Data to 14 Oct 2026

On the Intrinsic and Extrinsic Fairness Evaluation Metrics for Contextualized Language Representations

arXiv 2203.13928 ↗First posted 25 Mar 2022cs.CLcited by 75

Not on the map. Why?

Abstract (arXiv)

Multiple metrics have been introduced to measure fairness in various natural language processing tasks. These metrics can be roughly categorized into two categories: 1) extrinsic metrics for evaluating fairness in downstream applications and 2) intrinsic metrics for estimating fairness in upstream contextualized language representation models. In this paper, we conduct an extensive correlation study between intrinsic and extrinsic metrics across bias notions using 19 contextualized language models. We find that intrinsic and extrinsic metrics do not necessarily correlate in their original setting, even when correcting for metric misalignments, noise in evaluation datasets, and confounding factors such as experiment configuration for extrinsic metrics. %al

Key facts. Cited by 75 papers in the map (data through 14 Oct 2026). Builds on 19 of its 26 references that are in the map. First posted on arXiv 25 Mar 2022.

Builds on

19 of 26 references.

References · 19

2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
2019DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lightercited by 4,307
2019XLNet: Generalized Autoregressive Pretraining for Language Understandingcited by 2,811
2019ALBERT: A Lite BERT for Self-supervised Learning of Language Representationscited by 2,790
2020StereoSet: Measuring stereotypical bias in pretrained language modelscited by 924
2020CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked Language Modelscited by 710
2021BOLD: Dataset and Metrics for Measuring Biases in Open-Ended Language Generationcited by 412
2020HateXplain: A Benchmark Dataset for Explainable Hate Speech Detectioncited by 407
2019Nuanced Metrics for Measuring Unintended Bias with Real Data for Text Classificationcited by 404
2019Social Bias Frames: Reasoning about Social and Power Implications of Languagecited by 380
2019Bias in Bios: A Case Study of Semantic Representation Bias in a High-Stakes Settingcited by 349
2019Measuring Bias in Contextualized Word Representationscited by 276

7 more

2019Gender Bias in Contextualized Word Embeddingscited by 203
2019BERT has a Mouth, and It Must Speak: BERT as a Markov Random Field Language Modelcited by 185
2020Detecting Emergent Intersectional Biases: Contextualized Word Embeddings Contain a Distribution of Human-like Biasescited by 167
2019Reducing Sentiment Bias in Language Models via Counterfactual Evaluationcited by 148
2020Intrinsic Bias Metrics Do Not Correlate with Application Biascited by 148
2020Demographics Should Not Be the Reason of Toxicity: Mitigating Discrimination in Text Classifications with Instance Weightingcited by 41
2020Fairness and Robustness in Invariant Learning: A Case Study in Toxicity Classificationcited by 32

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 14; 2023: 23; 2024: 20; 2025: 14; 2026: 4.

75 papers.

Sort:

By area:

2023Bias and Fairness in Large Language Models: A Surveycited by 639
2022Prompting GPT-3 To Be Reliablecited by 260
2023From Pretraining Data to Language Models to Downstream Tasks: Tracking the Trails of Political Biases Leading to Unfair NLP Modelscited by 260
2023"Kelly is a Warm Person, Joseph is a Role Model": Gender Biases in LLM-Generated Reference Letterscited by 199
2022On Second Thought, Let's Not Think Step by Step! Bias and Toxicity in Zero-Shot Reasoningcited by 194
2022"I'm sorry to hear that": Finding New Biases in Language Models with a Holistic Descriptor Datasetcited by 162
2023A Survey on Fairness in Large Language Modelscited by 95
2024A Survey on Multilingual Large Language Models: Corpora, Alignment, and Biascited by 84
2022Perturbation Augmentation for Fairer NLPcited by 52
2024Measuring Implicit Bias in Explicitly Unbiased Large Language Modelscited by 51
2022Debiasing isn't enough! -- On the Effectiveness of Debiasing MLMs and their Social Biases in Downstream Taskscited by 48
2024Bias in Language Models: Beyond Trick Tests and Toward RUTEd Evaluationcited by 41

