Aylin CaliskanMy first name is pronounced as /aɪˈlin/
Assistant Professor |
|
@aylin_cim
I am a computer scientist with additional formal training in information systems and robotics. My research focuses on the societal impact of artificial intelligence (AI), particularly on empirical AI ethics in natural language processing, multimodal machine learning, and human-AI collaboration. Since 2016, I have made foundational contributions to understanding fairness in natural language processing and machine learning. I analyze the underpinning mechanisms of information transfer between human society and AI. To investigate the reasoning behind AI representations and outputs, I develop evaluation methods and transparency enhancing approaches that detect, quantify, and characterize human-like associations and biases learned by machines. I study human-AI bias interaction to uncover how machines that automatically learn implicit associations impact humans and society. As AI is co-evolving with society, my goal is to advance the responsible development and deployment of AI by integrating societal considerations to benefit everyone.
News |
|
source code Aylin Caliskan, Joanna J. Bryson, and Arvind Narayanan. A Story of Discrimination and Unfairness: Implicit Bias Embedded in Language Models. 9th Hot Topics in Privacy Enhancing Technologies
Accepted on 5/20/2016 |
Research |
|
Directionality and Representativeness are Differentiable Components of Stereotypes in Large Language Models PNAS Nexus, 2024 A Taxonomy of Stereotype Content in Large Language Models arXiv 2024 Gender, Race, and Intersectional Bias in Resume Screening via Language Model Retrieval AAAI/ACM Conference on Artificial Intelligence, Ethics, and Society (AAAI/ACM AIES 2024) "I don't see myself represented here at all": User Experiences of Stable Diffusion Outputs Containing Representational Harms across Gender Identities and Nationalities AAAI/ACM Conference on Artificial Intelligence, Ethics, and Society (AAAI/ACM AIES 2024) Do Generative AI Models Output Harm while Representing Non-Western Cultures: Evidence from A Community-Centered Approach AAAI/ACM Conference on Artificial Intelligence, Ethics, and Society (AAAI/ACM AIES 2024) Breaking Bias, Building Bridges: Evaluation and Mitigation of Social Biases in LLMs via Contact Hypothesis AAAI/ACM Conference on Artificial Intelligence, Ethics, and Society (AAAI/ACM AIES 2024) Effective AI regulation requires understanding general-purpose AI Brookings 2024 ChatGPT as Research Scientist: Probing GPT's Capabilities as a Research Librarian, Research Ethicist, Data Generator and Data Predictor Proceedings of the National Academy of Sciences (PNAS 2024) BiasDora: Exploring Hidden Biased Associations in Vision-Language Models In Findings of the Association for Computational Linguistics: EMNLP 2024 Science communication with generative AI Nature Human Behaviour, 2024 Extracting intersectional stereotypes from embeddings: Developing and validating the Flexible Intersectional Stereotype Extraction procedure PNAS Nexus, 2024 Global Gallery: The Fine Art of Painting Culture Portraits through Multilingual Instruction Tuning North American Chapter of the Association for Computational Linguistics (NAACL 2024) Safeguarding Human Values: Rethinking US Law for Generative AI's Societal Impacts AI and Ethics 2024 Label-Efficient Group Robustness via Out-of-Distribution Concept Curation Conference on Computer Vision and Pattern Recognition (CVPR 2024) Artificial Intelligence, Bias, and Ethics The 32nd International Joint Conference on Artificial Intelligence (IJCAI 2023) IJCAI Early Career Spotlight Paper Demographic Stereotypes in Text-to-Image Generation Stanford University Human-Centered Artificial Intelligence Policy Brief 2023 Pre-trained Speech Processing Models Contain Human-Like Biases that Propagate to Speech Emotion Recognition In Findings of the Association for Computational Linguistics: EMNLP 2023 'Person' == Light-skinned, Western Man, and Sexualization of Women of Color: Stereotypes in Stable Diffusion In Findings of the Association for Computational Linguistics: EMNLP 2023 ChatGPT Perpetuates Gender Bias in Machine Translation and Ignores Non-Gendered Pronouns: Findings across Bengali and Five other Low-Resource Languages AAAI/ACM Conference on Artificial Intelligence, Ethics, and Society (AAAI/ACM AIES 2023) Evaluating Biased Attitude Associations of Language Models in an Intersectional Context AAAI/ACM Conference on Artificial Intelligence, Ethics, and Society (AAAI/ACM AIES 2023) Contrastive Language-Vision AI Models Pretrained on Web-Scraped Multimodal Data Exhibit Sexual Objectification Bias The 2023 ACM Conference on Fairness, Accountability, and Transparency (ACM FAccT 2023) Easily Accessible Text-to-Image Generation Amplifies Demographic Stereotypes at Large Scale The 2023 ACM Conference on Fairness, Accountability, and Transparency (ACM FAccT 2023) Bias Against 93 Stigmatized Groups in Masked Language Models and Downstream Sentiment Classification Tasks The 2023 ACM Conference on Fairness, Accountability, and Transparency (ACM FAccT 2023) Envisioning Legal Mitigations for Intentional and Unintentional Harms Associated with Large Language Models (Extended Abstract) Fortieth International Conference on Machine Learning Workshop on Generative AI and Law (ICML GenLaw 2023) Regularizing Model Gradients with Concepts to Improve Robustness to Spurious Correlations (Poster) Fortieth International Conference on Machine Learning Workshop on Spurious Correlations, Invariance, and Stability (ICML SCIS 2023) Managing the risks of inevitably biased visual artificial intelligence systems Brookings 2022 Historical Representations of Social Groups Across 200 Years of Word Embeddings from Google Books Proceedings of the National Academy of Sciences (PNAS 2022) Gender Bias in Word Embeddings: A Comprehensive Analysis of Frequency, Syntax, and Semantics AAAI/ACM Conference on Artificial Intelligence, Ethics, and Society (AAAI/ACM AIES 2022) American == White in Multimodal Language-and-Image AI AAAI/ACM Conference on Artificial Intelligence, Ethics, and Society (AAAI/ACM AIES 2022) Measuring Gender Bias in Word Embeddings of Gendered Languages Requires Disentangling Grammatical Gender Signals AAAI/ACM Conference on Artificial Intelligence, Ethics, and Society (AAAI/ACM AIES 2022) Evidence for Hypodescent in Visual Semantic AI The 2022 ACM Conference on Fairness, Accountability, and Transparency (ACM FAccT 2022) Markedness in Visual Semantic AI The 2022 ACM Conference on Fairness, Accountability, and Transparency (ACM FAccT 2022) Contrastive Visual Semantic Pretraining Magnifies the Semantics of Natural Language Representations 60th Annual Meeting of the Association for Computational Linguistics (ACL 2022) VAST: The Valence-Assessing Semantics Test for Contextualizing Language Models Thirty-Sixth AAAI Conference on Artificial Intelligence (AAAI 2022) Detecting Emerging Associations and Behaviors With Regional and Diachronic Word Embeddings 16th IEEE International Conference on Semantic Computing (ICSC 2022) Learning to Behave: Improving Covert Channel Security with Behavior-Based Designs Privacy Enhancing Technologies Symposium (PETS 2022) Low Frequency Names Exhibit Bias and Overfitting in Contextualizing Language Models Empirical Methods in Natural Language Processing (EMNLP 2021) ValNorm Quantifies Semantics to Reveal Consistent Valence Biases Across Languages and Over Centuries Empirical Methods in Natural Language Processing (EMNLP 2021) Social biases in word embeddings and their relation to human cognition Book Chapter in The Handbook of Language Analysis in Psychology. Guilford Press, 2021 Editors Morteza Dehghani and Ryan Boyd Detecting and mitigating bias in natural language processing Brookings 2021 Disparate Impact of Artificial Intelligence Bias in Ridehailing Economy's Price Discrimination Algorithms AAAI/ACM Conference on Artificial Intelligence, Ethics, and Society (AAAI/ACM AIES 2021) Detecting Emergent Intersectional Biases: Contextualized Word Embeddings Contain a Distribution of Human-like Biases AAAI/ACM Conference on Artificial Intelligence, Ethics, and Society (AAAI/ACM AIES 2021) Image Representations Learned With Unsupervised Pre-Training Contain Human-like Biases The 2021 ACM Conference on Fairness, Accountability, and Transparency (ACM FAccT 2021) A Set of Distinct Facial Traits Learned by Machines Is Not Predictive of Appearance Bias in the Wild AI and Ethics, 2021 Automatically Characterizing Targeted Information Operations Through Biases Present in Discourse on Twitter 15th IEEE International Conference on Semantic Computing (ICSC 2021) |
Privacy | Terms