Fung2018 · 464 citations
38. Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons LearnedDeep Ganguli, Liane Lovitt, Jackson Kernion, et al.2022 · 456 citations
39. Large Legal Fictions: Profiling Legal Hallucinations in Large Language ModelsMatthew Dahl, Varun Magesh, Mirac Suzgun, et al.2024 · 431 citations
40. Ethical and social risks of harm from Language ModelsLaura Weidinger, John Mellor, Maribeth Rauh, et al.2021 · 421 citations
41. Gender and Dialect Bias in YouTube's Automatic CaptionsRachael Tatman2017 · 413 citations
42. Lipstick on a Pig: Debiasing Methods Cover up Systematic Gender Biases in Word Embeddings But do not Remove ThemHila Gonen, Yoav Goldberg2019 · 412 citations
43. Social Biases in NLP Models as Barriers for Persons with DisabilitiesBen Hutchinson, Vinodkumar Prabhakaran, Emily Denton, et al.2020 · 402 citations
44. Gender Bias in Neural Natural Language ProcessingKaiji Lu, Piotr Mardziel, Fangjing Wu, et al.2018 · 394 citations
45. Whose Opinions Do Language Models Reflect?Shibani Santurkar, Esin Durmus, Faisal Ladhak, et al.2023 · 390 citations
46. Representation Engineering: A Top-Down Approach to AI TransparencyAndy Zou, Long Phan, Sarah Chen, et al.2023 · 386 citations
47. Gender and Representation Bias in GPT-3 Generated StoriesLi Lucy, David Bamman2021 · 382 citations
48. Aligning AI With Shared Human ValuesDan Hendrycks, Collin Burns, Steven Basart, et al.2020 · 382 citations
49. Green AIRoy Schwartz, Jesse Dodge, Noah A. Smith, et al.2019 · 370 citations
50. Annotators with Attitudes: How Annotator Beliefs And Identities Bias Toxic Language DetectionMaarten Sap, Swabha Swayamdipta, Laura Vianna, et al.2022 · 364 citations
51. Identifying and Reducing Gender Bias in Word-Level Language ModelsShikha Bordia, Samuel R. Bowman2019 · 363 citations
52. Adversarial Removal of Demographic Attributes from Text DataYanai Elazar, Yoav Goldberg2018 · 357 citations
53. Towards Understanding Sycophancy in Language ModelsMrinank Sharma, Meg Tong, Tomasz Korbak, et al.2023 · 354 citations
54. BeaverTails: Towards Improved Safety Alignment of LLM via a Human-Preference DatasetJiaming Ji, Mickel Liu, Juntao Dai, et al.2023 · 351 citations
55. Toxicity in chatgpt: Analyzing persona-assigned language modelsAmeet Deshpande, Vishvak Murahari, Tanmay Rajpurohit, et al.2023 · 344 citations
56. Black is to Criminal as Caucasian is to Police: Detecting and Removing Multiclass Bias in Word EmbeddingsThomas Manzini, Lim Yao Chong, Alan W Black, et al.2019 · 332 citations
57. Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!Xiangyu Qi, Yi Zeng, Tinghao Xie, et al.2023 · 331 citations
58. Gender Bias in Machine TranslationBeatrice Savoldi, Marco Gaido, Luisa Bentivogli, et al.2021 · 310 citations
59. From Pretraining Data to Language Models to Downstream Tasks: Tracking the Trails of Political Biases Leading to Unfair NLP ModelsShangbin Feng, Chan Young Park, Yuhan Liu, et al.2023 · 305 citations
60. Discovering Language Model Behaviors with Model-Written EvaluationsEthan Perez, Sam Ringer, Kamilė Lukošiūtė, et al.2022 · 305 citations
61. Towards Debiasing Fact Verification ModelsTal Schuster, Darsh Shah, Yun Jie Serene Yeo, et al.2019 · 296 citations
62. Assessing Cross-Cultural Alignment between ChatGPT and Human Societies: An Empirical StudyYong Cao, Li Zhou, Seolhwa Lee, et al.2023 · 278 citations
63. Having Beer after Prayer? Measuring Cultural Bias in Large Language ModelsTarek Naous, Michael J. Ryan, Alan Ritter, et al.2023 · 268 citations
64. Hallucination-Free? Assessing the Reliability of Leading AI Legal Research ToolsVarun Magesh, Faiz Surani, Matthew Dahl, et al.2024 · 267 citations
65. Jailbreaking ChatGPT via Prompt Engineering: An Empirical StudyYi Liu, Gelei Deng, Zhengzi Xu, et al.2023 · 266 citations
66. Queens are Powerful too: Mitigating Gender Bias in Dialogue GenerationEmily Dinan, Angela Fan, Adina Williams, et al.2020 · 265 citations
67. Evaluating the Underlying Gender Bias in Contextualized Word EmbeddingsChristine Basta, Marta R. Costa-jussà, Noe Casas2019 · 264 citations
68. Towards Understanding and Mitigating Social Biases in Language ModelsPaul Pu Liang, Chiyu Wu, Louis-Philippe Morency, et al.2021 · 263 citations