38 more

2022Theory-Grounded Measurement of U.S. Social Stereotypes in English Language Modelscited by 35
2024Decoding Biases: Automated Methods and LLM Judges for Gender Bias Detection in Language Modelscited by 33
2023Overwriting Pretrained Bias with Finetuning Datacited by 32
2022The Tail Wagging the Dog: Dataset Construction Biases of Social Bias Benchmarkscited by 31
2022Choose Your Lenses: Flaws in Gender Bias Evaluationcited by 29
2023On the Independence of Association Bias and Empirical Fairness in Language Modelscited by 29
2023Sociodemographic Bias in Language Models: A Survey and Forward Pathcited by 28
2023FFT: Towards Harmlessness Evaluation and Analysis for LLMs with Factuality, Fairness, Toxicitycited by 25
2024Investigating Implicit Bias in Large Language Models: A Large-Scale Study of Over 50 LLMscited by 25
2022A Robust Bias Mitigation Procedure Based on the Stereotype Content Modelcited by 24
2023An Empirical Study of Metrics to Measure Representational Harms in Pre-Trained Language Modelscited by 24
2024Extrinsic Evaluation of Cultural Competence in Large Language Modelscited by 24
2025Fairness through Difference Awareness: Measuring Desired Group Discrimination in LLMscited by 20
2023Fairness in Language Models Beyond English: Gaps and Challengescited by 16
2023Evaluating Gender Bias of Pre-trained Language Models in Natural Language Inference by Considering All Labelscited by 16
2023Co2PT: Mitigating Bias in Pre-trained Language Models through Counterfactual Contrastive Prompt Tuningcited by 15
2024What the Harm? Quantifying the Tangible Impact of Gender Bias in Machine Translation with a Human-centered Studycited by 15
2022How Gender Debiasing Affects Internal Model Representations, and Why It Matterscited by 12
2023Learning to Generate Equitable Text in Dialogue from Biased Training Datacited by 12
2023Women Are Beautiful, Men Are Leaders: Gender Stereotypes in Machine Translation and Language Modelingcited by 12
2023Robust Concept Erasure via Kernelized Rate-Distortion Maximizationcited by 11
2024Tox-BART: Leveraging Toxicity Attributes for Explanation Generation of Implicit Hate Speechcited by 11
2024Certifying Counterfactual Bias in LLMscited by 10
2023This Prompt is Measuring <MASK>: Evaluating Bias Evaluation in Language Modelscited by 9
2023In-Depth Look at Word Filling Societal Bias Measurescited by 8
2023Mitigating Bias for Question Answering Models by Tracking Bias Influencecited by 8
2025Evaluating the Effect of Retrieval Augmentation on Social Biasescited by 7
2025Rethinking LLM Bias Probing Using Lessons from the Social Sciencescited by 7
2025Agree to Disagree? A Meta-Evaluation of LLM Misgenderingcited by 7
2022Gender Bias in Meta-Embeddingscited by 6
2023Evaluating Bias and Fairness in Gender-Neutral Pretrained Vision-and-Language Modelscited by 6
2024Investigating Annotator Bias in Large Language Models for Hate Speech Detectioncited by 6
2024Unveiling and Mitigating Bias in Large Language Model Recommendations: A Path to Fairnesscited by 6
2025Biases Propagate in Encoder-based Vision-Language Models: A Systematic Analysis From Intrinsic Measures to Zero-shot Retrieval Outcomescited by 6
2023On the Origins of Bias in NLP through the Lens of the Jim Codecited by 5
2024The Gaps between Pre-train and Downstream Settings in Bias Evaluation and Debiasingcited by 5
2022Undesirable Biases in NLP: Addressing Challenges of Measurementcited by 4
2024Why Don't Prompt-Based Fairness Metrics Correlate?cited by 4

50 most cited of 75 All 75 (JSON)