Measuring Bias in Contextualized Word Representations

2019/06/18 by Keita Kurita, Nidhi Vyas, Kurita, Keita +7 · 18 citations
Computer Science · #Computation and Language (cs.CL) #FOS: Computer and information sciences #Hate Speech and Cyberbullying Detection #Natural Language Processing Techniques #Topic Modeling

paper · pdf · doi:10.48550/arxiv.1906.07337

openalex publication_date 2019/06/18 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Contextual word embeddings such as BERT have achieved state of the art performance in numerous NLP tasks. Since they are optimized to capture the statistical properties of training data, they tend to pick up on and amplify social stereotypes present in the data as well. In this study, we (1)~propose a template-based method to quantify bias in BERT; (2)~show that this method obtains more consistent results in capturing social biases than the traditional cosine based method; and (3)~conduct a case study, evaluating gender bias in a downstream task of Gender Pronoun Resolution. Although our case study focuses on gender bias, the proposed technique is generalizable to unveiling other biases, including in multiclass settings, such as racial and religious biases.

Citations

• Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word\n Embeddings

Cited by

• Bias Testing and Mitigation in Black Box LLMs using Metamorphic Relations
• Exploring and Mitigating Gender Bias in Encoder-Based Transformer Models
• A word association network methodology for evaluating implicit biases in LLMs compared to humans
• Education Paradigm Shift To Maintain Human Competitive Advantage Over AI
• MEDEQUALQA: Evaluating Biases in LLMs with Counterfactual Reasoning
• Probing Social Identity Bias in Chinese LLMs with Gendered Pronouns and Social Groups
• BiasBusters: Uncovering and Mitigating Tool Selection Bias in Large Language Models
• Mechanistic Interpretability with SAEs: Probing Religion, Violence, and Geography in Large Language Models
• Simulating a Bias Mitigation Scenario in Large Language Models
• From Charts to Fair Narratives: Uncovering and Mitigating Geo-Economic Biases in Chart-to-Text
• Do Biased Models Have Biased Thoughts?
• FairLangProc: A Python package for fairness in NLP
• Investigating Gender Bias in LLM-Generated Stories via Psychological Stereotypes
• Explainable Mapper: Charting LLM Embedding Spaces Using Perturbation-Based Explanation and Verification Agents
• Taking a Stance on Fake News: Towards Automatic Disinformation Assessment via Deep Bidirectional Transformer Language Models for Stance Detection
• PRIDE -- Parameter-Efficient Reduction of Identity Discrimination for Equality in LLMs
• EsBBQ and CaBBQ: The Spanish and Catalan Bias Benchmarks for Question Answering
• Evaluating Metrics for Bias in Word Embeddings

Related

• Assessing Social and Intersectional Biases in Contextualized Word Representations
• Unmasking Contextual Stereotypes: Measuring and Mitigating BERT's Gender Bias
• Mitigating Gender Bias in Contextual Word Embeddings
• An Analysis of Social Biases Present in BERT Variants Across Multiple Languages
• Stereotype and Skew: Quantifying Gender Bias in Pre-trained and Fine-tuned Language Models