69. Towards Measuring the Representation of Subjective Global Opinions in Language ModelsEsin Durmus, Karina Nguyen, Thomas I. Liao, et al.2023 · 262 citations
70. Towards Debiasing Sentence RepresentationsPaul Pu Liang, Irene Mengze Li, Emily Zheng, et al.2020 · 259 citations
71. Decolonising Speech and Language TechnologySteven Bird2020 · 254 citations
72. Societal Biases in Language Generation: Progress and ChallengesEmily Sheng, Kai-Wei Chang, Prem Natarajan, et al.2021 · 252 citations
73. Measuring and Reducing Gendered Correlations in Pre-trained ModelsKellie Webster, Xuezhi Wang, Ian Tenney, et al.2020 · 252 citations
74. Interpreting Pretrained Contextualized Representations via Reductions to Static EmbeddingsRishi Bommasani, Kelly Davis, Claire Cardie2020 · 248 citations
75. DecodingTrust: A Comprehensive Assessment of Trustworthiness in GPT ModelsBoxin Wang, Weixin Chen, Hengzhi Pei, et al.2023 · 246 citations
76. Red teaming ChatGPT via Jailbreaking: Bias, Robustness, Reliability and ToxicityTerry Yue Zhuo, Yujin Huang, Chunyang Chen, et al.2023 · 243 citations
77. RAFT: Reward rAnked FineTuning for Generative Foundation Model AlignmentHanze Dong, Wei Xiong, Deepanshu Goyal, et al.2023 · 239 citations
78. Marked Personas: Using Natural Language Prompts to Measure Stereotypes in Language ModelsMyra Cheng, Esin Durmus, Dan Jurafsky2023 · 236 citations
79. ChatGPT-4 Outperforms Experts and Crowd Workers in Annotating Political Twitter Messages with Zero-Shot LearningPetter Tornberg2023 · 230 citations
80. Harms of Gender Exclusivity and Challenges in Non-Binary Representation in Language TechnologiesSunipa Dev, Masoud Monajatipoor, Anaelia Ovalle, et al.2021 · 228 citations
81. TrustLLM: Trustworthiness in Large Language ModelsYue Huang, Lichao Sun, Haoran Wang, et al.2024 · 226 citations
82. Trustworthy LLMs: a Survey and Guideline for Evaluating Large Language Models' AlignmentYang Liu; Yuanshun Yao; Jean-Francois Ton; Xiaoying Zhang; Ruocheng Guo; Hao Cheng; Yegor Klochkov; Muhammad Faaiz Taufiq;, Hang Li2023 · 219 citations
83. Systematic Inequalities in Language Technology Performance across the World's LanguagesDamian Blasi, Antonios Anastasopoulos, Graham Neubig2022 · 214 citations
84. An Empirical Survey of the Effectiveness of Debiasing Techniques for Pre-trained Language ModelsNicholas Meade, Elinor Poole-Dayan, Siva Reddy2022 · 211 citations
85. Safe RLHF: Safe Reinforcement Learning from Human FeedbackJosef Dai, Xuehai Pan, Ruiyang Sun, et al.2023 · 207 citations
86. Self-Diagnosis and Self-Debiasing: A Proposal for Reducing Corpus-Based Bias in NLPTimo Schick, Sahana Udupa, Hinrich Schutze2021 · 207 citations
87. Prompting GPT-3 To Be ReliableChenglei Si, Zhe Gan, Zhengyuan Yang, et al.2022 · 206 citations
88. Challenges in Detoxifying Language ModelsJohannes Welbl, Amelia Glaese, Jonathan Uesato, et al.2021 · 206 citations
89. Contextualizing Hate Speech Classifiers with Post-hoc ExplanationBrendan Kennedy, Xisen Jin, Aida Mostafazadeh Davani, et al.2020 · 203 citations
90. How Johnny Can Persuade LLMs to Jailbreak Them: Rethinking Persuasion to Challenge AI Safety by Humanizing LLMsYi Zeng, Hongpeng Lin, Jingwen Zhang, et al.2024 · 202 citations
91. HateCheck: Functional Tests for Hate Speech Detection ModelsPaul Röttger, Bertie Vidgen, Dong Nguyen, et al.2021 · 202 citations
92. Predictive Biases in Natural Language Processing Models: A Conceptual Framework and OverviewDeven Shah, H. Andrew Schwartz, Dirk Hovy2019 · 201 citations
93. Large Language Model as Attributed Training Data Generator: A Tale of Diversity and BiasYue Yu, Yuchen Zhuang, Jieyu Zhang, et al.2023 · 198 citations
94. Auto-Debias: Debiasing Masked Language Models with Automated Biased PromptsYue Guo, Yi Yang, Ahmed Abbasi2022 · 191 citations
95. Challenges in Automated Debiasing for Toxic Language DetectionXuhui Zhou, Maarten Sap, Swabha Swayamdipta, et al.2021 · 190 citations
96. HONEST: Measuring Hurtful Sentence Completion in Language ModelsDebora Nozza, Federico Bianchi, Dirk Hovy2021 · 190 citations
