Results 11 to 20 of about 3,053,264 (345)

Biased Humans, (Un)Biased Algorithms? [PDF]

Journal of Business Ethics, 2022
AbstractPrevious research has shown that algorithmic decisions can reflect gender bias. The increasingly widespread utilization of algorithms in critical decision-making domains (e.g., healthcare or hiring) can thus lead to broad and structural disadvantages for women.
Florian Pethig, Julia Kroenung
openaire   +3 more sources

From Pretraining Data to Language Models to Downstream Tasks: Tracking the Trails of Political Biases Leading to Unfair NLP Models [PDF]

Annual Meeting of the Association for Computational Linguistics, 2023
Language models (LMs) are pretrained on diverse data sources—news, discussion forums, books, online encyclopedias. A significant portion of this data includes facts and opinions which, on one hand, celebrate democracy and diversity of ideas, and on the ...
Shangbin Feng, Chan Young Park, Yuhan Liu, Yulia Tsvetkov   +3 more
semanticscholar   +1 more source

"Kelly is a Warm Person, Joseph is a Role Model": Gender Biases in LLM-Generated Reference Letters [PDF]

Conference on Empirical Methods in Natural Language Processing, 2023
Large Language Models (LLMs) have recently emerged as an effective tool to assist individuals in writing various types of content, including professional documents such as recommendation letters.
Yixin Wan, George Pu, Jiao Sun, Aparna Garimella, Kai-Wei Chang, Nanyun Peng   +5 more
semanticscholar   +1 more source

BOLD: Dataset and Metrics for Measuring Biases in Open-Ended Language Generation [PDF]

Conference on Fairness, Accountability and Transparency, 2021
Recent advances in deep learning techniques have enabled machines to generate cohesive open-ended text when prompted with a sequence of words as context.
J. Dhamala, Tony Sun, Varun Kumar, Satyapriya Krishna, Yada Pruksachatkun, Kai-Wei Chang, Rahul Gupta   +6 more
semanticscholar   +1 more source

Social Biases through the Text-to-Image Generation Lens [PDF]

AAAI/ACM Conference on AI, Ethics, and Society, 2023
Text-to-Image (T2I) generation is enabling new applications that support creators, designers, and general end users of productivity software by generating illustrative content with high photorealism starting from a given descriptive text as a prompt ...
Ranjita Naik, Besmira Nushi
semanticscholar   +1 more source

Graph Inductive Biases in Transformers without Message Passing [PDF]

International Conference on Machine Learning, 2023
Transformers for graph data are increasingly widely studied and successful in numerous learning tasks. Graph inductive biases are crucial for Graph Transformers, and previous works incorporate them using message-passing modules and/or positional ...
Liheng Ma, Chen Lin, Derek Lim, Adriana Romero-Soriano, P. Dokania, Mark Coates, Philip H. S. Torr, S. Lim   +7 more
semanticscholar   +1 more source

CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked Language Models [PDF]

Conference on Empirical Methods in Natural Language Processing, 2020
Pretrained language models, especially masked language models (MLMs) have seen success across many NLP tasks. However, there is ample evidence that they use the cultural biases that are undoubtedly present in the corpora they are trained on, implicitly ...
Nikita Nangia, Clara Vania, Rasika Bhalerao, Samuel R. Bowman   +3 more
semanticscholar   +1 more source

Generative language models exhibit social identity biases [PDF]

Nature Computational Science, 2023
Social identity biases, particularly the tendency to favor one’s own group (ingroup solidarity) and derogate other groups (outgroup hostility), are deeply rooted in human psychology and social behavior.
Tiancheng Hu, Y. Kyrychenko, Steven Rathje, Nigel Collier, S. V. D. Linden, J. Roozenbeek   +5 more
semanticscholar   +1 more source

Semantics derived automatically from language corpora contain human-like biases [PDF]

Science, 2016
Machines learn what people know implicitly AlphaGo has demonstrated that a machine can learn how to do things that people spend many years of concentrated study learning, and it can rapidly learn how to do them better than any human can.
Aylin Caliskan, J. Bryson, Arvind Narayanan   +2 more
semanticscholar   +1 more source

Developing biases [PDF]

Frontiers in Psychology, 2014
German nouns may alternate from singular to plural in two different ways. Some singular forms that end in a voiceless obstruent have a plural in which this obstruent is voiced. Another alternation concerns the vowel. Some singular forms with a back vowel have a plural form in which this back vowel is front.
van de Vijver, Ruben, Baer-Henney, Dinah
openaire   +3 more sources