97. SemEval-2023 Task 10: Explainable Detection of Online SexismHannah Kirk, Wenjie Yin, Bertie Vidgen, et al.2023 · 185 citations
98. Preference Ranking Optimization for Human AlignmentFeifan Song, Bowen Yu, Minghao Li, et al.2023 · 183 citations
99. On Second Thought, Let's Not Think Step by Step! Bias and Toxicity in Zero-Shot ReasoningOmar Shaikh, Hongxin Zhang, William Held, et al.2023 · 180 citations
100. Global MMLU: Understanding and Addressing Cultural and Linguistic Biases in Multilingual EvaluationShivalika Singh, Angelika Romanou, Clémentine Fourrier, et al.2025 · 179 citations
101. Investigating Cultural Alignment of Large Language ModelsBadr AlKhamissi, Muhammad ElNokrashy, Mai Alkhamissi, et al.2024 · 177 citations
102. Debiasing Pre-trained Contextualised EmbeddingsMasahiro Kaneko, Danushka Bollegala2021 · 174 citations
103. Low-Resource Languages Jailbreak GPT-4Zheng-Xin Yong, Cristina Menghini, Stephen H. Bach2023 · 168 citations
104. On Releasing Annotator-Level Labels and Information in DatasetsVinodkumar Prabhakaran, Aida Mostafazadeh Davani, Mark Diaz2021 · 168 citations
105. Social Bias in Elicited Natural Language InferencesRachel Rudinger, Chandler May, Benjamin Van Durme2017 · 168 citations
106. The Capacity for Moral Self-Correction in Large Language ModelsDeep Ganguli, Amanda Askell, Nicholas Schiefer, et al.2023 · 167 citations
107. Speak, Memory: An Archaeology of Books Known to ChatGPT/GPT-4Kent K. Chang, Mackenzie Cramer, Sandeep Soni, et al.2023 · 166 citations
108. Towards Controllable Biases in Language GenerationEmily Sheng, Kai-Wei Chang, Prem Natarajan, et al.2020 · 166 citations
109. “I'm sorry to hear that”: Finding New Biases in Language Models with a Holistic Descriptor DatasetEric Michael Smith, Melissa Hall, Melanie Kambadur, et al.2022 · 165 citations
110. Sustainable Modular Debiasing of Language ModelsAnne Lauscher, Tobias Lueken, Goran Glavaš2021 · 165 citations
111. Do All Languages Cost the Same? Tokenization in the Era of Commercial Language ModelsOrevaoghene Ahia, Sachin Kumar, Hila Gonen, et al.2023 · 164 citations
112. GPT-4 Is Too Smart To Be Safe: Stealthy Chat with LLMs via CipherYouliang Yuan, Wenxiang Jiao, Wenxuan Wang, et al.2023 · 158 citations
113. WildGuard: Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMsSeungju Han, Kavel Rao, Allyson Ettinger, et al.2024 · 155 citations
114. Probing Pre-Trained Language Models for Cross-Cultural Differences in ValuesArnav Arora, Lucie-aimée Kaffee, Isabelle Augenstein2023 · 155 citations
115. ProsocialDialog: A Prosocial Backbone for Conversational AgentsHyunwoo Kim, Youngjae Yu, Liwei Jiang, et al.2022 · 149 citations
116. UNQOVERing Stereotyping Biases via Underspecified QuestionsTao Li, Daniel Khashabi, Tushar Khot, et al.2020 · 144 citations
117. Justice or Prejudice? Quantifying Biases in LLM-as-a-JudgeJiayi Ye, Yanbo Wang, Yue Huang, et al.2024 · 142 citations
118. A Roadmap to Pluralistic AlignmentTaylor Sorensen, Jared Moore, Jillian Fisher, et al.2024 · 135 citations
119. SafetyBench: Evaluating the Safety of Large Language ModelsZhexin Zhang, Leqi Lei, Lindong Wu, et al.2023 · 135 citations
120. Political Compass or Spinning Arrow? Towards More Meaningful Evaluations for Values and Opinions in Large Language ModelsPaul Röttger, Valentin Hofmann, Valentina Pyatkin, et al.2024 · 134 citations
121. TOFU: A Task of Fictitious Unlearning for LLMsPratyush Maini, Zhili Feng, Avi Schwarzschild, et al.2024 · 133 citations
122. Toward Gender-Inclusive Coreference ResolutionYang Trista Cao, Hal Daumé III2020 · 125 citations
123. Safety-Tuned LLaMAs: Lessons From Improving the Safety of Large Language Models that Follow InstructionsFederico Bianchi, Mirac Suzgun, Giuseppe Attanasio, et al.2023 · 124 citations
124. Large Language Model Alignment: A SurveyTianhao Shen, Renren Jin, Yufei Huang, et al.2023 · 123 citations
125. Intrinsic Bias Metrics Do Not Correlate with Application BiasSeraphina Goldfarb-Tarrant, Rebecca Marchant, Ricardo Muñoz Sánchez, et al.2021 · 116 citations

← prev 1 2 3 next →

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/shelves/social-sciences/safety-research/page/2/